Health at A Glance 2015 PDF

Health at a Glance 2015
OECD INDICATORS
OECD INDICATORS
This work is published under the responsibility of the Secretary-General of the OECD. The
opinions expressed and arguments employed herein do not necessarily reflect the official
views of OECD member countries.
This document and any map included herein are without prejudice to the status of or
sovereignty over any territory, to the delimitation of international frontiers and boundaries
and to the name of any territory, city or area.
Please cite this publication as:

OECD (2015), Health at a Glance 2015: OECD Indicators, OECD Publishing, Paris.
http://dx.doi.org/10.1787/health_glance-2015-en
ISBN 978-92-64-23257-0 (print)

ISBN 978-92-64-24351-4 (PDF)
Series: Health at a Glance

ISSN 1995-3992 (print)
ISSN 1999-1312 (online)
The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use
of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli
settlements in the West Bank under the terms of international law.
Photo credits: Cover Oleksiy Mark/Shutterstock.com; Chapter 3: Comstock/Jupiterimages; Chapter 4: Comstock/

Jupiterimages; Chapter 5: Randy Faris/Corbis; Chapter 6: Vincent Hazat/Photo Alto; Chapter 7: onoky Fotolia.com;
Chapter 8: CREATAS/Jupiterimages; Chapter 9: Tetraimages/Inmagine; Chapter 10: RGtimeline/Shutterstock.com;
Chapter 11: Thinkstock/iStockphoto.com.
Corrigenda to OECD publications may be found on line at: www.oecd.org/about/publishing/corrigenda.htm.

OECD 2015
You can copy, download or print OECD content for your own use, and you can include excerpts from OECD publications, databases and
multimedia products in your own documents, presentations, blogs, websites and teaching materials, provided that suitable
acknowledgement of OECD as source and copyright owner is given. All requests for public or commercial use and translation rights should
be submitted to rights@oecd.org. Requests for permission to photocopy portions of this material for public or commercial use shall be
addressed directly to the Copyright Clearance Center (CCC) at info@copyright.com or the Centre franais dexploitation du droit de copie (CFC)
at contact@cfcopies.com.
FOREWORD
Foreword
T his 2015 edition of Health at a Glance OECD Indicators presents the most recent comparable
data on key indicators of health and health systems across the 34 OECD member countries. For a
subset of indicators, it also reports data for partner countries, including Brazil, China, Columbia,
Costa Rica, India, Indonesia, Latvia, Lithuania, the Russian Federation and South Africa. This edition
includes two new features: a set of dashboard indicators on health and health systems, presented in
Chapter 1, summarising the comparative performance of OECD countries, and a special chapter on
recent trends in pharmaceutical spending across OECD countries, presented in Chapter 2.
The production of Health at a Glance would not have been possible without the contribution
of OECD Health Data National Correspondents, Health Accounts Experts, and Health Care Quality
Indicators Experts from the 34 OECD countries. The OECD gratefully acknowledges their effort in
supplying most of the data contained in this publication. The OECD also acknowledges the
contribution of other international organisations, especially the World Health Organization and
Eurostat, for sharing some of the data presented here, and the European Commission for supporting
data development work.
This publication was prepared by a team from the OECD Health Division under the co-ordination
of Gatan Lafortune. Chapter 1 was prepared by Gatan Lafortune and Nelly Biondi; Chapter 2 by
Valrie Paris, Annalisa Belloni, David Morgan and Michael Mueller; Chapter 3 by Nelly Biondi and
Gatan Lafortune; Chapter 4 by Marion Devaux, Nelly Biondi and Franco Sassi; Chapter 5 by Gatan
Lafortune, Frdric Daniel, Liliane Moreira and Michael Gmeinder; Chapter 6 by Gatan Lafortune,
Frdric Daniel and Nelly Biondi; Chapter 7 by Gatan Lafortune, Marion Devaux, Michael Mueller,
Marie-Clmence Canaud, Frdric Daniel and Nelly Biondi; Chapter 8 by Ian Brownwood, Ian Forde,
Rie Fujisawa, Nelly Biondi, Emily Hewlett, Carol Nader, Luke Slawomirski and Niek Klazinga; Chapter 9
by David Morgan, Michael Mueller, Yuki Murakami and Michael Gmeinder; Chapter 10 by Valrie Paris,
Annalisa Belloni, David Morgan, Michael Mueller, Luke Slawomirski and Marie-Clmence Canaud;
Chapter 11 by Tim Muir, Yuki Murakami, Gatan Lafortune, Marie-Clmence Canaud and Nelly Biondi.
This publication benefited from useful comments from Francesca Colombo.
HEALTH AT A GLANCE 2015 OECD 2015 3

TABLE OF CONTENTS
Table of contents
Executive summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Readers guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Chapter 1. Dashboards of health indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Health status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Risk factors to health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Access to care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Quality of care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Health care resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Chapter 2. Pharmaceutical spending trends and future challenges . . . . . . . . . . . . . . . . 29

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
One in every five health dollars is spent on pharmaceuticals . . . . . . . . . . . . . . . . . . 30
The share of private funding of pharmaceuticals increases . . . . . . . . . . . . . . . . . . . 32
Pharmaceutical expenditure growth is driven by changes in quantity, prices
and therapeutic mix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Drivers of spending growth vary across therapeutic areas. . . . . . . . . . . . . . . . . . . . . 38
New challenges in the pharmaceutical market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Chapter 3. Health status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

Life expectancy at birth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Life expectancy by sex and education level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Mortality from cardiovascular diseases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Mortality from cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Mortality from transport accidents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Suicide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Infant mortality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Infant health: Low birth weight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Perceived health status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Cancer incidence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Chapter 4. Non-medical determinants of health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Tobacco consumption among adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Alcohol consumption among adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Fruit and vegetable consumption among adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Obesity among adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Overweight and obesity among children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

TABLE OF CONTENTS
Chapter 5. Health workforce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Doctors (overall number) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Doctors by age, sex and category . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Medical graduates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
International migration of doctors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Remuneration of doctors (general practitioners and specialists) . . . . . . . . . . . . . . . 88
Nurses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Nursing graduates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
International migration of nurses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Remuneration of nurses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Chapter 6. Health care activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

Consultations with doctors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Medical technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Hospital beds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Hospital discharges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Average length of stay in hospitals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Cardiac procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Hip and knee replacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Caesarean sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Ambulatory surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Chapter 7. Access to care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Coverage for health care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Unmet needs for medical care and dental care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Out-of-pocket medical expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Geographic distribution of doctors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Waiting times for elective surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Chapter 8. Quality of care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

Avoidable hospital admissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
Diabetes care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Prescribing in primary care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Mortality following acute myocardial infarction (AMI) . . . . . . . . . . . . . . . . . . . . . . . . 138
Mortality following stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Waiting times for hip fracture surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Surgical complications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Obstetric trauma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Care for people with mental health disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Screening, survival and mortality for cervical cancer . . . . . . . . . . . . . . . . . . . . . . . . 150
Screening, survival and mortality for breast cancer . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Survival and mortality for colorectal cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Childhood vaccination programme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Influenza vaccination for older people . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Patient experience with ambulatory care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Chapter 9. Health expenditure and financing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

Health expenditure per capita . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Health expenditure in relation to GDP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Health expenditure by function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Financing of health care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
6 HEALTH AT A GLANCE 2015 OECD 2015

TABLE OF CONTENTS
Expenditure by disease and age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

Capital expenditure in the health sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Chapter 10. Pharmaceutical sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177

Pharmaceutical expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Financing of pharmaceutical expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Pharmacists and pharmacies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Pharmaceutical consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Share of generic market. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Research and development in the pharmaceutical sector . . . . . . . . . . . . . . . . . . . . . 188
Chapter 11. Ageing and long-term care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Demographic trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
Life expectancy and healthy life expectancy at age 65 . . . . . . . . . . . . . . . . . . . . . . . 194
Self-reported health and disability at age 65 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Dementia prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Recipients of long-term care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
Informal carers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Long-term care workers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
Long-term care beds in institutions and hospitals . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Long-term care expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Annex A. Additional information on demographic and economic context,

and health expenditure and financing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Follow OECD Publications on:

http://twitter.com/OECD_Pubs
http://www.facebook.com/OECDPublications
http://www.linkedin.com/groups/OECD-Publications-4645871
http://www.youtube.com/oecdilibrary
OECD
Alerts http://www.oecd.org/oecddirect/
This book has... StatLinks2

A service that delivers Excel files from the printed page!
Look for the StatLinks2at the bottom of the tables or graphs in this book.
To download the matching Excel spreadsheet, just type the link into your
Internet browser, starting with the http://dx.doi.org prefix, or click on the link from
the e-book edition.

EXECUTIVE SUMMARY
Executive Summary
H ealth at a Glance 2015 presents cross-country comparisons of the health status of

populations and the performance of health systems in OECD countries, candidate
countries and key emerging economies. This edition offers two new features: a set of
dashboard indicators on health outcomes and health systems (presented in Chapter 1),
which summarise the comparative performance of OECD countries; and a special chapter
on recent trends in pharmaceutical spending across OECD countries. The key findings of
this publication are as follows.
New drugs will push up pharmaceutical spending unless policy adapts

Across OECD countries, pharmaceutical spending reached around USD 800 billion in
2013. This amounts to about 20% of total health spending on average when
pharmaceutical consumption in hospital is added to the purchase of pharmaceutical
drugs in the retail sector.
The growth of retail pharmaceutical spending has slowed down in recent years in most
OECD countries, while spending on pharmaceuticals in hospital has generally increased.
The emergence of new high-cost, specialty medicines targeting small populations and/or
complex conditions has prompted new debate on the long-term sustainability and
efficiency of pharmaceutical spending.
Life expectancy continues to rise, but widespread differences persist across

countries and socio-demographic groups
Life expectancy continues to increase steadily in OECD countries, rising on average by 3-
4 months each year. In 2013, life expectancy at birth reached 80.5 years on average, an
increase of over ten years since 1970. Japan, Spain and Switzerland lead a group of eight
OECD countries in which life expectancy now exceeds 82 years.
Life expectancy in key emerging economies, such as India, Indonesia, Brazil and China,
has increased over the past few decades, converging rapidly towards the OECD average.
There has been much less progress in countries such as South Africa (due mainly to the
epidemic of HIV/AIDS) and the Russian Federation (due mainly to a rise in risk-
increasing behaviours among men).
Across OECD countries, women can expect to live more than 5 years longer than men,
but this gap has narrowed by 1.5 years since 1990.
People with the highest level of education can expect to live six years longer on average
than those with the lowest level. This difference is particularly pronounced for men,
with an average gap of almost eight years.

EXECUTIVE SUMMARY
The number of doctors and nurses has never been higher in OECD countries
Since 2000, the number of doctors and nurses has grown in nearly all OECD countries,
both in absolute number and on a per capita basis. The growth was particularly rapid in
some countries that had fewer doctors in 2000 (e.g., Turkey, Korea, Mexico and the United
Kingdom), but there was also a strong rise in countries that already had a relatively large
number of doctors (e.g., Greece, Austria and Australia).
Growth was pushed by increased student intakes in domestic medical and nursing
education programmes, as well as by more foreign-trained doctors and nurses working
in OECD countries in response to short-term needs.
There are more than two specialist doctors for every generalist on average across the
OECD. In several countries, the slow growth in the number of generalists raises concerns
about access to primary care for all the population.
Out-of-pocket spending remains a barrier to accessing care

All OECD countries have universal health coverage for a core set of services, except
Greece, the United States and Poland. In Greece, the economic crisis led to a loss in
health insurance coverage among long-term unemployed and many self-employed
workers. However, since June 2014, measures have been taken to provide the uninsured
population with access to prescribed pharmaceuticals and emergency services. In the
United States, the percentage of the population uninsured has come down from 14.4% in
2013 to 11.5% in 2014 following the implementation of the Affordable Care Act and is
expected to diminish further in 2015.
Out-of-pocket spending by households can create barriers to health care access. On
average across OECD countries, about 20% of health spending is paid directly by patients,
ranging from less than 10% in France and the United Kingdom to over 30% in Mexico,
Korea, Chile and Greece. In Greece, the share of health spending paid directly by
households has increased by 4 percentage points since 2009, as public spending was
reduced.
Low-income households are four to six times more likely to report unmet needs for
medical care and dental care for financial or other reasons than those with high income.
In some countries, like Greece, the share of the population reporting some unmet
medical care needs has more than doubled during the economic crisis.
Too many lives are still lost because quality of care is not improving
fast enough
Better treatment of life-threatening conditions such as heart attack and stroke has led to
lower mortality rates in most OECD countries. On average, mortality rates following
hospital admissions for heart attack fell by about 30% between 2003 and 2013 and for
stroke by about 20%. Despite the progress achieved so far, there is still room in many
countries to improve the implementation of best practices in acute care to further
reduce mortality after heart attack and stroke.
Survival has also improved for many types of cancer in most countries, due to earlier
diagnosis and better treatment. For example, the relative five-year survival for breast
cancer and colorectal cancer has increased from around 55% on average for people
diagnosed and followed up in the period 1998-2003 to over 60% for those diagnosed and
followed up ten years later (2008-13). Still, several countries such as Chile, Poland and the

EXECUTIVE SUMMARY
United Kingdom are still lagging behind the best performers in survival following
diagnosis for different types of cancer.
The quality of primary care has improved in many countries, as illustrated by the
continuing reduction in avoidable hospital admissions for chronic diseases. Still, there is
room in all countries to improve primary care to further reduce costly hospital
admissions, in the context of population ageing and a growing number of people with
one or more chronic diseases.
Pharmaceutical prescribing practices can also be used as indicators of health care
quality. For example, antibiotics should be prescribed only where there is an evidence-
based need, to reduce the risk of antimicrobial resistance. Total volumes of antibiotic
consumption vary more than four-fold across OECD countries, with Chile, the
Netherlands and Estonia reporting the lowest, and Turkey and Greece reporting the
highest. Reducing unnecessary antibiotic use is a pressing, yet complex problem,
requiring multiple co-ordinated initiatives including surveillance, regulation and
education of professionals and patients.

READERS GUIDE
Readers guide
H ealth at a Glance 2015 presents comparisons of key indicators of health and health
system performance across the 34 OECD countries, as well as for candidate and key
partner countries where possible (Brazil, China, Colombia, Costa Rica, India, Indonesia,
Latvia, Lithuania, the Russian Federation and South Africa). The data presented in this
publication come mainly from official national statistics, unless otherwise indicated.
Content of the publication

This new edition of Health at a Glance contains two main new features: 1) a series of
dashboards are presented in Chapter 1 to summarise, in a clear and user-friendly way, the
relative strengths and weaknesses of OECD countries on a selected set of key indicators on
health and health system performance which are presented in other chapters of this
publication; and 2) a special focus is put on the pharmaceutical sector, including an
analysis of recent trends and future challenges in the management of pharmaceutical
expenditure in Chapter 2, as well as a new chapter on the pharmaceutical sector
(Chapter 10), combining both indicators that were previously shown in other chapters and
some new indicators based on the two-page format used in most of this publication.
The general framework underlying the indicators presented in this publication
assesses the performance of health systems in the context of a broader view of public
health (Figure 0.1). It is based on a framework that was endorsed for the OECD Health Care
Quality Indicators project (Kelley and Hurst, 2006; Arah et al., 2006). This framework
recognises that the goal of health systems is to improve the health status of the population.
Many factors influence health status, including a number that fall outside health care
systems, such as the physical environment in which people live, and individual lifestyles
and behaviours. The performance of health care systems also contributes obviously to the
health status of the population. This performance includes several dimensions, including
the degree of access to care and the quality of care provided. Performance measurement
also needs to take into account the financial resources required to achieve these access and
quality goals. The performance of health systems depends also greatly on the health
workers providing the services, and the training and equipment at their disposal. Finally, a
number of contextual factors also affect the health status of the population and the
demand for and supply of health services also need to be taken into account, including the
demographic context, and economic and social development.
Health at a Glance 2015 compares OECD countries on each component of this general
framework.

Figure 0.1. Conceptual framework for health system performance assessment
Health status
(Chapter 3)
Non-medical determinants of health

(Chapter 4)
Health care system performance

How does the health system perform?
What is the level of quality of care and access to services?
What does this performance cost?
Quality of care Access to care Health expenditure and financing
(Chapter 8) (Chapter 7) (Chapter 9)
Health care resources and activities
Health workforce Health care activities

(Chapter 5) (Chapter 6)
Demographic and economic context, and health expenditure and financing

(Annex A)
Source: Adapted from Kelley, E. and J. Hurst (2006).
Following the first two new chapters presenting the set of dashboards of indicators
and the special focus on pharmaceutical expenditure, Chapter 3 on health status highlights
variations across countries in life expectancy, some of the main causes of mortality and
other measures of population health status. This chapter also includes measures of
inequality in health status by education and income level for key indicators such as life
expectancy and perceived health status.
Chapter 4 on non-medical determinants of health focuses on health-related lifestyles
and behaviours, including tobacco smoking, alcohol drinking, nutrition, and overweight
and obesity problems among children and adults. Most of these factors can be modified by
public health and prevention policies.
Chapter 5 looks at the health workforce, focusing on the supply and remuneration of
doctors and nurses in OECD countries. This chapter presents trends in the number of new
graduates from medical and nursing education programmes and also features new
indicators on the international migration of doctors and nurses, highlighting the fact that
the number and share of foreign-trained doctors and nurses has increased in many OECD
countries over the past decade.
Chapter 6 on health care activities describes some of the main characteristics of health
service delivery in different OECD countries, starting with the number of consultations
with doctors (which is often the entry point of patients to health care systems),
hospitalisation rates, the utilisation rates of different diagnostic and surgical procedures,
as well as the development of ambulatory surgery for interventions such as cataract
surgery and tonsillectomy.
Chapter 7 on access to care presents a set of indicators related to financial access to
care, geographic access, and timely access (waiting times), as well as indicators of self-
reported unmet needs for medical care and dental care.
READERS GUIDE
Chapter 8 examines quality of care or the degree to which care is delivered in

accordance with established standards and improves health outcomes. It provides
comparisons on quality of care for chronic conditions and pharmaceutical prescriptions,
acute care for life-threatening diseases such as heart attack and stroke, patient safety,
mental health care, cancer care, the prevention of communicable diseases, as well as some
important aspects of patient experiences.
Chapter 9 on health expenditure and financing compares how much OECD countries
spend on health, both on a per capita basis and in relation to GDP. The chapter also
provides an analysis of the different types of health services and goods consumed across
OECD countries. It also looks at how these health services and goods are paid for and the
mix between public funding, private health insurance and direct out-of-pocket payments
by households in different countries.
Chapter 10 is a new chapter on the pharmaceutical sector, which combines some
indicators that were previously shown in other chapters and some new indicators. The
chapter begins by comparing recent trends and levels of pharmaceutical expenditure
across countries and how these expenditure are paid for, and then goes on to compare the
consumption of certain high-volume pharmaceutical drugs and the share of the generic
market in different countries. It concludes by reviewing spending on research and
development (R&D) to develop new products in the pharmaceutical sector.
Chapter 11 focuses on ageing and long-term care, starting by a review of demographic
trends which highlights the steady growth in the share of the population aged over 65 and
80 in all OECD countries. The chapter presents the most recent data on life expectancy and
life expectancy free of disability at age 65, along with data on self-reported health and
disability status, as important factors affecting the current and future demand for long-
term care. It then focuses on people currently receiving long-term care at home or in
institutions and people providing formal or informal care, and concludes with a review of
levels and trends in long-term care expenditure in different countries.
A statistical annex provides additional information on the demographic and economic
context within which health and long-term care systems operate.
Presentation of indicators
With the exception of the first two chapters, each of the indicators covered in the rest
of the publication is presented over two pages. The first provides a brief commentary
highlighting the key findings conveyed by the data, defines the indicator and signals any
significant national variation from the definition which might affect data comparability.
On the facing page is a set of figures. These typically show current levels of the indicator
and, where possible, trends over time. Where an OECD average is included in a figure, it is
the unweighted average of the OECD countries presented, unless otherwise specified.
Data limitations
Limitations in data comparability are indicated both in the text (in the box related to
Definition and comparability) as well as in footnotes to figures.
Data sources
Readers interested in using the data presented in this publication for further analysis
and research are encouraged to consult the full documentation of definitions, sources and
methods presented in OECD Health Statistics on OECD.Stat (http://stats.oecd.org/index.aspx,

READERS GUIDE
then choose Health). More information on OECD Health Statistics is available at

www.oecd.org/health/health-data.htm.
Population figures
The population figures presented in the Annex and used to calculate rates per capita
throughout this publication come from the OECD Historical Population Data and
Projections (as of end of May 2015), and refer to mid-year estimates. Population estimates
are subject to revision, so they may differ from the latest population figures released by the
national statistical offices of OECD member countries.
Note that some countries such as France, the United Kingdom and the United States
have overseas colonies, protectorates or territories. These populations are generally
excluded. The calculation of GDP per capita and other economic measures may, however,
be based on a different population in these countries, depending on the data coverage.
OECD country ISO codes
Australia AUS Japan JPN

Austria AUT Korea KOR
Belgium BEL Luxembourg LUX
Canada CAN Mexico MEX
Chile CHL Netherlands NLD
Czech Republic CZE New Zealand NZL
Denmark DNK Norway NOR
Estonia EST Poland POL
Finland FIN Portugal PRT
France FRA Slovak Republic SVK
Germany DEU Slovenia SVN
Greece GRC Spain ESP
Hungary HUN Sweden SWE
Iceland ISL Switzerland CHE
Ireland IRL Turkey TUR
Israel ISR United Kingdom GBR
Italy ITA United States USA
Partner country ISO codes
Brazil BRA Indonesia IDN

China CHN Latvia LVA
Colombia COL Lithuania LTU
Costa Rica CRI Russian Federation RUS
India IND South Africa ZAF
References
Arah, O. et al. (2006), A Conceptual Framework for the OECD Health Care Quality Indicators Project,
International Journal for Quality in Health Care, Vol. 18, Supplement No. 1, pp. 5-13.
Kelley, E. and J. Hurst (2006), Health Care Quality Indicators Project: Conceptual Framework, OECD
Health Working Paper, No. 23, OECD Publishing, Paris, http://dx.doi.org/10.1787/440134737301.

OECD 2015
Chapter 1
Dashboards of health indicators
This chapter presents, for the first time, a set of dashboards which are designed to
shed light on how well OECD countries do in promoting the health of their
population and improving their health system performance. These dashboards do
not have the ambition of identifying which countries have the best health system
overall. They summarise some of the relative strengths and weaknesses of countries
on a selected set of indicators on health and health system performance, to help
identify possible priority areas for actions. These dashboards, which take the form
of summary tables, highlight how well OECD countries are doing along five
dimensions: 1) health status; 2) risk factors to health; 3) access to care; 4) quality of
care; and 5) health care resources. For each of these five dimensions, a selected set
of key indicators are presented. The selection of these indicators is based on three
main criteria: 1) policy relevance; 2) data availability; and 3) data interpretability
(i.e., no ambiguity that a higher/lower value means a better/worse performance).
There is, however, one exception to the application of this third criterion: for the fifth
dashboard on health care resources, more health spending or more human or
physical resources does not necessarily mean better performance. This is why the
ranking of countries is displayed differently.
The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli
authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights,
East Jerusalem and Israeli settlements in the West Bank under the terms of international law.
17
1. DASHBOARDS OF HEALTH INDICATORS
A cross the OECD, policy makers have a keen interest to understand how good the health
of their people is, and how well their health systems are able to deliver good results. A look
at indicators contained in this publication shows that much progress has already been
achieved. People in OECD countries are living longer than ever before, with life expectancy
now exceeding 80 years on average, thanks to improvements in living conditions and
educational attainments, but also to progress in health care. In most countries, universal
health coverage provides financial protection against the cost of illness and promotes
access to care for the whole population. The quality of care has also generally improved, as
illustrated by the reduction in deaths after heart attacks and strokes, and the earlier
detection and improved treatments for serious diseases such as diabetes and cancer. But
these improvements have come at a cost. Health spending now accounts for about 9% of
GDP on average in OECD countries, and exceeds 10% in many countries. Higher health
spending is not a problem if the benefits exceed the costs, but there is ample evidence of
inequities and inefficiencies in health systems which need to be addressed. There is also a
need to achieve a proper balance between spending on disease prevention and treatment.
Despite these improvements, important questions about how successful countries are
in achieving good results on different dimensions of health system performance remain.
What are the main factors explaining differences in health status and life expectancy
across OECD countries? Is the increase in certain risk factors such as inactivity and obesity
offsetting some of the gains from the reduction in other risk factors like smoking? To what
extent do all citizens have adequate and timely access to care, and good financial
protection against the cost of health care? What do we know about the quality and safety
of care provided to people with different health conditions? What are the financial, human
and technical resources allocated to health systems in different countries? And how does
this translate into beneficial activities and better health outcomes?
Answering these questions is by no mean an easy task. But the dashboards presented
in this chapter can help shed light on how well countries do in promoting the health of
their population and on several dimensions of health system performance. These
dashboards do not have the ambition of identifying which countries have the best health
system overall. However, they summarise some of the relative strengths and weaknesses
of OECD countries on a selected set of indicators on health and health system
performance, and can be useful to identify possible priority areas for actions.
These dashboards, which take the form of summary tables, highlight how well OECD
countries are doing along five dimensions: 1) health status; 2) risk factors to health;
3) access to care; 4) quality of care; and 5) health care resources. For each of these five
dimensions, a selected set of key indicators (ranging from 4 to 7) are presented in a
summary table. The selection of these indicators is based on three main criteria: 1) policy
relevance; 2) data availability; and 3) data interpretability (i.e., no ambiguity that a higher/
lower value means a better/worse performance). There is, however, one notable exception
to the application of this third criterion: for the fifth dashboard on health care resources,

more health spending or more human or physical resources does not necessarily mean
better performance. This is why the ranking of countries is displayed differently (through
different colours) in this last dashboard. Box 1.1 at the end of this chapter summarises
some of the main limitations in interpreting these dashboards.
In most of the dashboards, countries are classified in three groups: 1) top third
performer; 2) middle third performer; and 3) bottom third performer. In addition, the
specific ranking of countries is indicated in each cell to provide further information on how
close countries may be to the other group. The ranking is based on the number of countries
for which data are available for each indicator (with a maximum of 34, when all countries
are covered), with countries separated in three equal groups. For the first indicator related
to access to care (the percentage of the population with health coverage), the grouping of
countries is based on a different method because most countries are at or close to 100%
coverage: the top countries are defined as those with a population coverage rate between
95% and 100%, the middle countries with a coverage between 90% and 95%, and the bottom
countries with a coverage of less than 90%. The availability of comparable data is also more
limited for indicators of access to care, either because of a lack of harmonisation in survey
instruments (for indicators related to unmet care needs) or limitations in administrative
data (for indicators on waiting times).
Health status
The broad measures of population health status shown in Table 1.1, such as life
expectancy at various ages, are not only related to health spending and the performance of
health systems, but also to a wide range of non-medical determinants of health (with some
of the lifestyle and behavioural factors presented in Table 1.2). Countries that perform well
on life expectancy at birth for men and women usually also tend to do well on life
expectancy at older ages, and typically have lower rates of mortality from cardiovascular
diseases (the main causes of death in nearly all OECD countries).
Japan, Spain, Switzerland, Italy and France are among the countries that have the
highest life expectancy at birth and at older ages, although France does not perform so well
in terms of life expectancy at birth for men, reflecting higher mortality rates among
younger and middle-aged men.
Mexico, Hungary, the Slovak Republic and Turkey have the lowest life expectancy at
birth and older ages, although Turkey has achieved huge gains in longevity over the past
few decades and is quickly moving towards the OECD average (see the first indicator on life
expectancy in Chapter 3 for trends over time).
While higher health spending per capita is generally associated with higher life
expectancy, this relationship is less pronounced in countries with the highest health
spending per capita. Japan, Spain and Korea stand out as having relatively high life
expectancies, and the United States relatively low life expectancies, given their levels of
health spending (see Table 1.5). Life expectancy in the United States is lower than in most
other OECD countries because of higher mortality rates from various health-related
behaviors (including higher calorie consumption and obesity rates, higher consumption of
legal and illegal drugs, higher deaths from road traffic accidents and homicides), adverse
socio-economic conditions affecting a large segment of the US population, and poor access
and co-ordination of care for certain population groups.

Risk factors to health

Most countries do not perform well for at least one or more indicators of risk factors to
health, whether that is the proportion of their population smoking tobacco, alcohol
consumption, or overweight and obesity among children and adults (Table 1.2). This
highlights the importance of countries putting a higher priority on health promotion and
disease prevention policies to reduce modifiable risk factors to health and mortality from
related diseases.
The United States, Canada, Australia and Mexico have achieved remarkable progress
over the past few decades in reducing tobacco smoking among adults and have very low
rates now, but they face the challenge of tackling relatively high rates of overweight and
obesity among children and adults. Some countries like Italy and Portugal currently have a
relatively low rate of obesity among adults, but the current high rate of overweight and
obesity among children is likely to translate into higher rates among adults in the future.
Other countries like Turkey and Greece have relatively low levels of alcohol consumption,
but still have a way to go to reduce tobacco smoking. Alcohol consumption remains high in
Austria, Estonia, the Czech Republic, Hungary, France and Germany, although the overall
level of consumption has come down in many of these countries over the past few decades
(see the indicator on alcohol consumption in Chapter 4).
Access to care
Most OECD countries have achieved universal (or near-universal) coverage of health
care costs for a core set of services, with the exception of Greece, the United States and
Poland, where a sizeable proportion of the population is still not covered (Table 1.3). In the
United States, the percentage of the population uninsured has started to decrease
significantly in 2014, following the implementation of the Affordable Care Act which is
designed to expand health insurance coverage. In Greece, the response to the economic
crisis has reduced health insurance coverage among people who have become long-term
unemployed, and many self-employed workers have also not renewed their health
insurance plans because of reduced disposable income. However, since June 2014,
uninsured people are covered for prescribed pharmaceuticals and for services in
emergency departments in public hospitals, as well as for non-emergency hospital care
under certain conditions.
The financial protection that people have against the cost of illness depends not only
on whether they have a health insurance, but also on the range of goods and services
covered and the extent to which these goods and services are covered. In countries like
France and the United Kingdom, the amount that households have to pay directly for
health services and goods as a share of their total consumption is relatively low, because
most such goods and services are provided free or are fully covered by public and private
insurance, with only small additional payments required. Some other countries, such as
Korea and Mexico, have achieved universal (or quasi-universal) health coverage, but a
relatively small share of the cost of different health services and goods are covered,
leaving a significant amount to be paid by households. Direct out-of-pocket payments
can create financial barriers to health care, dental care, prescribed pharmaceutical drugs
or other health goods or services, particularly for low-income households. The share of
household consumption spent on direct medical expenditure is highest in Korea,
Switzerland, Portugal, Greece and Mexico, although some of these countries have put in
place proper safeguards to protect access to care for people with lower income.

Access to health care may be restricted not only because of financial reasons, but also
because of geographic barriers, waiting times and other reasons. In Europe, around 3% of
the population on average in countries that are OECD members reported unmet needs for
medical examination due to cost, travel distance or waiting lists in 2013, according to the
EU-SILC survey. The share of the population reporting such unmet medical care needs was
highest in Greece and Poland, and lowest in the Netherlands and Austria. In nearly all
countries, a higher proportion of the population reports some unmet needs for dental care,
reflecting that public coverage for dental care is generally lower. People in Portugal, Iceland,
Italy and Greece reported the highest rates of unmet needs for dental care among
European countries that are OECD members in 2013.
Waiting times for different health services indicate the extent to which people have
timely access to care for specific interventions such as elective surgery. Denmark, Canada
and Israel have relatively low waiting times for interventions such as cataract surgery and
knee replacement among the limited group of countries that provide these data, while
Poland, Estonia and Norway have relatively long waiting times.
Quality of care
Improving quality of care is a high priority in most OECD countries. Based on the
available data, no country consistently performs in the top group on all indicators of
quality of care (Table 1.4), even those that spend much more on health. This suggests that
there is room for improvement in all countries in the governance of health care quality and
prevention, early diagnosis and treatment of different health problems.
The United States is doing well in providing acute care for people having a heart attack
or a stroke and preventing them from dying, but is not performing very well in preventing
avoidable hospital admissions for people with chronic conditions such as asthma and
diabetes. The reverse is true in Portugal, Spain and Switzerland, which have relatively low
rates of hospital admissions for certain chronic conditions, but relatively high rates of
mortality for patients admitted to hospital for a heart attack or stroke.
Finland and Sweden do relatively well in having high survival of people following
diagnosis for cervical, breast or colorectal cancer, while the survival for these types of
cancer remains lower in Chile, Poland, the Czech Republic, the United Kingdom and
Ireland. An important pillar to achieve progress in the fight against cancer is to establish a
national cancer control plan to focus political and public attention on performance in
cancer prevention, early diagnosis and treatment.
Health care resources

Higher health spending is not always closely related to a higher supply of health
human resources or to a higher supply of physical and technical equipment in health
systems.
The United States continues to spend much more on health per capita than all other
OECD countries, but is not in the top group in terms of the number of doctors or nurses per
population. Following the United States, the next biggest spenders on health are
Switzerland, Norway, the Netherlands and Sweden, whereas the lowest per capita
spenders are Mexico and Turkey (Table 1.5). Health spending per capita is also relatively
low in Chile, Poland and Korea, although it has grown quite rapidly over the past decade.
Greece, Austria and Norway have the highest number of doctors per capita, while
Switzerland, Norway and Denmark have the highest number of nurses. The mix between
different categories of health workers varies widely, with some countries choosing to have

relatively more doctors (such as Greece and Austria) and others opting to rely more on
nurses and other health care providers to deliver some services (such as Finland and the
United States).
Some Central and Eastern European countries such as Hungary, Poland and the Slovak
Republic continue to have a relatively high number of hospital beds, reflecting an excessive
focus of activities in hospital. The number of hospital beds per capita is lowest in Mexico,
Chile, Sweden, Turkey, Canada and the United Kingdom. Relatively low number of hospital
beds may not create any capacity problem if primary care systems are sufficiently
developed to reduce the need for hospitalisation.
The availability of expensive technological equipment such as MRI and CT scanners is
highest in Japan and the United States, and much lower in Mexico, Hungary, Israel and the
United Kingdom. There is no ideal number of MRI units or CT scanners per population, and
there is also evidence in many countries of inappropriate and excessive use of these
expensive diagnostic technologies.
Higher health spending and other human or technical resources are not always
correlated with greater access to care or higher quality of care, as shown by the lack of any
consistent correlation in countries relative position between health spending and various
indicators of access or quality of care. For example, Norway has high levels of health
spending and also relatively high numbers of doctors and nurses, and does generally well
on many indicators of quality of care, but still faces some persisting issues in terms of
access to care, for instance, on waiting times for elective surgery. On the other hand, the
Czech Republic spends much less on health and is achieving good results for several
indicators on access to care, but could improve public health and prevention programmes
and improve the quality of care for people who have chronic diseases such as diabetes. The
performance of health systems in achieving the key policy goals of universal access and
quality depends not only on allocating more money on health care, but also on making a
more rational use of resources and providing the right incentives to ensure the best value
for money spent.

Table 1.1. Health status

Top third performers
Middle third performers
Bottom third performers
Note: Countries are listed in alphabetical order. The number in the cell indicates the position of each country among all countries
for which data is available. For the mortality indicator, the top performers are countries with the lowest rates.
Mortality from
Life expectancy Life expectancy Life expectancy Life expectancy
Indicator cardiovascular
at birth - Men at birth - Women at 65 - Men* at 65 - Women*
diseases**
Australia 8 7 3 7 7
Austria 18 13 16 13 26
Belgium 22 19 23 14 15
Canada 13 17 10 10 5
Chile 27 27 27 28 16
Czech Rep. 28 28 29 30 31
Denmark 21 25 25 26 10
Estonia 32 26 31 27 32
Finland 23 8 20 9 24
France 15 3 2 2 2
Germany 18 19 16 22 25
Greece 17 9 13 11 27
Hungary 33 33 34 34 33
Iceland 2 16 10 20 23
Ireland 15 23 19 24 21
Israel 3 11 3 17 3
Italy 3 4 8 4 17
Japan 5 1 6 1 1
Korea 20 5 20 5 4
Luxembourg 9 11 6 8 12
Mexico 34 34 28 32 22
Netherlands 11 19 16 20 8
New Zealand 11 19 8 17 18
Norway 9 13 15 14 11
Poland 30 29 30 28 30
Portugal 24 9 23 11 14
Slovak Rep. 31 31 33 31 34
Slovenia 25 17 26 14 28
Spain 5 2 3 3 6
Sweden 5 13 10 17 19
Switzerland 1 6 1 5 13
Turkey 29 32 32 33 29
United Kingdom 14 24 14 23 9
United States 26 29 22 25 20
* Life expectancy at 65 is not presented in chapter 3 on health status, but rather in chapter 11 on ageing and long-
term care.
** Mortality from cardiovascular diseases includes deaths from ischemic heart diseases and cerebrovascular diseases
shown in Chapter 3, as well as other cardiovascular diseases.
Source: Health at a Glance 2015.
1 2 http://dx.doi.org/10.1787/888933281467

Table 1.2. Risk factors

Note: Countries are listed in alphabetical order. The number in the cell indicates the position of each country among all countries for which
data is available.
Alcohol Overweight and

Indicator Smoking in adults Obesity in adults*
consumption obesity in children**
Australia 4 22 30* 20
Austria 26 34 8 14
Belgium 15 20 9 5
Canada 6 11 29* 21
Chile 33 10 28* 28
Czech Rep. 25 32 20* 5
Denmark 12 17 10 23
Estonia 31 33 18 7
Finland 10 14 26 17
France 30 30 11 13
Germany 23 28 25* 3
Greece 34 7 19 33
Hungary 32 30 31* 24
Iceland 2 6 21 9
Ireland 16 26 24* 11
Israel 11 2 13 18
Italy 24 4 4 31
Japan 17 7 1* 15
Korea 19 12 2* 16
Luxembourg 9 29 23* 19
Mexico 3 3 33* 30
Netherlands 13 14 6 7
New Zealand 8 16 32* 27
Norway 7 5 3 1
Poland 27 27 14 2
Portugal 14 25 12 25
Slovak Rep. 18 22 16* 3
Slovenia 22 17 17 22
Spain 29 20 15 26
Sweden 1 7 7 9
Switzerland 21 22 4 11
Turkey 27 1 22* n.a.
United Kingdom 20 19 27* 32
United States 5 13 34* 29
* Data on obesity in adults are based on measured height and weight for all the countries marked with an *. These
result in more accurate data and higher obesity rates compared with all other countries that are providing self-
reported height and weight.
** Data on overweight or obesity in children are all based on measured data, but refer to different age groups across
countries.
1 2 http://dx.doi.org/10.1787/888933281473

Table 1.3. Access to care
Top third performers (or between 95% and 100% for health care coverage)
Middle third performers (or between 90% and 95% for health care coverage)
Bottom third performers (or less than 90% for health care coverage)
Note: Countries are listed in alphabetical order. The number in the cell indicates the position of each country among all countries for
which data is available. For out-of-pocket medical expenditure, unmet care needs and the waiting times indicators, the top performers
in terms of access are countries with the lowest expenditure as a share of household consumption, the lowest unmet care needs
or lowest waiting times.
Share of out of
Waiting times Waiting times
Health care pocket medical
Unmet medical Unmet dental for cataract for knee
Indicator coverage expenditure in
care needs* care needs* surgery replacement
houselhold
- median - median
consumption
Australia 1 22 n.a. n.a. 8 12
Austria 1 18 1 2 n.a. n.a.
Belgium 1 20 11 8 n.a. n.a.
Canada 1 11 n.a. n.a. 2 4
Chile 1 28 n.a. n.a. 13 8
Czech Rep. 1 7 5 4 n.a. n.a.
Denmark 1 14 7 10 4 1
Estonia 2 12 21 19 9 13
Finland 1 18 19 11 10 7
France 1 3 15 15 n.a. n.a.
Germany 1 5 9 5 n.a. n.a.
Greece 3 32 23 20 n.a. n.a.
Hungary 1 30 14 9 1 6
Iceland 1 21 18 22 n.a. n.a.
Ireland 1 22 17 17 n.a. n.a.
Israel 1 16 n.a. n.a. 3 3
Italy 1 22 20 21 n.a. n.a.
Japan 1 9 n.a. n.a. n.a. n.a.
Korea 1 34 n.a. n.a. n.a. n.a.
Luxembourg 1 5 4 3 n.a. n.a.
Mexico 1 30 n.a. n.a. n.a. n.a.
Netherlands 1 2** 1 1 n.a. n.a.
New Zealand 1 9 n.a. n.a. 7 5
Norway 1 16 8 15 12 10
Poland 2 13 22 13 14 14
Portugal 1 29 16 23 6 11
Slovak Rep. 2 22 11 6 n.a. n.a.
Slovenia 1 7 n.a. n.a. n.a. n.a.
Spain 1 26 3 18 11 9
Sweden 1 26 11 14 n.a. n.a.
Switzerland 1 33 6 12 n.a. n.a.
Turkey 1 1 n.a. n.a. n.a. n.a.
United Kingdom 1 3 9 7 4 2
United States 3 14 n.a. n.a. n.a. n.a.
* Unmet medical or dental care needs may be for financial reasons, waiting times or long distance to travel to get
access to services. The data only cover European countries because they are based on the EU-SILC survey.
** The ranking for the Netherlands is overrated as it excludes compulsory co-payments to health insurers (if these
were included, this would move the Netherlands in the middle third category).
1 2 http://dx.doi.org/10.1787/888933281483

Table 1.4. Quality of care

Note: Countries are listed in alphabetical order. The number in the cell indicates the position of each country among all countries for which
data is available. For the indicators of avoidable hospital admissions and case-fatality rates, the top performers are countries with the
lowest rates.
Case-fatality
Case-fatality
Asthma and Diabetes for ischemic Cervical Breast Colorectal
for AMI
Indicator COPD hospital hospital stroke cancer cancer cancer
(admission-
admission admission (admission- survival survival survival
based)
based)
Australia 29 17 1 20 11 5 3
Austria 28 29 27 8 19 19 7
Belgium 16 20 19 20 16 12 4
Canada 18 10 11 26 12 8 13
Chile 6 27 31 16 25 23 n.a.
Czech Rep. 12 23 11 22 13 22 21
Denmark 26 14 7 17 5 11 18
Estonia 27 n.a. 28 29 8 25 22
Finland 10 15 9 4 6 4 7
France 7 21 17 13 n.a. n.a. n.a.
Germany 21 25 25 8 15 15 10
Greece n.a. n.a. n.a. n.a. n.a. n.a. n.a.
Hungary 31 11 30 22 n.a. n.a. n.a.
Iceland 14 4 15 14 7 10 n.a.
Ireland 32 16 8 24 20 20 19
Israel 19 9 11 6 10 7 2
Italy 2 1 5 7 3 15 12
Japan 1 18 29 1 4 9 4
Korea 24 30 24 2 2 14 1
Luxembourg 9 19 16 17 n.a. n.a. n.a.
Mexico 5 31 32 31 n.a. n.a. n.a.
Netherlands 11 6 20 12 16 16 11
New Zealand 30 22 10 14 14 12 15
Norway 17 7 11 5 1 2 13
Poland 20 28 3 n.a. 24 24 23
Portugal 3 8 26 27 18 6 16
Slovak Rep. 23 26 17 28 n.a. n.a. n.a.
Slovenia 8 13 4 30 23 18 17
Spain 15 3 23 24 n.a. n.a. n.a.
Sweden 13 12 2 8 9 1 6
Switzerland 4 2 22 11 n.a. n.a. n.a.
Turkey n.a. n.a. n.a. n.a. n.a. n.a. n.a.
United Kingdom 22 5 20 19 22 21 20
United States 25 24 5 3 21 2 9

1 2 http://dx.doi.org/10.1787/888933281494

Table 1.5. Health care resources

Top third in health spending or resources
Middle third in health spending or resources
Bottom third in health spending or resources
Note: Countries are listed in alphabetical order. The number in the cell indicates the position of each country among all countries for
which data is available. Although countries are ranked from highest health spending or availability of resources to lowest, this does not
necessarily mean better performance.
Doctors Nurses
Health expenditure Hospital beds MRI units CT scanners
Indicator per capita per capita
per capita per capita per capita* per capita*
(active) (active)
Australia 13 14 10 18 12* 2*
Austria 8 2 21 4 9 10
Belgium 11 21 15 9 19* 11*
Canada 10 28 16 29 22 23
Chile 30 33 27 32 26 26
Czech Rep. 27 10 20 7 24 22
Denmark 7 11 3 23 10 5
Estonia 31 18 23 12 17 15
Finland 17 20 5 13 6 13
France 12 16 17 8 21 24
Germany 6 5 6 3 15* 16*
Greece 25 1 32 14 5 8
Hungary 29 19 22 5 31* 31*
Iceland 15 11 4 21 7 4
Ireland 16 25 7 26 13 17
Israel 24 13 31 22 30 29
Italy 20 8 24 19 3 9
Japan 14 29 13 1 1 1
Korea 26 31 29 2 4 6
Luxembourg 9 22 9 11 14 12
Mexico 33 32 33 33 32 32
Netherlands 4 17 8 n.a. 16 28
New Zealand 18 22 14 26 18 20
Norway 3 3 2 17 n.a. n.a.
Poland 32 30 28 6 28 19
Portugal 22 4 25 20 27* 14*
Slovak Rep. 28 14 26 10 25 21
Slovenia 23 26 18 16 23 27
Spain 21 9 30 24 11 18
Sweden 5 7 11 31 n.a. n.a.
Switzerland 2 6 1 15 8* 7
Turkey 34 34 34 30 20 25
United Kingdom 19 24 19 26 29 30
United States 1 27 12 25 2 3
* Data for most countries marked with an * do not include MRI units and CT scanners installed outside hospitals,
leading to an under-estimation. In Australia and Hungary, the data only include MRI units and CT scanners eligible
for public reimbursement, also leading to an under-estimation.
1 2 http://dx.doi.org/10.1787/888933281500

Box 1.1. Limitations in the interpretation and use of the dashboards

The previous dashboards should be interpreted and used with caution for several reasons:
Due to limitations in data availability, the indicators selected on each topic do not generally provide a
complete coverage of all important aspects related to this topic. For instance, the indicators of health
status relate solely to mortality because mortality data are more widely available and comparable across
countries than morbidity data. While life expectancy undoubtedly is a key indicator of health status, the
lack of indicators about the physical and mental health status of people while they are alive is an
important limitation. The same limitations also apply to the dashboards on risk factors (which only
include some risk factors to health), access to care and quality of care.
There are limitations in data comparability for some indicators which should be kept in mind in
interpreting the ranking of countries. One notable example is the indicator on obesity rates among
adults, which in several countries are based on self-reported height and weight, resulting in an under-
estimation compared to those countries that provide more reliable data based on measured obesity.
The grouping of countries in three groups (tertiles) is based on a simple method using only the point
estimates of each country and dividing them in three equal groups. It does not take into account the
distribution of the data around the OECD average, nor the confidence intervals for those indicators
where these have been calculated (notably for several indicators of quality of care).
These dashboards only present the current situation and in this respect may hide the progress that some
OECD countries might have achieved over time and the fact that they may be moving quickly towards the
OECD average. These key trends are discussed in the publication.
Because of these limitations in data availability, comparability and statistical significance, there is no
attempt to calculate any summary indicator of performance for each of the dimensions or across
dimensions. These dashboards should be used to get a first impression on the relative strengths and
weaknesses of different OECD countries on the set of indicators selected. It should be complemented by a
more in-depth review of the data and the factors influencing the cross-country variations presented in the
following chapters of this publication.

OECD 2015
Chapter 2
Pharmaceutical spending trends

and future challenges
Across OECD countries, pharmaceutical spending reached around USD 800 billion
in 2013, accounting for about 20% of total health spending on average when
pharmaceutical consumption in hospital is added to the purchase of pharmaceutical
drugs in the retail sector. This chapter looks at recent trends in pharmaceutical
spending across OECD countries. It examines the drivers of recent spending trends,
highlighting differences across therapeutic classes. It shows that while the
consumption of medicines continues to increase and to push pharmaceutical
spending up, cost-containment policies and patent expiries of a number of top-
selling products have put downward pressure on pharmaceutical prices in recent
years. This resulted in a slower pace of growth over the past decade.
The chapter then looks at emerging challenges for policy makers in the management
of pharmaceutical spending. The proliferation of high-cost specialty medicines will
be a major driver of health spending growth in the coming years. While some of
these medicines bring great benefits to patients, others provide only marginal
improvements. This challenges the efficiency of pharmaceutical spending.
29
2. PHARMACEUTICAL SPENDING TRENDS AND FUTURE CHALLENGES
Introduction
Pharmaceutical spending across OECD countries reached around USD 800 billion in
2013, accounting for about 20% of total health spending on average when pharmaceutical
consumption in hospitals is added to the purchase of pharmaceutical drugs in the retail
sector. Retail pharmaceutical spending growth has slowed down in most OECD countries in
the last decade, while spending on pharmaceuticals used in hospital has increased in most
countries where this information is available. Current market developments, such as the
multiplication of high-cost medicines targeting small populations and/or complex
conditions, have prompted new debates on the sustainability and efficiency of
pharmaceutical spending. Will OECD countries be able to afford access to these high-cost
medicines to all patients who need them and at what price? Will they get value for the
money they will spend?
This chapter looks first at recent trends in pharmaceutical spending and financing
across OECD countries. Then, it examines the drivers of recent spending trends,
highlighting differences across drug classes. Finally, it focuses on current and predicted
trends in pharmaceutical markets and associated challenges in the management of
pharmaceutical expenditure.
One in every five health dollars is spent on pharmaceuticals

In 2013, OECD countries spent an average of more than 500 USD per person on retail
pharmaceuticals1 (Figure 2.1). In the United States, the level of spending was twice the
OECD average, and more than 35% higher than in Japan, the next highest spender. At the
other end of the scale, Denmark spent less than half the OECD average.
The data on pharmaceutical spending shown in Figure 2.1 only include those
purchased in the retail sector, as many countries are not able to supply data on the cost of
pharmaceuticals consumed in hospitals and other health care facilities. In those countries
that are able to supply these data, the inclusion of pharmaceutical expenditure in hospital
and other facilities adds another 10% on top of the retail pharmaceutical spending in the
case of Germany, Canada and Australia, and more than 25% in countries such as Spain,
Czech Republic and Portugal (Figure 2.2). Such differences stem from the budgetary and
distributional channels within a country. On average, the use of pharmaceuticals in
hospitals and other health care facilities raises the pharmaceutical bill by around 20%,
meaning that a little more than one health dollar in five goes towards purchasing
pharmaceuticals.
Prior to 2005, spending on retail pharmaceuticals grew at a faster rate than other key
components of health care, such as inpatient and outpatient care, and was a major
contributor in driving up overall health expenditures (see Figure 2.3). Over the subsequent
decade, however, retail pharmaceutical spending growth was seriously affected by patent
expiries of several blockbuster drugs and cost-containment policies, particularly as a
consequence of the economic crisis. As a result, retail pharmaceutical spending decreased
dramatically in some countries, for example in Portugal, Denmark and Greece.

Figure 2.1. Expenditure on retail pharmaceuticals per capita and as a share of GDP, 2013
(or nearest year)
Pharmaceutical expenditure per capita Pharmaceutical expenditure as a share of GDP
1026 United States 1.9
752 Japan 2.1
721 Greece 1 2.8
713 Canada 1.7
678 Germany 1.5
666 Switzerland 1.2
652 Ireland1 1.4
603 Belgium 1.4
596 France 1.6
590 Australia 1.3
572 Italy1 1.6
536 Austria 1.2
533 Slovak Republic1 2.0
526 Spain 1.6
515 OECD29 1.4
503 Hungary 2.2
481 Slovenia 1.7
459 Finland 1.2
459 Sweden 1.0
436 Korea 1.3
397 Netherlands1 0.9
396 Iceland 0.9
392 Portugal1 1.4
381 Czech Republic 1.3
367 Norway 0.6
364 Luxembourg 2 0.6
326 Poland 1.4
287 Israel1 1.0
273 Estonia 1.1
240 Denmark 0.5
1200 1000 800 600 400 200 0 0 1 2 3 4

USD PPP % GDP
1. Includes medical non-durables.
2. Excludes over-the-counter drugs (OTC).
Source: OECD Health Statistics 2015
1 2 http://dx.doi.org/10.1787/888933280639
Figure 2.2. Total (retail and hospital) pharmaceutical spending,

per capita USD PPP, 2013 (or nearest year)
Retail pharmaceuticals Hospital pharmaceuticals
USD PPP
1 000
800 +9%
+10%
+27% +10%
600 +44%
+9%
+33%
400
+19%
200
0
Canada Germany Spain Portugal Australia Czech Republic Korea Estonia
(2011) (2010) (2011)
Note: Data for Portugal are OECD estimates based on adjusted total and retail pharmaceutical spending figures.
Source: OECD Health Statistics 2015.
1 2 http://dx.doi.org/10.1787/888933280642

Figure 2.3. Average annual growth in pharmaceutical and total health expenditure
per capita, in real terms, average across OECD countries, 1990 to 2013
(or nearest year)
% Health expenditure (including pharmaceutical spending) Pharmaceutical expenditure

8
-2
-4
1990 1995 2000 2005 2010 2013
1 2 http://dx.doi.org/10.1787/888933280658
Over the same period, spending on hospital medicines grew faster in several countries
(see Figure 2.4). The multiplication of specialty drugs2 offers a partial explanation, as these
are often delivered in a hospital setting (including in an outpatient department) rather
than dispensed via pharmacies (Hirsch et al., 2014) and are coming to the market with
increasingly high prices.
Figure 2.4. Annual average growth in retail and hospital pharmaceutical

expenditure, in real terms, 2005 and 2013 (or nearest year)
% Retail pharmaceuticals Hospital pharmaceuticals

12
10.1
6
4.7
3.0
2.2 2.2 2.1
1.5 1.5
0
-0.2
-0.8
-1.8
-5.1
-6
Czech Republic Korea Spain Germany Canada Portugal
Note: OECD estimates for Portugal exclude expenditure on other medical products from reported total and retail
spending.
1 2 http://dx.doi.org/10.1787/888933280663
The share of private funding of pharmaceuticals increases

Private funding in the purchasing of pharmaceuticals is greater than for other
categories of health care. On average in OECD countries, 43% of retail pharmaceutical

spending is paid for from private sources (private health insurance or out-of-pocket),
compared with 21% for inpatient and outpatient care. Most of the private spending for
drugs (37%) comes directly from households pockets, reflecting both the high cost-sharing
requirements and the extent of self-consumption of over-the-counter (OTC) medicines (see
the indicator on pharmaceutical expenditure in Chapter 10). Countries such as France,
Germany and Japan report a relatively low private share of pharmaceutical spending of
around 25-30%, whereas the United States and Canada (both countries where private
health insurance plays a large role in financing pharmaceutical spending), as well as
Poland (where spending on OTC drugs is significant), all report more than 60% of the
pharmaceutical bill being covered by private sources.
In a majority of OECD countries, private spending on pharmaceuticals has grown
faster than public spending over the last decade (Figure 2.5). In particular, since 2009,
private spending on drugs did not decline to the same extent as public spending. This
is due in part to an observed shift of some of the cost-burden to households. For
example, in Hungary, the out-of-pocket share of spending on prescribed medicines rose
from 40% to 45% between 2010 and 2013 (Figure 2.6). The Czech Republic and Slovak
Republic also reported increases in the households' share of medicines to 38% and 33%
respectively.
Figure 2.5. Annual growth in public and total retail pharmaceutical spending, OECD countries,
2005-2013
% 2005-2009 2009-2013 %
Total expenditure on pharmaceuticals, annual average growth rate
15 15
10 GRC 10
Private spending Private spending
EST
growth higher SVK KOR growth higher
than public JPN than public CHE
5 5
IRL KOR JPN
USA
HUN NLD NOR

0 0
ESP USA
ISL
NOR EST
-5 -5
Private spending DNK Private spending
growth lower GRC growth lower
LUX
than public PRT than public
-10 -10
-15 -15
-15 -10 -5 0 5 10 15 % -15 -10 -5 0 5 10 15 %
Public expenditure on pharmaceuticals, annual average growth rate
1 2 http://dx.doi.org/10.1787/888933280679
The trends in public and private spending are partly explained by a range of policy
measures adopted by countries to contain public spending on pharmaceuticals, such as
increases in cost-sharing, as well as the increasing use of OTC drugs (usually not
reimbursed) compared with prescription drugs (usually reimbursed) in several countries. In
Slovenia, Poland and Spain, the OTC share of pharmaceutical spending has significantly
increased.

Figure 2.6. Expenditure on retail pharmaceuticals by type of financing, 2013

(or nearest year)
Public Private insurance Private out-of-pocket Other
% of retail pharmaceutical expenditure
100
13
19 18 17
90
28 25
5 32 32 33 33 33 30
33 34
80 1 38 37
7 42 44
14 46 47 48 46
50 51
1
70 1
45
2 58 68
27
60 5
1
1 2 36
50 30
5 6
40 82 80
75
71 69 68 68 67 67 66 65
30 62
58 57 55 54 53 52 52 49 48
43 43
20 38 36 34 32
10
Au gal
St a
Ge nds
i t z um
ec nd
Es a
Po en
Po s
t h ur g
Fr n
ce
ain
nd
ak a
G r p.
Be ce
N o p.
OE ay
26
Fi ia
Sl li a
Sw d
Hu ni a
De ar y
Ic r k
Un C a d
an
i te nad
e
ov r i
re
an
pa
an
n
Re
Re
a
rw
at
an
ee
ra
ed
Cz rla
la
CD
Sl us t
Sp
to
r tu
e
Ko
nm
S w l gi
ng
Ne bo
rm
la
el
Ja
nl
ov
st
h
e
er
A
m
d
xe
Lu

1 2 http://dx.doi.org/10.1787/888933280689
Pharmaceutical expenditure growth is driven by changes in quantity, prices

and therapeutic mix
The increasing demand for medicines and the introduction of new drugs into the
market are the main drivers of spending growth. At the same time, the availability of
generics and biosimilars combined with the introduction and strengthening of cost-
containment policies have exerted a downward pressure on spending in recent years
(Belloni et al., forthcoming).
An increasing demand for pharmaceuticals and new treatment opportunities push

pharmaceutical spending up
The quantity of drugs consumed has increased over time in many therapeutic classes.
Between 2000 and 2013, among countries for which data are available, the use of
antihypertensive, antidiabetic and anti-depressant medications nearly doubled, while the
use of cholesterol-lowering drugs tripled (see indicator on Pharmaceutical consumption
in Chapter 10). These trends reflect an increasing demand for pharmaceuticals, resulting
from the rising prevalence of chronic diseases, population ageing, changes in clinical
practices and coverage extensions, as well as new treatment opportunities.
The prevalence of many chronic diseases, such as cancer, diabetes and mental illness has
increased, leading to an increased demand for medical treatments. Improvements in
diagnosis, leading to earlier recognition of conditions and earlier treatment with
medicines, as well as the development of more medicines (both prescribed and OTC) to
treat common conditions have also contributed to increase the consumption of medicines.
Population ageing also increases the demand for pharmaceutical treatments. With age,
the tendency to develop health conditions which require some kind of medication
increases. As shown in Figure 2.7 for Korea and the Netherlands, per capita spending on
pharmaceuticals increases rapidly with age.

Figure 2.7. Per capita spending on retail pharmaceuticals by age, Korea and the Netherlands,
2011
Males Females
Korea (won) Netherlands (euros)
2 000
1 200 000
1 500
800 000
1 000
400 000
500
0 0
25 24
5 5
10 to 9
15 o 14
20 o 19
30 o 29
35 34
4 0 o 39
45 44
5 0 o 49
55 54
6 0 o 59
65 64
70 6 9
75 o 74
8 0 o 79
Ov 84
85
5 5
10 to 9
15 o 14
25 24
20 o 19
30 o 29
35 34
4 0 o 39
45 44
5 0 o 49
55 54
6 0 o 59
65 64
70 6 9
75 o 74
8 0 o 79
Ov 84
85
an
an
to
to
to
to
er
to
er
to
to
to
to
to
to
to
to
to
t
t
th
th
t
t
t
t
t
t
t
ss
ss
Le
Le
Source: OECD Database on Expenditure by Disease, Age and Gender (unpublished).

1 2 http://dx.doi.org/10.1787/888933280694
New and innovative drugs expand treatment options and increase treatment costs. New
drugs can be new chemical entities or new formulations of existing drugs. Both categories
may increase treatment options, for instance, for previously unmet needs or for new
population targets (e.g. children), increasing the quantity of drugs consumed. While the
approval of new drugs in existing market segments can increase competition and lead to
potential savings, usually new drugs offering therapeutic advantages for patients are
priced higher than their competitors and contribute significantly to pharmaceutical
spending growth.
In recent years, the proliferation of specialty pharmaceuticals with high prices, in
particular oral cancer drugs and immune modulators,3 has played an increasing role in
pharmaceutical spending growth (Express Scripts, 2015; Trish et al., 2014). In the United
States, specialty drugs represented just 1% of total prescriptions but accounted for 25% of
total prescription drug spending in 2012 (Express Scripts, 2015).
Changes in clinical practice guidelines also influence the consumption of pharmaceuticals
upward. Updated guidelines have often recommended earlier treatments, higher dosages
or longer treatment durations for secondary prevention or management of chronic
diseases, leading to increases in volume consumed. This is the case for instance for
guidelines for cholesterol-lowering drugs (e.g. statins), one of the fastest-growing
therapeutic classes of prescription drugs all over the world. Prescription guidelines have
been updated several times since the end of the 1990s, recommending wider screening and
lower lipid level targets as an indication for prescription in Canada, the United Kingdom
and the United States (CIHI, 2012; ACC/AHA, 2014; NICE, 2014).
In a few countries, coverage expansion has contributed to pharmaceutical spending
growth. In the United States, Medicare Part D was introduced in 2006 and the Affordable
Care Act was implemented in 2014, contributing to a substantial reduction in the number
of people uninsured. In Korea, with the establishment of the National Health Insurance

(NHI) in 1989 and successive steps in coverage expansion, pharmaceutical spending

increased rapidly at a rate of more than 10% each year on average between 2000 and 2004
(Yang et al., 2008) and continued to grow since then, albeit at a slower rate.
Cost-containment policies and patent losses have put downward pressure on

spending growth
Pharmaceutical policies have the potential to influence spending trends and the
efficiency (cost-effectiveness) of pharmaceutical spending. In recent years, and in
particular after the economic crisis in 2008, OECD countries have implemented or
strengthened a number of cost-containment policies (Table 2.1).
Table 2.1. Pharmaceutical cost-containment policies introduced since 2008

in a selection of OECD countries
Policies Examples Extent of implementation
Pricing policies One-off cut in ex-factory prices of on-patent Austria, Belgium, Czech Republic, France, Germany,
medicines Greece, Ireland, Italy, Portugal, Spain, Switzerland,
United Kingdom
Implementation of external price referencing or Greece, Portugal, Slovak Republic, Spain, Switzerland
change in the method or basket of countries
Reduction in value-added tax (VAT) rates Austria, Czech Republic, Greece
Reduction of mark-ups for distributors Czech Republic, Estonia, Greece, Hungary, Ireland,
Portugal, Spain
Increase of rebates paid by manufacturers or Germany
distributors
Extra-ordinary price reviews Greece, Ireland, Portugal, Slovak Republic, Spain,
Switzerland
Pressure on prices of branded medicines (e.g. group Canada
purchasing or negotiation)
Reimbursement policies Change in the reference price system (max. Estonia, Greece, Ireland, Portugal, Slovak Republic,
reimbursement price by cluster) Spain
Delisting of products Czech Republic, Greece, Ireland, Portugal, Spain
Increase in cost-sharing Austria, Czech Republic, Estonia, France, Greece,
Ireland, Italy, Portugal, Slovenia, Slovak Republic,
Spain, Sweden
Introduction of health-technology assessment (HTA) Germany
to inform coverage/pricing decisions
Managed-entry agreements Belgium, Italy, United Kingdom
Policies to exploit the potential Implementation of voluntary or mandatory International Belgium, Estonia, France, Italy, Luxembourg,
of off-patent drugs Non-proprietary Name (INN) prescribing Portugal, Slovak Republic, Spain
Incentives for physicians to prescribe generics Belgium, France, Greece, Hungary, Japan
Incentives for pharmacists to dispense generics Belgium, France, Ireland, Japan
Incentives and information for patients to purchase Austria, Estonia, France, Iceland, Ireland,
generics Luxembourg, Portugal, Spain
Pressure on generic prices (e.g. tendering, price cuts) Canada, France, Greece, Portugal
Source: Belloni et al. (forthcoming), complemented by Thomson et al. (2014) on cost-sharing policies.
Since 2008, price cuts have been very common. At least one third of OECD countries
implemented measures to reduce regulated prices of pharmaceuticals. They most often
imposed cuts on ex-factory prices of on-patent and/or generic drugs (e.g. Greece, Ireland,
Portugal and Spain), but many of these countries also reduced distribution margins at least
for some categories of medicines. Germany increased temporarily the mandatory rebates
imposed on pharmaceutical companies from 6% to 16% between 2010 and 2013. In
April 2014, the mandatory rebate was set at 7% for all medicines except generics. In
Canada, several provinces and territories entered in joint price negotiations for brand-
name drugs covered by public plans. Finally, five countries changed VAT rates imposed on

medicine, either to reduce pharmaceutical spending (e.g. Austria, Czech Republic and
Greece) or to increase public revenues (e.g. Estonia, Portugal) resulting in increased spending.
Greece, Portugal, the Slovak Republic, Spain and Switzerland reformed their external
reference price system, expanding or reducing the basket of countries used for international
benchmarking or revising the method for setting prices. For example, the Slovak Republic
included Greece in the basket of benchmarked countries in 2010.
A range of policy measures have shifted some of the burden of pharmaceutical spending to
private payers (households or complementary private insurance). These rarely took the form
of delisting products (i.e. excluding them from reimbursement), with the notable
exceptions of Greece, where 49 medicines were delisted after a price review in 2011, Czech
Republic, Ireland, Portugal and Spain. At least a dozen of countries introduced or increased
user charges for retail prescription drugs (Austria, Czech Republic, Estonia, France, Greece,
Ireland, Italy, Portugal, Slovak Republic, Slovenia, Spain and Sweden) (see Thomson et al.,
2014; and Belloni et al., forthcoming).
Some countries decided to give a greater role to health technology assessment (HTA) in their
reimbursement and/or pricing process. In Germany, for instance, a new law, which took
effect in January 2011, introduced a systematic and formal assessment of the added
therapeutic benefit of new medicines after market entry to allow negotiation of a
reimbursement price where needed. Expected savings for health insurance funds are up to
several million Euros for some individual products (Henschke, 2013).
In parallel, many OECD countries have introduced or expanded the use of managed
entry agreements (MEAs), which are arrangements between the manufacturer and the
payer that allow coverage of drugs subject to defined conditions. Managed-entry
agreements cover a wide range of contractual arrangements, which can be just financial
or performance-based (i.e. reimbursement and pricing conditions are linked to observed
performance of a product in real life). They take the form of price-volume agreements,
coverage with evidence development, performance-based outcome guarantees, patient
access scheme, etc. Their implementation varies across countries. The United
Kingdom, Italy, Germany and Poland have taken the lead in using these arrangements
(Ferrario and Kanavos, 2013). In Italy, the amounts recouped by the government from
manufacturers through performance-based arrangements are modest and represent 5%
of total expenditure for the relevant indications. This is due, at least partly, to high
administrative and management costs of the scheme (Garattini et al., 2015, Navarria et
al., 2015, van de Vooren et al., 2014). Their impact in other jurisdictions has not yet been
evaluated.
Since the onset of the economic crisis, several countries have strengthened their generic
policies (see Table 2.1 and Figures 10.12 and 10.13 in Chapter 10). While no formal
evaluation is available, these policies associated with the patent cliff have certainly
contributed to the significant increase in the generic market share observed over the past
decade in most countries.
From the mid-2000s, a number of blockbuster drugs lost patent protection, contributing to
the decline of pharmaceutical spending growth. Several products worth more than
USD 30 billion a year in US sales lost their patents in 2011-12, among which Plavix
(antiplatelet agent), Lipitor (anti-cholesterol) and Actos (diabetes), which accounted
together for nearly USD 15 billion in sales (Managed Care, 2011).
Patent expiries offer huge opportunities to make savings without affecting the quality
of care. In the United States, for instance, where the generic market is very dynamic, the

price of a generic drug is on average 80 to 85 % lower than that of the brand name product.
In 2012, 84% of all prescriptions filled in the United States were for generic drugs (IMS
Institute for Healthcare Informatics, 2013, see also indicator on Share of generic market
in Chapter 10).
Biosimilars can also lead to significant savings, although the potential is perhaps not
as high as with generics of small molecules, due to longer and costlier development and
production costs. Entry barriers are higher: Europe established a pathway for the
approval of biosimilars in 2005, Japan approved biosimilars regulation in 2009 and Korea
in 2010. The United States approved the legislative framework for licensing follow-on
biologic products in 2010, but the FDA only recently approved the first biosimilar in March
2015. In addition, countries regulations often restrict market growth potential and price
competition. In many countries, prescribing by International Non-proprietary
Names (INN) is not allowed, patients cannot be switched to a biosimilar and substitution
by the pharmacist is not allowed (European Biopharmaceutical Enterprises, 2015).
Drivers of spending growth vary across therapeutic areas

All the drivers of spending growth listed before interact differently across therapeutic
classes, leading to contrasting trends.
In the case of antidiabetic medicines for instance, where use has been steadily
increasing in line with the increasing prevalence of type-2 diabetes, the existence of long-
standing treatments with generic versions resulted in a 'cost of treatment' which remained
relatively stable over a number of years. However, the arrival of new and more expensive
treatments in recent years significantly increased the average daily treatment cost. The
shift from existing medications to new drugs has therefore been the main contributor to
pharmaceutical spending growth in this therapeutic class in the recent period, as shown
for Denmark between 2005 and 2013 in Figure 2.8.
Figure 2.8. Annual growth in sales, volumes and cost per defined daily dosage
(DDD) of antidiabetic drugs, Denmark, 2005-13
% Volume (DDD per 1000/day) Cost per DDD Sales (per capita)
16
12
-4
2005 2006 2007 2008 2009 2010 2011 2012 2013
1 2 http://dx.doi.org/10.1787/888933280701

By contrast, in the class of cholesterol lowering medications, the expiry of the patent
for some of the top selling statins in the mid-2000s and the introduction of generics has led
to a pattern of decreasing treatment costs in many countries in recent years. For example,
costs per defined daily dose (DDD) typically fell by more than 10% per year, on average,
since 2005 in Germany (Figure 2.9).
Figure 2.9. Annual growth in sales, volumes and cost per defined daily dosage
(DDD) of lipid-lowering drugs, Germany, 2005-13
% Volume (DDD per 1000/day) Cost per DDD Sales (per capita)
20
10
-10
-20
-30
2005 2006 2007 2008 2009 2010 2011 2012 2013
1 2 http://dx.doi.org/10.1787/888933280715
The high price of new drugs has been the main driver of spending growth in other
therapeutic areas.
In the area of cancer for instance, the price of specialty medicines has steadily
increased, especially since 2000. In the United States, the median monthly price of cancer
treatment for Medicare patients has increased from around USD 5 000 in 2000-05 to around
USD 10 000 in 2010-15.4 In 2012, 12 out of 13 cancer-approved drugs cost more than
USD 100 000 per year (Light and Kantarjian, 2013). These price increases are observed
everywhere. In Australia, the average reimbursement price per anticancer prescription
drug more than doubled in real terms between 1999-2000 and 2011-12, while the price of
all other prescription drugs only increased by about one-third during that period (Karikios
et al., 2014).
Treatment costs for multiple sclerosis and pulmonary hypertension are also very
high and increasing (Lotvin et al., 2014). The first generation of multiple sclerosis therapies,
originally costing USD 8 000 to USD 11 000 per year in 1993-96, now cost about USD 60 000
per year, reflecting an increase five to seven times higher than prescription drug inflation
over the period 1993-2013. Newer therapies entered the market with a cost 25%-60% higher
than existing ones (Hartung et al., 2015).
In 2013 and 2014, new treatments for hepatitis C became available, posing an
unprecedented challenge to many OECD countries. These medicines represent a great
medical advancement: they are much better tolerated than previous treatments and reach
cure rates of 95% or higher for sub-groups of patients with hepatitis C. For these target
groups, these treatments are even cost-effective. The immediate budget impact of treating
the entire population affected proved to be unaffordable for OECD countries, due to high
prices and high prevalence of the disease. In reaction, many countries sought to reach

agreements with manufacturers to limit the budget impact and to recommend priority use
for the most severely affected patients, generating frustration for physicians, patients and
decision makers alike.
Orphan drugs5 also typically have high prices. The median cost per patient and per
year is 19 times higher for an orphan drug than for a non-orphan drug (EvaluatePharma,
2014). The premium for ultra-rare indications is very high. The number of newly approved
molecular entities classified as orphans has been increasing since the implementation of
policies designed to encourage their development and medicines with orphan designation
now account for one-third of new chemical entities approved by the FDA (IMS Institute for
Healthcare Informatics, 2014).
New challenges in the pharmaceutical market

Changes in the pharmaceutical market, with the increased availability of high-cost
drugs, suggest that future pharmaceutical spending growth may pick up again, instead of
continuing its recent path, at least in some countries. Countries will face a number of
challenges to make new high-cost medicines available to patients, contain spending
growth and ensure value for money.
The IMS Institute for Healthcare Informatics predicts worldwide pharmaceutical
sales6 to be 30% higher in 2018 than in 2013 (IMS Institute for Healthcare Informatics, 2014).
The average annual growth rate is slightly higher than in previous years due to a smaller
number of patent expiries and a higher number of new specialty drugs. Emerging markets,
in addition to the United States, are expected to contribute most of this growth, while
European markets will make more modest contributions.
The United States is the largest pharmaceutical market, accounting for one third of
global sales, and is expected to continue to grow. The IMS Institute for Healthcare
Informatics predicted peaks in US spending growth of 14% in 2014 and 8% in 2015, followed
by annual growth rates of 4-5% until 2018. According to CMS projections, prescription drug
spending is expected to grow at an average annual rate of over 6% per year between 2016
and 2024 (Keehan, 2015).
The largest European markets are predicted to experience lower levels of growth.
According to the IMS Institute for Healthcare Informatics, the top 5 European markets
(Germany, France, the United Kingdom, Italy and Spain) will see annual growth rates of
between 1 and 4% during the period 2014 to 2018. Pharmaceutical spending in the United
Kingdom and Germany should experience the highest growth, while France and Spain will
have zero to negative growth (IMS Institute for Healthcare Informatics, 2014). In an earlier
study, Urbinati et al. (2014) had predicted a decrease in pharmaceutical spending in all
European countries studied except Poland between 2012 and 2016.
Specialty drugs will continue to be a major contributor to pharmaceutical spending growth.
Since 2010, one out of every two FDA approvals is a specialty drug and, as the population
ages, the number of patients eligible for specialty drugs such as treatments for rheumatoid
arthritis and cancer is increasing (Lotvin et al., 2014). Increased spending on these drugs is
projected to account for 53% of total growth in North America between 2013 and 2018,
while in Europe it is expected to account for 94% of the (much slower) growth over the same
period (IMS Institute for Healthcare Informatics, 2014). The huge contribution of specialty
medicines to pharmaceutical spending growth is explained by the fact that there will be
more of them, priced at very high levels, with more patients needing them.

Cancer is the therapeutic area with the highest expected spending growth, driven by
new drug approvals and the increasing incidence of cancer worldwide (IMS Institute for
Healthcare Informatics, 2014). Many orphan drugs approvals are also expected in the years
to come. Their predicted budget impact by 2020 in several European countries ranges
from 4-5% to 9-11% of pharmaceutical spending, depending on the success rate of
products in development (Schey et al., 2011; Hutchings et al., 2014). Another study
estimated that the share of orphan drugs in the worldwide pharmaceutical market for
non-generic prescription drugs is expected to increase from 14% in 2014 to 19% in 2020
(EvaluatePharma, 2014).
High prices of drugs are an important barrier to access, and this does not concern
developing countries only. The results of a recent survey conducted among policy makers
(reported in WHO, 2015) show that policy makers in European countries consider the high
price of drugs as the main challenge to provide access to new medicines given the
budgetary constraints they have. Many drugs, including drugs providing important
benefits, are not available at all, or not accessible to all patients who need them. For
example, as already noted, a lot of countries restricted access to the new hepatitis C
treatments to the most severely affected patients and a few countries have not yet
reimbursed the new medicines at all (e.g. Poland).
A further challenge is that high prices of new medicines do not always appear to be
justified by high clinical benefits (Howard et al., 2015; Light and Kantarjian, 2013). For
example, many new cancer drugs provide small added benefits over existing ones. Among
the 12 new anticancer drugs approved by the FDA in 2012, only one provides survival gains
that exceed two months. Sometimes cancer drugs are used for several indications with
varying levels of efficacy, but the price is usually unique (Bach, 2014). Examining the launch
prices of cancer drugs approved between 1995 and 2013, Howard et al. (2015) observed that
patients and insurers paid USD 54 100 for a year of life gained in 1995, USD 139 100 a
decade later and USD 207 000 in 2013 for the same benefit (in constant 2013 dollars,
adjusting earlier costs for inflation).
Similarly, many orphan drugs do not pass the test of cost-effectiveness. In the
Netherlands, medicines used for the treatment of Pompes and Fabrys disease have been
assessed to cost several million Euros per QALY gained, which triggered a discussion about
the opportunity to maintain health insurance coverage of these products. However, they
were not delisted, since these medicines are used for severe diseases for which no
alternative treatment is available (van den Brink, 2014).
Conclusions
Retail pharmaceutical spending has increased at a slower pace than before or even
decreased in recent years due to patent losses of several blockbusters and cost-
containment policies, while pharmaceutical spending in hospital has increased in most
countries for which data are available.
New high-cost specialty drugs are coming to the market and are expected to account
for 50% or more of pharmaceutical spending growth in the near future. Their increasing
availability, combined with population ageing, suggests that pharmaceutical expenditure
may pick up again after the recent stagnation or decline.
Pharmaceutical spending growth is not necessarily a problem in itself. Medicines play
an important role in the management of a number of chronic diseases (e.g. diabetes,
asthma) and, in some circumstances, they prevent complications and the use of costly
health care services. However, the increasing availability and sky-rocketing prices of new

medicines, especially in cancer, hepatitis C, pulmonary hypertension and multiple

sclerosis, or for rare diseases, have raised a number of questions about accessibility, budget
impact and the legitimacy of such high prices.
While some of these high-price medicines bring great benefits to patients, others
provide only marginal improvement of patients outcomes. In reality, prices seem more
determined by market conditions (high unmet medical need, small population target) than
by any conception of value in terms of clinical or wider benefits for patients. Many of these
medicines are not cost-effective, according to standard thresholds. This challenges both
the static and dynamic efficiency of pharmaceutical spending and raises questions about
the best ways to align societies interests with those of pharmaceutical companies and
investors.
Notes
1. Retail pharmaceuticals are delivered to patients via community pharmacies and other retail outlets.
Pharmaceuticals are also consumed in other care settings primarily the hospital sector where by
convention the pharmaceuticals used are considered as an input to the overall service treatment
and not separately accounted. That said, health accounts do allow for an additional reporting item
to report a total pharmaceutical spending estimate covering all modes of provision. Currently only
about one-third of OECD countries submit such figures.
2. Specialty medicines include most injectable and biologic agents used to treat complex conditions
such as rheumatoid arthritis, multiple sclerosis and cancer and often require special handling or
delivery mechanisms.
3. Biologics used in the treatment of certain types of immunologic and inflammatory diseases,
including rheumatoid arthritis, psoriasis, Crohns disease, and ulcerative colitis.
4. https://www.mskcc.org/research-areas/programs-centers/health-policy-outcomes/cost-drugs.
5. Orphan drugs refer to medicines developed for rare conditions. The United States and the
European Union have implemented policies to encourage private investments in R&D for rare
diseases (e.g. increased market exclusivity) and have consequently defined criteria to be met by a
medicine to be granted an orphan drug status. In the European Union, those criteria are: the
severity of the disease; the fact that it serves an unmet need; and either prevalence below one in
2 000 or a negative expected return on investment.
6. IMS data report market sales at ex-manufacturer prices and do not reflect off-invoice discounts
and rebates (IMS Institute for Healthcare Informatics, 2014). By contrast, pharmaceutical spending,
as reported in the System of Health Accounts, are estimated at retail prices (including VAT) and are
in principle net of off-invoice discounts and rebates. Both sets of data are not directly comparable
but are expected to show more or less consistent trends.
References
ACC/AHA American College of Cardiology/American Heart Association (2014), 2013 ACC/AHA
Guideline on the Treatment of Blood Cholesterol to Reduce Atherosclerotic Cardiovascular Risk in
Adults, Journal of the American College of Cardiology, Vol. 63(25_PA), pp. 2889-2934.
Bach, P.B. (2014), Indication-Specific Pricing for Cancer Drugs, Journal of American Medical Association,
Vol. 312, No. 16, pp. 1629-1630.
Belloni, A., D. Morgan and V. Paris (forthcoming), Pharmaceutical Expenditure and Policies: Past
Trends and Future Challenges, OECD Health Working Papers, OECD Publishing.
CIHI Canadian Institute for Health Information (2012), Drivers of Prescription Drug Spending in Canada,
Canadian Institute for Health Information, Ottawa, Ontario.
European Biopharmaceutical Enterprises (2015), What Pricing and Reimbursement Policies to Use for
Off-patent Biologicals? Results from the EBE 2014 Biological Medicines Policy Survey, Gabi Journal,
Vol. 4, No. 1, pp. 17-24.
EvaluatePharma (2014), Budget-busters: The Shift to High-priced Innovator Drugs in the USA.

EvaluatePharma (2014), Orphan Drug Report 2014,

http://info.evaluategroup.com/rs/evaluatepharmaltd/images/2014OD.pdf.
Express Scripts (2015), The 2014 Drug Trend Report Highlights, March.
Ferrario, A. and P. Kanavos (2013), Managed Entry Agreements for Pharmaceuticals: The European Experience,
EMINet.
Garattini, L., A. Curto and K. van de Vooren (2015), Italian Risk-sharing Agreements on Drugs: Are
They Worthwhile?, European Journal of Health Economics, Vol. 16, pp. 1-3.
Hartung, D. et al. (2015), The Cost of Multiple Sclerosis Drugs in the US and the Pharmaceutical
Industry Too Big to Fail?, Neurology, Vol. 84, May 26, pp. 1-8.
Henschke, C., L. Sundmacher and R. Busse (2013), Structural Changes in the German Pharmaceutical
Market: Price Setting Mechanisms Based on the Early Benefit Evaluation, Health Policy, Vol. 109,
pp. 263-269.
Hirsch, B.R., S. Balu and K.A. Schulman (2014), The Impact of Specialty Pharmaceuticals As Drivers of
Health Care Costs, Health Affairs, Vol. 33, No. 10, pp. 1714-1720.
Howard, D. et al. (2015), Pricing in the Market for Anticancer Drugs, Journal of Economic Perspectives,
Vol. 29, No. 1, pp. 139-162.
Hutchings, A. et al. (2014), Estimating the Budget Impact of Orphan Drugs in Sweden and France 2013-
2020, Orphanet Journal of Rare Diseases, Vol. 9, pp. 9-22.
IMS Institute for Healthcare Informatics (2013), Declining Medicine Use and Costs: For Better or For Worse?
A Review of the Use of Medicines in the United States in 2012.
IMS Institute for Healthcare Informatics (2014), Medicine Use and Shifting Costs of Healthcare. A Review of
the Use of Medicines in the United States in 2013, April 2014.
Karikios, D.J. et al. (2014), Rising Cost of Anticancer Drugs in Australia, Internal Medical Journal, Vol. 44,
No. 5, pp. 458-463.
Keehan, S.K. et al. (2015), National Health Expenditure Projections, 2014-24: Spending Growth Faster
Than Recent Trends, Health Affairs, Vol. 34, No. 8, pp. 1407-1417.
Light, D.W. and H. Kantarjian (2013), Market Spiral Pricing of Cancer Drugs, Cancer, Vol. 15,
No. 119(22), pp. 3900-3902, November.
Lotvin, A.M. et al. (2014), Specialty Medications: Traditional and Novel Tools Can Address Rising
Spending on These Costly Drugs, Health Affairs, Vol. 33, No. 10, pp. 1736-1744.
Managed Care (2011), Patent Cliff: Billions To Be Saved Starting Now,
http://www.managedcaremag.com/content/patent-cliff-billions-be-saved-%E2%80%94-starting-now.
Navarria, A. et al. (2015), Do Current Performance-based Schemes in Italy Really Work? Success Fee:
A Novel Measure for Cost-containment of Drug Expenditure, Value in Health, Vol. 18, pp. 131-136.
NICE National Institute for Health and Care Excellence (2014), NICE Clinical Guideline 181, Lipid
Modification: Cardiovascular Risk Assessment and the Modification of Blood Lipids for the Primary
and Secondary Prevention of Cardiovascular Disease, July 2014.
Schey, C., T. Milanova and A. Hutchings (2011), Estimating the Budget Impact of Orphan Medicines in
Europe: 2010-2020, Orphanet Journal of Rare Diseases, Vol. 6, No. 62, pp. 1-10.
Thomson, S. et al. (2014), Economic Crisis, Health Systems and Health in Europe: Impact and
Implications for Policy, WHO Regional Office for Europe and European Observatory on Health
Systems and Policies.
Trish, E., G. Joyce and D.P. Goldman (2014), Specialty Drug Spending Trends Among Medicare and
Medicare Advantage Enrollees, 2007-11, Health Affairs, Vol. 33, No. 11, November, pp. 2018-2024.
van den Brink, R. (2014), Reimbursement of Orphan Drugs: The Pompe and Fabry Case in the
Netherlands, Orphanet Journal of Rare Diseases, Vol. 9, Suppl. 1, O17.
van de Vooren, K. et al. (2014), Market-access Agreements for Anti-cancer Drugs, Journal of the Royal
Society of Medicine, Vol. 108, No. 5, pp. 166-170.
WHO World Health Organization (2015), Access to New Medicines in Europe: Technical Review of
Policy Initiatives and Opportunities for Collaboration and Research, WHO Regional Office for
Europe, Copenhagen.
Yang B., E. Bae and J. Kim (2008), Economic Evaluation and Pharmaceutical Reimbursement Reform In
South Koreas National Health Insurance, Health Affairs, Vol. 27, No. 1, pp. 179-187.

3. HEALTH STATUS
Life expectancy at birth
Life expectancy by sex and education level
Mortality from cardiovascular diseases
Mortality from cancer
Mortality from transport accidents
Suicide
Infant mortality
Infant health: Low birth weight
Perceived health status
Cancer incidence
authorities. The use of such data by the OECD is without prejudice to the status of the Golan
Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.

3. HEALTH STATUS
Life expectancy at birth continues to increase steadily in Federation (due mainly to the impact of the economic
OECD countries, going up on average by 3 to 4 months each transition in the 1990s and a rise in risk increasing behaviours
year, with no sign of slowing down. These gains in longe- among men, notably rising alcohol consumption).
vity can be attributed to a number of factors including Higher national income (as measured by GDP per capita) is
improved lifestyle and better education, and progress in generally associated with higher life expectancy at birth,
health care. although the relationship is less pronounced at the highest
In 2013, life expectancy on average across OECD countries levels of national income (Figure 3.2). There are also notable
reached 80.5 years, an increase of more than ten years differences in life expectancy between countries with
since 1970 (Figure 3.1). Japan, Spain and Switzerland lead a similar income per capita. For example, Japan, Spain and
large group of 25 OECD countries in which life expectancy Italy have higher, and the United States and the Russian
at birth now exceeds 80 years. A second group, including Federation have lower life expectancies than would be pre-
the United States, Chile and a number of Central and dicted by their GDP per capita alone.
Eastern European countries, has a life expectancy between Figure 3.3 shows the relationship between life expectancy
75 and 80 years. at birth and current health expenditure per capita (excluding
Among OECD countries, Mexico had the lowest life expec- capital investments) across OECD, candidate and partner
tancy in 2013, still slightly below 75 years. Since 2000, life countries. Higher health spending per capita is generally
expectancy in Mexico has increased more slowly than in associated with higher life expectancy at birth, although
other OECD countries, with a gain of just over a year (from this relationship tends to be less pronounced in countries
73.3 to 74.6 years) compared with an average gain of more with the highest health spending per capita. Japan, Spain
than three years across OECD countries. The gap in longe- and Korea stand out as having relatively high life expec-
vity between Mexico and other OECD countries has there- tancies, and the United States and the Russian Federation
fore widened from about four years to six years between relatively low life expectancies, given their levels of health
2000 and 2013. The slow progress in life expectancy in spending.
Mexico is due to a number of factors, including harmful Variation in life expectancy across countries can be
health-related behaviours such as poor nutrition and very explained by many factors beyond national income and
high obesity rates, a lack of progress in reducing mortality total health spending.
from cardiovascular diseases, very high death rates from
road traffic accidents and homicides, as well as persistent
barriers of access to high-quality care.
In the United States, the gains in life expectancy over the Definition and comparability
past few decades have also been more modest than in most
Life expectancy at birth measures how long, on average,
other OECD countries. While life expectancy in the United
people would live based on a given set of age-specific
States used to be one year above the OECD average in 1970,
death rates. However, the actual age-specific death
it is now more than one year below the average. Many
rates of any particular birth cohort cannot be known in
factors can explain these lower gains in life expectancy,
advance. If age-specific death rates are falling (as has
including: 1) the highly fragmented nature of the US health
been the case over the past decades), actual life spans
system, with relatively few resources devoted to public
will be higher than life expectancy calculated with
health and primary care, and a large share of the popula-
current death rates.
tion uninsured; 2) health-related behaviours, including
higher calorie consumption per capita and greater obesity The methodology used to calculate life expectancy
rates, higher consumption of prescription and illegal drugs, can vary slightly between countries. This can change
more deaths from road traffic accidents and higher homi- a countrys estimates by a fraction of a year.
cide rates; and 3) adverse socio-economic conditions Life expectancy at birth for the total population is calcu-
affecting large segments of the US population, with higher lated by the OECD Secretariat for all OECD countries,
rates of poverty and income inequality than in most other using the unweighted average of life expectancy of
OECD countries (National Research Council and Institute of men and women.
Medicine, 2013).
Although the life expectancy in partner countries such as
India, Indonesia, Brazil and China remains well below the
OECD average, these countries have achieved considerable
References
gains in longevity over the past decades, with the level National Research Council and Institute of Medicine,
converging rapidly towards the OECD average. There has S. Woolf and L. Aron (eds) (2013), U.S. Health in Interna-
been much less progress in countries such as South Africa tional Perspective: Shorter Lives, Poorer Health, National
(due mainly to the epidemic of HIV/AIDS), and the Russian Academies Press, Washington, DC.

3. HEALTH STATUS
3.1. Life expectancy at birth, 1970 and 2013 (or nearest years)
Years 1970 2013

90
83.4
83.2
82.9
82.8
82.3
82.2
82.1
82.1
82.0
81.9
81.8
81.8
81.5
81.4
81.4
81.4
81.2
81.1
81.1
81.1
80.9
80.8
80.7
80.5
80.4
80.4
80.0
78.8
78.8
78.3
77.3
80
77.1
76.6
76.5
75.7
75.4
75.2
75.0
74.6
73.9
73.5
70.9
70.7
70
66.5
60
56.8
50
40
lg l
a
h s
lo na
Br i a
M z il
th ia
y
L co
Ru n d o n i a
h ia
nd
N e Gr d a
Ze ds
i te A land
rm d
i te C a
Cz ta e
Po ni a
Sl Tu nd
ak ey
ng .
i a si a
u t In d .
Sw S an
er n
Fr l y
Is d
Lu S w r ael
m en
Ca rea
N o ur g
Ko a y
N e er c e
ng i a
rt y
O E ium
ov 4
C o nm i a
s t ar k
E s e p.
st e
Ic a li a
F i om
Ir n d
B e uga
Hu Rep
d
d hil
ec te
Po an
ric
Au anc
ic
ar
Ge elan
i t z p ai
an
Sl D 3
b
De en
Li at v
K i tr
It a
Fe
a
ov r k
rw
th ee
Co Chi
xe e d
na
w lan
i
p
la
la
ss ne
aR
m
to
I ua
R
ex
d
d us
bo
r
el
nl
Ja
Af
C
a
n
S
So
Un
Un
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en.

1 2 http://dx.doi.org/10.1787/888933280727
3.2. Life expectancy at birth and GDP per capita, 2013 3.3. Life expectancy at birth and health spending
(or latest year) per capita, 2013 (or latest year)
Life expectancy in years Life expectancy in years
85 85
ESP ITA JPN ISL CAN ESP JPN AUS
FRA AUS
SWE CHE NOR ISR ITA FRA SWE CHE
ISR KOR ISL
GRC
FIN NLD GRC NZL LUXCAN AUT NOR
KOR IRL AUT NLD
NZL CRI PRT GBR
80 PRT DEU 80 BEL DEU
CHL SVN DNK SVN FIN DNK
CHL
CZE GBR USA
EST
IRL USA
BEL
TUR POL EST TUR POL
CZE
COL BRA SVK COL CHN SVK
75 HUN 75 HUN
CHN MEX BRA
MEX
LVA LTU LVA LTU
IDN
IDN RUS RUS
70 70
IND R = 0.58 R = 0.51

IND
65 65
0 10 000 20 000 30 000 40 000 50 000 60 000 70 000 0 2 000 4 000 6 000 8 000 10 000
GDP per capita (USD PPP) Health spending per capita (USD PPP)
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en. Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en.
1 2 http://dx.doi.org/10.1787/888933280727 1 2 http://dx.doi.org/10.1787/888933280727
Information on data for Israel: http://oe.cd/israel-disclaimer

3. HEALTH STATUS
There remain large gaps in life expectancy between women people with the highest level of education can expect to live
and men in all OECD countries. On average across OECD six years longer than people with the lowest level of educa-
countries, life expectancy at birth for women reached 83.1 tion at age 30 (53 years versus 47 years). These differences
years in 2013, compared with 77.8 years for men, a gap of in life expectancy by education level are particularly
5.3 years (Figure 3.4). pronounced for men, with an average gap of almost
The gender gap in life expectancy increased substantially eight years. The differences are especially large in Central
in many OECD countries during the 1970s and early 1980s and Eastern European countries (Czech Republic, Estonia,
to reach a peak of almost seven years in the mid-1980s, but Hungary and Poland), where the life expectancy gap
it has narrowed during the past 25 years, reflecting higher between higher and lower educated men is more than
gains in life expectancy among men than among women. ten years. This is largely explained by the greater preva-
This can be attributed at least partly to the narrowing of lence of risk factors among men, such as tobacco and alco-
differences in risk-increasing behaviours, such as smoking, hol use. Differences in other countries such as Portugal,
accompanied by sharp reductions in mortality rates from Sweden, Switzerland and Italy are less pronounced.
cardiovascular diseases among men.
In 2013, the life expectancy for women in OECD countries
ranged from less than 80 years in Turkey, Hungary and
Mexico to more than 85 years in Japan, Spain, France, Italy Definition and comparability
and Switzerland. Life expectancy for men ranged from less
Life expectancy at birth measures how long, on aver-
than 75 years in Mexico, Hungary, Estonia, the Slovak
age, people would live based on a given set of age-
Republic, Poland and Turkey to over 80 years in Switzerland,
specific death rates. However, the actual age-specific
Iceland, Italy, Israel, Japan, Spain, Sweden and Australia.
death rates of any particular birth cohort cannot be
In the United States, the life expectancy for both women known in advance. If age-specific death rates are fall-
and men is now slightly shorter than the OECD average, ing (as has been the case over the past decades),
and the gap with leading countries has been widening. The actual life spans will be higher than life expectancy
life expectancy for US men in 2013 was 4.3 years shorter calculated with current death rates.
than in Switzerland (up from less than three years in 1970);
The methodology used to calculate life expectancy
for US women, it was 5.4 years shorter than in Japan in
can vary slightly between countries. This can change
2013 (there was no gap in 1970). Possible explanations for
a countrys estimates by a fraction of a year.
this slower progress are provided under the indicator Life
expectancy at birth. To calculate life expectancies by education level,
detailed data on deaths by sex, age and education
Among OECD countries, the gender gap in life expectancy is
level are needed. However, not all countries have
relatively narrow in Iceland, Israel, Sweden, the Netherlands,
information on education as part of their deaths data.
New Zealand and the United Kingdom (a gap of less than
Data linkage to another source (e.g. a census) which
four years), but much larger in Estonia (around nine years),
does have information on education may be required
Poland (around eight years), the Slovak Republic and Hungary
(Corsini, 2010).
(around seven years).
Life expectancy in OECD countries varies not only by gender,
but also by socio-economic status as measured, for
instance, by education level (Figure 3.5). Higher education
level not only provides the means to improve the socio- References
economic conditions in which people live and work, but Corsini, V. (2010), Highly Educated Men and Women Likely
may also promote the adoption of healthier lifestyles and to Live Longer: Life Expectancy by Educational Attain-
facilitate access to appropriate health care. On average ment, Eurostat Statistics in Focus 24/2010, European Com-
among 15 OECD countries for which recent data are available, mission, Luxembourg.

3. HEALTH STATUS
3.4. Life expectancy at birth by sex, 2013 (or latest year)
Years Men Women

90
86.6
86.1
85.6
85.2
85.1
85.0
84.3
84.1
84.0
84.0
83.9
83.9
83.8
83.8
83.8
83.7
83.6
83.6
83.2
83.2
83.2
83.2
83.1
83.1
82.9
82.4
81.7
81.4
81.3
81.2
81.2
80.7
80.5
80.3
80.3
80.2
80.2
80.2
80.1
80.1
79.8
79.8
79.5
79.5
79.4
79.3
79.2
79.1
79.0
79.0
78.7
78.6
78.6
78.5
78.3
78.1
78.0
77.8
77.6
77.4
77.2
80
76.4
76.3
75.2
73.7
73.0
72.9
72.8
72.2
71.7
70
60
50
40
nd
n
i t z a in
Fr y
S w ael
m n
Au c e
Ic i a
d
Ca a
da
Be gal
No r g
ay
Ze s
th ece
i te Au d
Ge nd
ng ia
Fi m
ec es
Po i a
M ry
Ir e d
Po any
O E um
Sl 3 4
S e
o
De ni a
E s p.
nd
ak ey
H u e p.
w nd
l
re
l
ar
xe e d e
pa
an
an
an
ic
l
i t e Chi
n
It a
d s tr
Re
do
a
u
Sl Turk
rw
an
Cz t at
ra
na
la
la
la
CD
r
S w Sp
r tu
to
e
ex
Ko
nm
i
ng
bo
N e Gr e
rm
Ja
el
N e er l a
al
nl
Is
lg
ov
er
st
h
d
Ki
ov
Lu
Un
Un
Note: Countries are ranked in descending order of life expectancy for the whole population.
1 2 http://dx.doi.org/10.1787/888933280737
3.5. Gap in life expectancy at age 30 by sex and educational level, 2012 (or latest year)
Men Women
3.6 Italy 1.8
3.7 Sweden 3.0
4.3 Portugal 1.8
4.5 Netherlands (2011) 4.2
5.1 Norway 3.9
5.2 Mexico 4.6
5.4 Denmark 3.8
5.5 Finland 3.5
5.8 Israel 4.7
6.9 Slovak Rep. 3.5
7.7 OECD15 4.2
9.1 Slovenia 4.4
11.6 Poland 5.0
12.1 Hungary 5.5
15.0 Estonia 8.1
17.8 Czech Rep. 5.2
20 15 10 5 0 0 2 4 6 8 10
Gap in years Gap in years
Note: The figures show the gap in the expected years of life remaining at age 30 between adults with the highest level (tertiary education) and the
lowest level (below upper secondary education) of education.
Source: Eurostat database complemented with national data for Israel, Mexico and Netherlands.
1 2 http://dx.doi.org/10.1787/888933280737

3. HEALTH STATUS
Despite substantial declines in recent decades, cardiovas- There are large variations in cerebrovascular disease mor-
cular diseases remain the main cause of mortality in most tality rates across countries (Figure 3.7). The Slovak Repub-
OECD countries, accounting for nearly one-third (32.3%) of lic and Hungary report a cerebrovascular mortality that is
all deaths in 2013. Prospects for further reductions may be more than three times higher than that of Switzerland,
hampered by a rise in certain risk factors such as obesity Canada and France, and have the highest mortality rates
and diabetes (OECD, 2015). Cardiovascular diseases cover a for both IHD and cerebrovascular disease. The high preva-
range of illnesses related to the circulatory system, includ- lence of risk factors common to both diseases (such as
ing ischemic heart disease (often referred to as heart smoking and high blood pressure) may explain this link.
attack) and cerebrovascular diseases such as stroke. Since 1990, cerebrovascular disease mortality has
Ischemic heart disease (IHD) is caused by the accumulation decreased in all OECD countries, although to a lesser extent
of fatty deposits lining the inner wall of a coronary artery, in Poland and the Slovak Republic. On average, the mortal-
restricting blood flow to the heart. IHD alone was responsi- ity burden from cerebrovascular disease has halved across
ble for nearly 20% of all deaths in OECD countries in 2013. OECD countries. In Estonia, Luxembourg and Portugal, the
However, mortality from IHD varies considerably across rates have been cut by at least two-thirds, although in Estonia
countries (Figure 3.6). Central and Eastern European coun- this is partly due to a change in death recording practices
tries report the highest IHD mortality rates; Japan, France with a greater recording of other related causes of death
and Korea report the lowest rates. Across OECD countries, such as hypertension. As with IHD, the reduction in mortal-
IHD mortality rates in 2013 were around 84% higher for ity from cerebrovascular disease can be attributed at least
men than women. partly to a reduction in risk factors as well as improve-
IHD mortality rates have declined in nearly all OECD coun- ments in medical treatments (see indicator Mortality
tries, with an average reduction of 45% since 1990, contrib- following stroke in Chapter 8), but rising obesity and diabe-
uting greatly to gains in life expectancy, particularly among tes threatens progress in tackling cerebrovascular disease
men. The decline has been most remarkable in Denmark, (OECD, 2015).
the Netherlands, and Norway, where rates fell by two-
thirds or more. Declining tobacco consumption contributed
significantly to reducing the incidence of IHD (see Indicator
Tobacco consumption among adults in Chapter 4), and con- Definition and comparability
sequently to reducing mortality rates. Improvements in med-
ical care have also contributed to reduced mortality rates (see Mortality rates are based on numbers of deaths regis-
the indicators on Cardiac procedures in Chapter 6 and tered in a country in a year divided by the size of the
Mortality following acute myocardial infarction in corresponding population. The rates have been
Chapter 8). directly age-standardised to the 2010 OECD popula-
tion to remove variations arising from differences in
In Korea, IHD mortality rates have increased substantially
age structures across countries and over time. The
since 1990, although they remain low compared with
source is the WHO Mortality Database.
nearly all other OECD countries and have started to fall
after peaking in 2006. The initial rise in IHD mortality rates Deaths from ischemic heart disease are classified to
in Korea has been attributed to changes in lifestyle and ICD-10 codes I20-I25, and cerebrovascular disease to
dietary patterns as well as environmental factors at the I60-I69.
time of birth, with people born between 1940 and 1950 facing
higher relative risks. In 2006, Korea introduced a Compre-
hensive Plan to tackle cardiovascular diseases that encom-
passed prevention and primary care as well as better acute
care, contributing to the reduction in mortality in recent References
years (OECD, 2012).
Murray, C.J.L. et al. (2015), Global, Regional, and National
Cerebrovascular disease was the underlying cause for
Disability-adjusted Life Years (DALYs) for 306 Diseases
about 7% of all deaths in OECD countries in 2013. Cerebro-
and Injuries and Healthy Life Expectancy (HALE) for 188
vascular disease refers to a group of diseases that relate to
Countries, 1990-2013: Quantifying the Epidemiological
problems with the blood vessels that supply the brain.
Transition, The Lancet, published online: 26 August 2015.
Common manifestations of cerebrovascular disease
include ischemic stroke, which develops when the brain's OECD (2015), Cardiovascular Disease and Diabetes: Policies for
blood supply is blocked or interrupted, and haemorrhagic Better Health and Quality of Care, OECD Publishing, Paris,
stroke which occurs when blood leaks from blood vessels http://dx.doi.org/10.1787/9789264233010-en.
into the surface of the brain. In addition to being an impor- OECD (2012), OECD Reviews of Health Care Quality: Korea:
tant cause of mortality, the disability burden from stroke Raising Standards, OECD Publishing, Paris,
and other cerebrovascular diseases is also substantial http://dx.doi.org/10.1787/9789264173446-en.
(Murray et al., 2015).

3. HEALTH STATUS
3.6. Ischemic heart disease mortality, 2013 and change 1990-2013 (or nearest years)
2013 Change 1990-2013
35 Japan -38
43 France -52
43 Korea 63
50 Netherlands -73
51 Portugal -57
56 Spain -47
63 Belgium -48
66 Luxembourg -56
68 Chile -58
70 Israel -68
71 Denmark -77
78 Norway -70
82 Switzerland -49
83 Greece -36
84 Italy -38
94 Slovenia -47
95 Canada -59
98 United Kingdom -67
98 Australia -64
105 Sweden -62
106 Poland -27
115 Germany -48
117 OECD34 -45
128 United States -50
133 Iceland -46
136 Ireland -59
138 New Zealand -53
140 Austria -38
140 Mexico -1
146 Turkey n.a.
154 Finland -55
260 Estonia -60
260 Czech Rep. -41
297 Hungary -10
404 Slovak Rep. 9
500 400 300 200 100 0 -100 -50 0 50 100

Age-standardised rates per 100 000 population Change in %

1 2 http://dx.doi.org/10.1787/888933280741
3.7. Cerebrovascular disease mortality, 2013 and change 1990-2013 (or nearest years)
2013 Change 1990-2013
37 Switzerland -60
38 Canada -52
38 France -56
44 Israel -58
44 United States -42
45 Spain -69
46 Luxembourg -77
49 Austria -69
49 Norway -64
51 Belgium -55
51 Netherlands -52
51 Australia -55
52 Germany -63
53 United Kingdom -62
54 Japan -61
54 Denmark -51
55 Sweden -50
60 Mexico -38
60 Iceland -41
61 Ireland -54
65 Finland -57
66 OECD34 -54
67 New Zealand -44
67 Italy -54
68 Estonia -79
77 Korea -56
80 Chile -42
86 Poland -14
88 Portugal -73
92 Slovenia -51
97 Czech Rep. -69
101 Turkey n.a.
106 Greece -51
118 Hungary -54
137 Slovak Rep. -2
150 100 50 0 -100 -50 0 50 100

Age-standardised rates per 100 000 population Change in %

1 2 http://dx.doi.org/10.1787/888933280741

3. HEALTH STATUS
Cancer is the second leading cause of mortality in OECD Colorectal cancer is a major cause of cancer mortality
countries after cardiovascular diseases, accounting for 25% among both men and women (second-highest cause of
of all deaths in 2013, up from 15% in 1960. In a number of cancer mortality in men and third in women). In 2013,
countries, cancer is now the most frequent cause of death. colorectal cancer mortality was lowest in Mexico and
The rising share of deaths due to cancer reflects the fact Turkey, and highest in Hungary and the Slovak Republic
that mortality from other causes, particularly cardiovascu- (see indicator Survival and mortality for colorectal cancer
lar diseases, has been declining more rapidly than mortal- in Chapter 8).
ity from cancer. Prostate cancer has become the most common cancer
There are more than 100 different types of cancers, with among men in many OECD countries, particularly among
most named for the organ in which they start. For a large men aged 65 years and over. Mortality from prostate cancer
number of cancer types, the risk of developing the disease remains lower than for lung cancer in all countries except
rises with age. While genetics is a risk factor, only about 5% in Chile and Mexico, where it is the leading cause of cancer
to 10% of all cancers are inherited. Modifiable risk factors deaths in men, and in some Nordic countries (Iceland,
such as smoking, obesity, lack of exercise and excess sun Norway and Sweden) where mortality from prostate and
exposure, as well as environmental exposures, explain up lung cancer are almost equal. Mortality rates from prostate
to 90-95% of all cancer cases (Anand et al., 2008). Preven- cancer in 2013 were lowest in Japan and Korea, and highest
tion, early detection and treatment remain at the forefront in Estonia and Iceland.
in the battle to reduce the burden of cancer (OECD, 2013). In most OECD countries, cancer-related mortality rates
In 2013, the average rate of mortality attributable to cancer have fallen since 1990. On average, rates fell by 17%
across OECD countries was just over 200 per 100 000 popula- between 1990 and 2013. Substantial declines in mortality
tion (Figure 3.8). Mortality due to cancer was lowest in from stomach cancer, colorectal cancer, lung cancer for
Mexico, Turkey, Finland, Switzerland and Japan, with rates men, breast, cervical and ovarian cancer for women, as
less than 180 per 100 000 population. Hungary, Slovenia, well as prostate cancer for men contributed to this reduc-
the Slovak Republic and Denmark bear the highest cancer tion. However, these gains were partially offset by increases
mortality burden, with rates in excess of 240 per in the number of deaths due to cancer of the liver, skin and
100 000 population. pancreas for both sexes, as well as lung cancer for women.
Mortality due to cancer is consistently higher for men than
for women in all countries. The gender gap is particularly
wide in Korea, Turkey, Estonia, Spain and Portugal, with
rates among men more than twice those for women. This
Definition and comparability
gender gap can be explained partly by the greater preva-
lence of risk factors among men, notably smoking rates. Mortality rates are based on numbers of deaths regis-
Among men, lung cancer imposes the highest mortality tered in a country in a year divided by the size of the
burden, accounting for 26% of all cancer-related deaths corresponding population. The rates have been
(Figure 3.9). In Turkey, Greece, Poland, Hungary and directly age-standardised to the 2010 OECD popula-
Belgium, this percentage was over 30%. For women, lung tion to remove variations arising from differences in
cancer accounted for 17% of all cancer-related deaths. In age structures across countries and over time. The
many countries, lung cancer mortality rates for men have source is the WHO Mortality Database. Deaths from all
decreased over the last 20 years. But lung cancer mortality cancers are classified to ICD-10 codes C00-C97. The
has risen for women in several countries such as France international comparability of cancer mortality data
and Spain where it has more than doubled since 1990. can be affected by differences in medical training
These conflicting trends are, to a large degree, explained by and practices as well as in death certification across
the high number of females who started smoking several countries.
decades later than males (in the 1980s and 1990s).
Breast cancer is the second most common cause of cancer
mortality in women in many OECD countries. While there
has been an increase in the incidence of breast cancer over
the past decade, mortality has declined in most countries References
due to earlier diagnosis and better treatment. Mortality Anand, P. et al. (2008), Cancer is a Preventable Disease that
from breast cancer increased somewhat in Korea and Requires Major Lifestyle Changes, Pharmaceutical
Japan, although the rates there remained the lowest in Research, Vol. 25, No. 9, pp. 2097-2116.
2013. Mortality rates from breast cancer in 2013 were high- OECD (2013), Cancer Care: Assuring Quality to Improve Sur-
est in Denmark, Hungary, Belgium, Ireland, Slovenia and vival, OECD Publishing, Paris,
the Netherlands (see indicator Screening, survival and http://dx.doi.org/10.1787/9789264181052-en.
mortality for breast cancer in Chapter 8).

3. HEALTH STATUS
3.8. Cancer mortality, 2013 (or nearest year)
Men Women Total

Age-standardised rates per 100 000 population
400
350
300
250
200
150
100
50
0
No al
nd
da
ec ds
Tu o
F y
Lu we l
m n
Gr s
Fr e
Ze d
i t e Ir e d
K i and
i t z nd
g
Po in
Au y
Un us t a
St a
ce
Ge nc e
OE ly
Be 3 4
y
Ca m
t h om
E s p.
Po i a
De nd
S l e p.
y
ak k
Hu ni a
e
e
il
an
re
A ri
ov ar
ar
xe d e
w an
an
ic
pa
ur
g
ra
i te r al
n
It a
a
Re
iu
rk
at
rw
ee
Ch
na
C z lan
la
la
S w inl a
st
CD
r tu
to
Sp
e
R
ex
Ko
Sl nm
Ne ngd
ng
bo
rm
a
l
N e Ic el
Ja
al
Is
lg
ov
er
h
M
er
S
d
Un
1 2 http://dx.doi.org/10.1787/888933280758
3.9. Main causes of cancer deaths among men and women in OECD countries, 2013
Men Women
Stomach Stomach
6% 5%
Colorectal Others Colorectal

11% 40% 11%
Others
36%
Liver
Pancreas
6%
7%
Pancreas
6%
Lung
17%
Lung
Prostate Ovary
26%
9% 5%
Breast
15%

1 2 http://dx.doi.org/10.1787/888933280758

3. HEALTH STATUS
Injuries from transport accidents most of which are due rose during the 1990s as the economy and the number of
to road traffic are a major public health problem in OECD vehicles grew and has remained relatively high since then
countries, causing the premature deaths of more than (Nghiem et al., 2013).
100 000 people in 2013 (more than 1% of all deaths). Almost Declines in mortality rates for vulnerable road users such
three-quarters of these deaths occurred among men. In as pedestrians, cyclists and motorcyclists were substan-
addition, more than 5 million people were injured in road tially less than those for car occupants. Reductions in
accidents. The direct and indirect financial costs of trans- deaths among pedestrians, cyclists and motorcyclists have
port accidents are substantial, with estimates ranging from levelled-off and some increases have been recorded. As a
1 to 3% of GDP annually (OECD/ITF, 2015). consequence, road safety priorities in many countries have
Most fatal traffic injuries occur in passenger vehicles, recently shifted to vulnerable road users in urban areas
although other road users also face substantial risks. In (OECD/ITF, 2015).
Korea, Japan, Israel and Poland, pedestrians account for The economic crisis has contributed to the reduction in
over one third of all road user fatalities. Cyclists in the road traffic deaths in many countries, by reducing the dis-
Netherlands and motorcyclists in Greece, Italy and France tance travelled (especially by young men and by trucks).
account for over one fourth of road traffic accident deaths However, this impact is likely to be short-lived and, over the
in these countries (OECD/ITF, 2015). longer term, effective road safety policies will remain the
The average OECD mortality rate due to transport acci- primary contributor to reduced mortality (OECD/ITF, 2015).
dents was 7 per 100 000 population in 2013 (Figure 3.10).
There is considerable variation between countries with
transport accidents claiming more than five times as
many lives per 100 000 population in Mexico compared to Definition and comparability
the United Kingdom and Sweden. Mortality rates from road
transport accidents were also relatively high in Korea, Chile Mortality rates are based on numbers of deaths regis-
and the United States. tered in a country in a year divided by the size of the
Much transport accident injury and mortality is prevent- corresponding population. The rates have been
able. Road safety for car occupants has increased greatly directly age-standardised to the 2010 OECD popula-
over the past decades in many countries through improve- tion to remove variations arising from differences in
ments of road systems, education and prevention cam- age structures across countries and over time. The
paigns as well as vehicle design. In addition, the adoption source is the WHO Mortality Database.
of new laws and regulations and the enforcement of these Deaths from transport accidents are classified to
laws to improve compliance with speed limits, seatbelt use ICD-10 codes V01-V89.
and drink-driving rules have had a major impact on reduc-
ing the burden of road transport accidents.
As a result, deaths due to transport accidents have
decreased in almost all countries over the last few decades.
References
Since 1990, the average OECD mortality rate due to transport
accidents has fallen by more than 70% (Figure 3.11). These Nghiem, H., L. Connelly and S. Gargett (2013), Are Road
gains are even more impressive when considering the Traffic Crash Fatality Rates Converging among OECD
increase in the number of vehicle kilometers travelled over Countries?, Accident Analysis & Prevention, Vol. 52,
this period (OECD/ITF, 2015). Chile is the only country pp. 162-170.
where deaths due to transport accidents have increased. In OECD/ITF (2015), IRTAD Road Safety 2015 Annual Report,
1990, Chiles mortality rate was comparatively low, but then OECD Publishing.

3. HEALTH STATUS
3.10. Transport accident mortality, 2013 (or nearest year)

Men Women Total
30
25
20
15
10
0
m d1
Fi g 1
C a al
s
Ir e d
es
it z an
nd
Ge a in
Au el
Es ly
B e ni a
Z e p.
Hu d
Sl ar y
il e
M a
o
Sw m
De en
th ar k
No ny
ay
Lu c el a
Fr d
ce
A u ium
Po a li a
OE da
e 34
ak a
Po .
nd
Un Gr y
St e
p
nd
ri
e
ov ni
i te ec
re
n
an
an
ic
ra
g
It a
Re
do
xe a n
ur
rk
at
rw
an
Ch
ed
Sw Jap
na
la
la
la
st
Cz CD
to
r tu
Sp
Sl ve
Ne ch R
ex
Ko
N e nm
ng
e
r
rm
la
nl
al
Is
lg
bo
Tu
er
st
ng
o
er
d
Ki
w
d
i te
Un
1. Three-year average.
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en and Ministry of Health for New Zealand.
1 2 http://dx.doi.org/10.1787/888933280766
3.11. Trends in transport accident mortality, selected OECD countries, 1990-2013
Canada Chile Mexico France Greece Spain

United States OECD United Kingdom OECD
Age-standardised rates per 100 000 population Age-standardised rates per 100 000 population
30 30
25 25
20 20
15 15
10 10
5 5
0 0
0
00
02
04
06
08
10
12
00
02
04
06
08
10
12
9
9
9
9
9
9
9
9
20
20
20
20
19
19
20
19
20
20
19
20
19
19
20
19
20
19
20
19
20
20
19
20

1 2 http://dx.doi.org/10.1787/888933280766

3. HEALTH STATUS
Suicide
Suicide is a significant cause of death in many OECD coun- also needed to remove the stigma associated with seeking
tries, accounting for over 150 000 deaths in 2013. A complex care (OECD, 2014).
set of reasons may explain why some people choose to Previous studies have shown a strong link between adverse
attempt or commit suicide. A high proportion of people economic conditions and higher levels of suicide (Van Gool
who have attempted or committed suicide are suffering and Pearson, 2014). Suicide rates rose slightly at the start of
from psychiatric disorders such as severe depression, bipolar the economic crisis in 2008-2009 in a number of countries,
disorder and schizophrenia. The social context in which an but this trend did not persist in most. In Greece, suicide
individual lives is also important. Low income, alcohol and rates were stable in 2009 and 2010, but have increased since
drug abuse, unemployment and social isolation are all 2011 (Figure 3.13). All countries need to continue monitoring
associated with higher rates of suicide. developments closely in order to be able to respond quickly,
Figure 3.12 shows that suicide rates in 2013 were lowest in including monitoring high-risk populations such as the
Turkey, Greece, Mexico, Italy and Israel, at seven or fewer unemployed and those with psychiatric disorders (see
deaths per 100 000 population, although the number of sui- indicator Mental health care in Chapter 8).
cides in certain countries may be under-reported because
of the stigma associated with the act or data unreliability
associated with reporting criteria (see Definition and com-
parability). Korea had the highest suicide rate with nearly Definition and comparability
30 deaths per 100 000 population, followed by Japan,
Hungary and Slovenia with nearly 20 deaths per 100 000 The World Health Organization defines suicide as an
population. Mortality rates from suicide are three-to-four act deliberately initiated and performed by a person in
times greater for men than for women across OECD coun- the full knowledge or expectation of its fatal outcome.
tries (Figure 3.12). In Poland and the Slovak Republic, men Comparability of data between countries is affected by
are seven times more likely to commit suicide than a number of reporting criteria, including how a per-
women. The gender gap is narrower for attempted suicides, sons intention of killing themselves is ascertained,
reflecting the fact that women tend to use less fatal methods who is responsible for completing the death certifi-
than men. Suicide is also related to age, with young people cate, whether a forensic investigation is carried out,
aged under 25 and elderly people especially at risk. While and the provisions for confidentiality of the cause of
suicide rates among the latter have generally declined over death. Caution is required therefore in interpreting
the past two decades, less progress has been observed variations across countries.
among younger people.
Mortality rates are based on numbers of deaths regis-
Since 1990, suicide rates have decreased by around 30% tered in a country in a year divided by the size of the
across OECD countries, with the rates being halved in corresponding population. The rates have been
countries such as Hungary and Finland (Figure 3.13). In directly age-standardised to the 2010 OECD popula-
Estonia, after an initial rise in the early 1990s, the rates tion to remove variations arising from differences in
have also fallen sharply. On the other hand, death rates age structures across countries and over time. The
from suicides have increased in Korea and Japan. In Japan, source is the WHO Mortality Database. Deaths from sui-
there was a sharp rise in the mid-to-late 1990s, coinciding cide are classified to ICD-10 codes X60-X84.
with the Asian financial crisis, but rates have started to
come down in recent years. In Korea, suicide rates rose
steadily over the past two decades peaking around 2010,
before starting to come down (Lim et al., 2014). Suicide is References
the number one cause of death among teenagers in Korea.
Lim, D. et al. (2014), Trends in the Leading Causes of Death
Suicide is often linked with depression and the abuse of alco-
in Korea, 1983-2012, Journal of Korean Medical Science,
hol and other substances. Early detection of these psycho-
Vol. 29, No. 12, pp. 1597-1603.
social problems in high-risk groups by families and health
professionals is an important part of suicide prevention OECD (2014), Making Mental Health Count: The Social and Eco-
campaigns, together with the provision of effective support nomic costs of Neglecting Mental Health Care, OECD Publish-
and treatment. Many countries are developing national ing, Paris, http://dx.doi.org/10.1787/9789264208445-en.
strategies for prevention, focusing on at-risk groups. Mental Van Gool, K. and M. Pearson (2014), Health, Austerity and
health services in Korea lag behind those of other countries Economic Crisis: Assessing the Short-term Impact in
with fragmented support, focused largely around institutions, OECD Countries, OECD Health Working Papers, No. 76,
and insufficient or ineffective support services provided to OECD Publishing, Paris,
those who remain in the community. Further efforts are http://dx.doi.org/10.1787/5jxx71lt1zg6-en.

3. HEALTH STATUS
Suicide
3.12. Suicide, 2013 (or nearest year)

Men Women Total
45
40
35
30
25
20
15
10
0
Au rg 1
OE nd 1
m al
Un S w d
th ada
Ge nds
St n
Au s
Sl um
y
o
ly
el
a
ce
K i ain
C li a
il e
Ze d
it z 3 4
ec ia
nd
Hu n
y
P om
No ny
ay
De and
ak k
I c p.
Po .
Fi e
Es d
B e ni a
ia
p
e
ke
re
ov ar
ar
n
i te ede
w lan
an
pa
ic
ra
xe u g
en
It a
C z s tr
Re
Re
at
rw
ee
an
Ch
ra
la
la
Sw CD
u
to
p
ex
Ko
Sl nm
i
d
ng
a
r
Ne an
rm
la
nl
Ja
al
Is
lg
ov
bo
Tu
Lu or t
N e Ir e
er
st
el
ng
Gr
Fr
h
M
er
d
d
i te
Un
1. Three-year average.
1 2 http://dx.doi.org/10.1787/888933280778
3.13. Trends in suicide, selected OECD countries, 1990-2013
Estonia Finland Greece Japan Korea Mexico

Hungary OECD United States OECD
Age-standardised rates per 100 000 population Age-standardised rates per 100 000 population
45 45
40 40
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
0
00
02
04
06
08
10
12
00
02
04
06
08
10
12
9
9
9
9
9
9
20
20
20
20
19
19
20
19
20
19
20
19
19
20
20
19
20
19
20
20
19
19
20
20

1 2 http://dx.doi.org/10.1787/888933280778

3. HEALTH STATUS
Infant mortality
Infant mortality, the rate at which babies and children of and with infant mortality more than double that for White
less than one year of age die, reflects the effect of economic women (10.9 vs. 5.1 in 2012) (NCHS, 2015).
and social conditions on the health of mothers and new- Many studies use infant mortality as a health outcome to
borns, the social environment, individual lifestyles as well examine the effect of a variety of medical and non-medical
as the characteristics and effectiveness of health systems. determinants of health. Although most analyses show that
In most OECD countries, infant mortality is low and there is higher health spending tends to be associated with lower
little difference in rates (Figure 3.14). In 2013, the average in infant mortality, the fact that some countries with a high
OECD countries was less than four deaths per 1 000 live level of health expenditure do not exhibit low levels of
births, with rates being the lowest in Iceland, Slovenia, infant mortality suggests that more health spending is not
Finland, Estonia and Japan. A small group of OECD coun- necessarily required to obtain better results (Retzlaff-Roberts
tries still have comparatively high infant mortality (Mexico, et al., 2004).
Turkey and Chile), although in these three countries infant
mortality has reduced considerably over the past few
decades (Figure 3.15).
In some large partner countries (India, South Africa and Definition and comparability
Indonesia), infant mortality remains above 20 deaths per
1 000 live births. In India, one-in-twenty-five children die The infant mortality rate is the number of deaths of
before their first birthday, although the rates have fallen children under one year of age, expressed per
sharply over the past few decades. Infant mortality rates 1 000 live births. Some of the international variation
have also reduced greatly in Indonesia. in infant mortality rates is related to variations in reg-
istering practices for very premature infants. While
In OECD countries, around two-thirds of the deaths that
some countries register all live births including very
occur during the first year of life are neonatal deaths (i.e.,
small babies with low odds of survival, several coun-
during the first four weeks). Birth defects, prematurity and
tries apply a minimum threshold of a gestation period
other conditions arising during pregnancy are the main
of 22 weeks (or a birth weight threshold of 500 grams)
factors contributing to neonatal mortality in developed
for babies to be registered as live births (Euro-Peristat,
countries. With an increasing number of women deferring
2013). To remove this data comparability limitation,
childbearing and a rise in multiple births linked with fertil-
the data presented in this section are now based on a
ity treatments, the number of pre-term births has tended
minimum threshold of 22 weeks of gestation period
to increase (see indicator Infant health: low birth weight). In
(or 500 grams birth weight) for a majority of OECD
a number of higher-income countries, this has contributed to
countries that have provided these data. However, the
a levelling-off of the downward trend in infant mortality
data for some countries (e.g., Canada and Australia)
over the past few years. For deaths beyond a month (post-
continue to be based on all registered live births,
neonatal mortality), there tends to be a greater range of
resulting in some over-estimation.
causes the most common being SIDS (sudden infant
death syndrome), birth defects, infections and accidents.
In the United States, the reduction in infant mortality has
been slower than in most other OECD countries. In 2000,
the US rate was below the OECD average, but it is now References
higher (Figure 3.14). One of the explanations that have been Euro-Peristat (2013), European Perinatal Health Report: The
given for that the high rate of infant mortality in the United Health and Care of Pregnant Women and their Babies in
States is that it is based on a more complete registration of 2010, Luxembourg.
very premature and low birth weight babies than in many
Joseph, K.S. et al. (2012), Influence of Definition Based
other countries (Joseph et al., 2012). In order to remove the
Versus Pragmatic Registration on International Compar-
impact of differences in registration practices of very small
isons of Perinatal and Infant Mortality: Population Based
babies, the figures shown in Figure 3.14 for a majority of
Retrospective study, British Medical Journal, Vol. 344,
countries (including the United States) exclude deaths of
e746.
babies of less than 22 weeks of gestation period or
500 grams birth weight. The rate in the United States none- NCHS (2015), Health, United States, 2014, with Special
theless remains higher than the OECD average, especially Feature on Adults Aged 55-64, NCHS, Hyattsville, United
for post-neonatal mortality (deaths after one month) which States.
is greater in the United States than in most other OECD Retzlaff-Roberts, D., C. Chang and R. Rubin (2004), Techni-
countries. There are large differences in infant mortality cal Efficiency in the Use of Health Care Resources: A
among racial groups in the United States, with Black Comparison of OECD Countries, Health Policy, Vol. 69,
women more likely to give birth to low birth weight infants, pp. 55-72.

3. HEALTH STATUS
Infant mortality
3.14. Infant mortality, 2013 (or nearest year)

Deaths per 1 000 live births
45
41.4
40
32.8
35
30
24.5
25
20
17.5
13.0
15
12.3
10.9
10.2
10
8.4
8.2
7.0
5.1
5.0
5.0
4.8
4.5
4.4
4.4
4.0
3.8
3.7
3.7
3.6
3.6
3.5
3.5
3.5
5
3.3
3.1
2.9
2.9
2.9
2.8
2.6
2.5
2.5
2.5
2.4
2.4
2.3
2.0
2.0
1.7
1.7
1.3
0
Sl and 1
Fr r g 1
Un w s tr al
ng d
h n
bo d
Au el
Ze ds
Po i a
i te ng a
ak es
s s Ch .
i a il e
Tu ica
Ch y
Br a
C o ex i l
l co
S o do bi a
h ia
In a
F i ni a
E s and
Ja ia
No pan
S y
S w p ain
n m p.
Is r k
rm ia
y
Ko l y
r a
i te i t z a li a
B e dom
L u I ium
G ce
th ce
N e O E C ni a
N e h er 3 4
L a nd
C a and
Sl d S t r y
a
aR .
M a zi
p
st ed
a
an
Un u d
e
in
di
Po r e
ric
ec de
K i an
m an
Au tug
n
ra
ut nes
tv
Ge s tr
It a
Re
De Re
a
rk
ov a t
rw
an
Li ree
w lan
na
a
t D
u
In o m
e
to
ua
Co F
Cz e
nl
d er l
xe r e l
al
l
lg
Af
ov
el
n
Ic
Ru
S
Note: The data for most countries are based on a minimum threshold of 22 weeks of gestation period (or 500 grams birthweight) to remove the impact
of different registration practices of extremely premature babies across countries.
1. Three-year average (2011-13).
1 2 http://dx.doi.org/10.1787/888933280782
3.15. Trends in infant mortality, selected OECD countries, 2000-13

Chile Mexico Turkey United States OECD
Deaths per 1 000 live births
40
35
30
25
20
15
10
0
00
01
02
03
04
05
06
07
08
09
10
11
12
13
20
20
20
20
20
20
20
20
20
20
20
20
20
20

1 2 http://dx.doi.org/10.1787/888933280782

3. HEALTH STATUS
Low birth weight defined as newborns weighing less than weight babies between 2009 and 2012 were linked to the
2 500 grams is an important indicator of infant health economic crisis and its impact on unemployment rates and
because of the close relationship between birth weight and lowering family incomes in Greece (Kentikelenis, 2014). In
infant morbidity and mortality. There are two categories of 2013, the rate came down to levels observed before the crisis.
low birth weight babies: those occurring as a result of Comparisons of different population groups within coun-
restricted foetal growth and those resulting from pre-term tries indicate that the proportion of low birth weight
birth. Low birth weight infants have a greater risk of poor infants may also be influenced by differences in education
health or death, require a longer period of hospitalisation level, income and associated living conditions. In the
after birth, and are more likely to develop significant United States, there are marked differences in the proportion
disabilities. Risk factors for low birth weight include mater- of low birth weight infants among racial groups, with black
nal smoking, excessive alcohol consumption, poor nutri- infants having a rate almost double that of white infants
tion, low body mass index, lower socio-economic status, and (13% versus 7% in 2013) (NCHS, 2015). Similar differences
having had in-vitro fertilisation treatment and multiple have also been observed among the indigenous and non-
births. indigenous populations in Australia, Mexico and New
One in 15 babies born in OECD countries in 2013 or 6.6% of Zealand, often reflecting the disadvantaged living condi-
all births weighed less than 2 500 grams at birth tions of many of these mothers.
(Figure 3.16). The proportions of low-weight births were The proportion of low birth weight infants is also generally
lowest in Nordic countries (Iceland, Finland, Sweden, higher among women who smoke than for non-smokers.
Norway, with the exception of Denmark) and Estonia, with
less than 5% of live births defined as low birth weight.
Japan, had the highest proportion of low birth weight
infants among OECD countries, with rates close to 10%, Definition and comparability
followed by Greece, Hungary and Portugal.
Low birth weight is defined by the World Health
Despite the widespread use of a 2 500 grams limit for low Organization (WHO) as the weight of an infant at birth
birth weight, physiological variations in size occur across of less than 2 500 grams (5.5 pounds) irrespective of
different countries and population groups, and these need the gestational age of the infant. This threshold is
to be taken into account when interpreting differences based on epidemiological observations regarding the
(Euro-Peristat, 2013). Some populations may have lower increased risk of death to the infant and serves for
than average birth weights than others because of genetic international comparative health statistics. The number
differences. of low weight births is expressed as a percentage of
In almost all OECD countries, the proportion of low birth total live births.
weight infants has increased over the past two decades,
mainly due to increases in pre-term births (Euro-Peristat,
2013). There are several reasons for this rise, including a
growing number of multiple pregnancies mainly as a result References
of the rise in fertility treatment, and a rise in maternal age Delnord, M. et al. (2015), What Contributes to Disparities
(Delnord et al., 2015). Another factor which may explain the in the Preterm Birth Rate in European Countries?, Cur-
rise in low birth weight infants is the increased use of deliv- rent Opinion in Obstetrics and Gynecology, Vol. 27, No. 2,
ery management techniques such as induction of labour pp. 133-142, April.
and caesarean delivery, which have increased the survival Euro-Peristat (2013), European Perinatal Health Report: The
rates of low birth weight babies. Health and Care of Pregnant Women and their Babies in 2010,
Korea, Greece, Spain, Portugal and Japan have seen large Luxembourg.
increases of low birth weight babies over the past two Kentikelenis, A. (2014), Greeces Health Crisis: From
decades, although the proportions remain below the OECD Austerity to Denialism, The Lancet, Vol. 383, No. 9918,
average in Korea (Figure 3.17). In Japan, this increase can be pp. 748-753.
explained by changes in obstetric interventions, in particu-
NCHS National Center for Health Statistics (2015), Health,
lar the greater use of caesarean sections, along with
United States, 2014, With Special Feature on Adults Aged
changes in maternal socio-demographic and behavioural
55-64, NCHS, Hyattsville, United States.
factors (Yorifuji et al., 2012). In Greece, the rise in the pro-
portion of low birth weight babies started in the mid-1990s, Yorifuji, T. et al. (2012), Trends of Preterm Birth and Low
well before the economic crisis, and peaked in 2010. Some Birth Weight in Japan: A One Hospital-Based Study,
researchers have suggested that the high rates of low birth BMC Pregnancy and Childbirth, Vol. 12:162.

3. HEALTH STATUS
3.16. Low birth weight infants, 2013 (or nearest year)

% of newborns weighing less than 2 500 g
10
9.6
8.9
8.8
8.7
8.4
8.1
8.0
7.9
7.8
7.6
8
7.5
7.3
7.0
7.0
6.8
6.8
6.7
6.6
6.6
6.6
6.2
6.1
6.0
6.0
6.0
6.0
5.8
5.8
6
5.5
5.5
4.6
4.3
4.3
4.1
3.7
0
d
No en
Sw ia
M a
Hu g al
d
ay
o
th nd
Ze s
d
il e
OE nd
Sl nd
Ca ia
Au da
Ge a li a
i t z ny
Fr 4
h s
Au e
i te Belg a
K i ium
xe om
De l y
g
ak k
p.
ain
S l
Tu .
Po ke y
Gr y
ce
n
e
p
w nd
re
e
i
c
ov ar
ar
an
an
an
ic
pa
ur
3
n
en
i t e sr a
r
It a
Re
Re
rw
Cz t at
Ch
an
ee
ed
na
Sw rma
a
la
la
st
CD
to
r tu
Sp
ex
Ko
Sl nm
Lu ngd
ng
bo
r
r
el
nl
l
N e er l a
al
Ja
ov
e
Po
er
st
Un I
Es
Ic
r
Fi
m
I
ec
d
Ne
d
Un

1 2 http://dx.doi.org/10.1787/888933280794
3.17. Trends in low birth weight infants, selected OECD countries, 1990-2013
Greece Japan Korea Portugal Spain OECD
% of newborns weighing less than 2 500 g
10
0
1990 1995 2000 2005 2010

1 2 http://dx.doi.org/10.1787/888933280794

3. HEALTH STATUS
Most OECD countries conduct regular health surveys which Greater emphasis on public health and disease prevention
allow respondents to report on different aspects of their among disadvantaged groups, and improving access to
health. A commonly asked question relates to self- health services may contribute to further improvements in
perceived health status, of the type: How is your health in population health status in general and reducing health
general?. Despite the subjective nature of this question, inequalities.
indicators of perceived general health have been found to
be a good predictor of peoples future health care use and
mortality (DeSalvo et al., 2005).
For the purpose of international comparisons, cross-country
variations in perceived health status are difficult to interpret Perceived health status reflects peoples overall percep-
because responses may be affected by the formulation of tion of their health. Survey respondents are typically
survey questions and responses, and by social and cultural asked a question such as: How is your health in
factors. In addition, since older people report poor health general? very good, good, fair, poor, very poor.
more often than younger people, countries with a larger
Caution is required in making cross-country compari-
proportion of aged persons will also have a lower propor-
sons of perceived health status, for at least two reasons.
tion of people reporting to be in good health.
First, peoples assessment of their health is subjective
With these limitations in mind, in almost all OECD coun- and can be affected by cultural factors. Second, there
tries, a majority of adults reports being in good health are variations in the question and answer categories
(Figure 3.18). New Zealand, Canada, the United States and used to measure perceived health across surveys and
Australia are the four leading countries, with more than countries. In particular, the response scale used in the
85% of people reporting to be in good health. However, the United States, Canada, New Zealand, Australia and
response categories offered to survey respondents in these Chile is asymmetric (skewed on the positive side),
four countries are different from those used in European including the following response categories: excellent,
countries and Asian OECD countries, which introduce an very good, good, fair, poor. The data in OECD Health
upward bias (see box on Definition and comparability). Statistics refer to respondents answering one of the
On the other hand, less than half of adults in Japan, Korea three positive responses (excellent, very good or
and Portugal rate their health as being good. The propor- good). By contrast, in most other OECD countries, the
tion is also relatively low in Estonia, Hungary, Poland, Chile response scale is symmetric, with response categories
and the Czech Republic, where less than 60% of adults con- being: very good, good, fair, poor, very poor. The
sider themselves to be in good health. data reported from these countries refer only to the
In all OECD countries, men are more likely than women to first two categories (very good, good). In Israel, the
report being in good health, except in Australia, New Zealand, scale is symmetric but there is no middle category
Canada and United Kingdom where the proportion is related to fair health. Such differences in response
almost equal. As expected, peoples rating of their own categories biases upward the results from those
health tends to decline with age. In many countries, there countries that are using an asymmetric scale or a
is a particularly marked decline in how people rate their symmetric scale but without any middle category.
health after age 45 and a further decline after age 65 Self-reported health by income level is reported for
(OECD, 2015). the first quintile (lowest 20% of income group) and
There are large disparities in self-reported health across the fifth quintile (highest 20%). Depending on the
different socio-economic groups, as measured by income surveys, the income may relate either to the individ-
or education level. Figure 3.19 shows that, in all countries, ual or the household (in which case the income is
people with a lower level of income tend to report poorer equivalised to take into account the number of persons
health than people with higher income, although the gap in the household).
varies. On average across OECD countries, nearly 80% of
people in the highest income quintile report being in good
health, compared with just over 60% for people in the low-
est income group. These disparities may be explained by
differences in living and working conditions, as well as References
differences in lifestyles (e.g., smoking, harmful alcohol
DeSalvo, K.B. et al. (2005), Predicting Mortality and Health-
drinking, physical inactivity, and obesity problems). In
care Utilization with a Single Question, Health Services
addition, people in low-income households may have limited
Research, Vol. 40, pp. 1234-1246.
access to certain health services for financial or other reasons
(see Chapter 7 on Access to care). A reverse causal link is OECD (2015), OECD Health Statistics 2015, online, OECD Pub-
also possible, with poor health status leading to lower lishing, Paris, www.oecd.org/health/health-data.
employment and lower income.

3. HEALTH STATUS
3.18. Perceived health status among adults, 2013 (or nearest year)
Good or very good Fair Bad or very bad
% of population aged 15 years and over
100 2 3 3 4 4 4 4 6 5 8 9 7 6
8 9 11 8 8 8 9 9 8 11
8 9 12 13 12 13 14
9 16 16 15 16
90 11 20 19
14 15 16
18 16 19
17 16 18 20 21 20
80 22 22 24 29
21 27 25
21 22 35
28
27
70 27
31
35
60 49 49
50
90 89 88 85
82 81 81
40 80
77 76 76 74 74 74 72 72 72 69 69 68 67 66 66 65 65 65
30 60 59 58 57
53
46
20
35 35
10
0
Un C a n 1
St 1
Au tes 1
Ir e 1
Ic l 1
il e 1
Ja l
S nd
i t z en
nd
N d
Be s
m
er y
Fr y
Es r y
Po ni a
a
xe om
D e ur g
k
O E a in
Au 3
Tu a
ce
ov It al y
G e e p.
Hu nd
Sl ny
Fi ia
d
p.
ga
d
li a
a
nd
e
d eec
ri
re
ar
an
C z nl a n
pa
3
e
en
i te ad
Re
iu
an
a
rk
N e or w
an
S w wed
a
ra
la
la
la
st
CD
to
r tu
Sp
R
ra
Ch
Ko
nm
Lu ngd
ng
bo
rm
el
la
a
lg
ov
Po
er
al
Is
i t e Gr
ak
h
st
m
Ze
ec
Ki
th
d
w
Sl
Ne
Un
1. Results for these countries are not directly comparable with those for other countries, due to methodological differences in the survey
questionnaire resulting in an upward bias. In Israel, there is no category related to fair health.
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en (EU-SILC for European countries).
1 2 http://dx.doi.org/10.1787/888933280801
3.19. Perceived health status by income level, 2013 (or nearest year)
Highest income Lowest income
% of population aged 15 years and over reporting to be in good health
100
96
95
96
94
91
93
89
89
89
87
86
90
85
84
84
83
83
83
82
81
80
80
79
79
78
78
77
77
77
77
76
76
75
80
74
74
73
73
70
69
69
69
69
67
67
67
67
66
70
64
63
62
62
62
61
60
60
58
60
54
54
53
52
52
49
48
50 42
40
40
39
40
31
28
30
20
10
0
Un C a n 1
St 1
Au tes 1
Ir e 1
Ic l 1
il e 1
Ja l
S nd
i t z en
nd
N d
Be s
m
er y
e
xe om
D e ur g
k
O E a in
Au 3
Tu a
Fr y
ce
ov It al y
G e e p.
Hu nd
Es r y
Po ni a
a
Sl ny
Fi ia
d
p.
ga
d
li a
a
nd
e
d eec
ri
re
ar
an
C z nl a n
pa
3
e
en
i te ad
Re
iu
an
a
rk
N e or w
an
S w wed
a
ra
la
la
la
st
CD
to
r tu
Sp
R
ra
Ch
Ko
nm
Lu ngd
ng
bo
rm
el
la
a
lg
ov
Po
er
al
Is
i t e Gr
ak
h
st
m
Ze
ec
Ki
th
d
w
Sl
Ne
Un
Note: Countries are ranked in descending order of perceived health status for the whole population.
1. Results for these countries are not directly comparable with those for other countries, due to methodological differences in the survey
questionnaire resulting in an upward bias. In Israel, data by income group relate to the employed population.
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en (EU-SILC for European countries).
1 2 http://dx.doi.org/10.1787/888933280801

3. HEALTH STATUS
Cancer incidence
In 2012, an estimated 5.8 million new cases of cancer were new cancer diagnoses in men in 2012, followed by lung
diagnosed in OECD countries, 54% (around 3.1 million) (14%) and colorectal (12%). Similar to breast cancer, the
occurring in men and 46% (around 2.7 million) in women. causes of prostate cancer are not well-understood but age,
The most common were breast cancer (12.9% of all new ethnic origin, family history, obesity, lack of exercise and
cancer cases) and prostate cancer (12.8%), followed by lung poor nutrition are the main risk factors. Incidence in 2012
cancer (12.3%) and colorectal cancer (11.9%). These four was highest in Norway, Sweden, Australia and Ireland, with
cancers represented half of the estimated overall burden of rates more than 50% higher than the OECD average
cancer in OECD countries (Ferlay et al., 2014). (Figure 3.22). Greece had the lowest rates, followed by
Large variations exist in cancer incidence across OECD Mexico, Korea and Japan. Prostate cancer incidence rates
countries. Cancer incidence rates are highest in Denmark, have increased in most OECD countries since the late 1990s
Australia, Belgium, Norway, United States, Ireland, Korea, with increased use of prostate specific antigen (PSA) tests
Netherlands and France registering more than 300 new having led to greater detection (Ferlay et al., 2014). Differ-
cancer cases per 100 000 population in 2012 (Figure 3.20). ences between countries rates can be partly attributed to
The lowest rates were reported in some Latin American differences in the use of PSA testing. Mortality rates from
and Mediterranean countries such as Mexico, Greece, Chile prostate cancer have decreased in some OECD countries as
and Turkey, with around 200 new cases or less per a consequence of early detection and improvements in
100 000 population. These variations reflect not only varia- treatments (see indicator Mortality from cancer).
tions in the prevalence of risk factors for cancer, but also
national policies regarding cancer screening and differ-
ences in quality of reporting.
Cancer incidence was higher for men in all OECD countries Definition and comparability
in 2012 except in Mexico. However, the gender gap varies
Cancer incidence rates are based on numbers of new
widely across countries. In Turkey, Estonia and Spain, inci-
cases of cancer registered in a country in a year per
dence among men were around 60% higher than among
100 000 population. The rates have been directly age-
women, whereas in the United Kingdom, Denmark and Ice-
standardised based on Segis world population to
land, the gap was less than 10%.
remove variations arising from differences in age struc-
Breast was by far the most common primary sites in tures across countries and over time. The data come
women (28% on average), followed by colorectal (12%), lung from the International Agency for Research on Cancer
(10%), and cervical (3%). The causes of breast cancer are not (IARC), GLOBOCAN 2012, available at globocan.iarc.fr.
fully understood, but the risk factors include age, family GLOBOCAN estimates for 2012 may differ from
history, breast density, exposure to oestrogen, being over- national estimates due to differences in methods.
weight or obese, alcohol intake, radiation and hormone
Cancer registration is well established in most OECD
replacement therapy. Incidence rates in 2012 were highest
countries, although the quality and completeness of
in Belgium, Denmark and Netherlands, with rates 25% or
cancer registry data may vary. In some countries, cancer
more than the OECD average (Figure 3.21). Chile and Mexico
registries only cover subnational areas. The interna-
had the lowest rate, followed by Turkey and Greece. The
tional comparability of cancer incidence data can also
variation in breast cancer incidence across OECD countries
be affected by differences in medical training and
may be at least partly attributed to variation in the extent
practice.
and type of screening activities. Although mortality rates
for breast cancer have declined in most OECD countries The incidence of all cancers is classified to ICD-10
since the 1990s due to earlier detection and improvements codes C00-C97 (excluding non-melanoma skin
in treatments, breast cancer continues to be the leading cancer C44). Breast cancer corresponds to C50, and
cause of death from cancer among women (see Indicator prostate cancer to C61.
Mortality from cancer in Chapter 3 and Screening, survival
and mortality from breast cancer in Chapter 8).
Prostate cancer has become the most commonly diagnosed
cancer among men in almost all OECD countries, except in References
Hungary, Poland, Turkey and Greece where lung cancer is Ferlay, J. et al. (2014), Cancer Incidence and Mortality
still predominant, and in Japan and Korea where colorectal Worldwide: Sources, Methods and Major Patterns in
cancer is the main cancer among men. On average across GLOBOCAN 2012, International Journal of Cancer, Vol. 136,
OECD countries, prostate cancer accounted for 24% of all No. 5, pp. E359-E386.

3. HEALTH STATUS
Cancer incidence
3.20. All cancers incidence, men and women, 2012

Men Women Total
400
350
300
250
200
150
100
50
0
Sp l
e nd
U n Ir e a
St d
No es
A u ium
Un O d en
Ge ael
Hu d
Sl d a
il e
Au in
i t z ar y
Z e p.
Ca d
B e ay
De a li a
k
ia
th nce
Gr o
ce
ey
Po n
Es d
Po ni a
Fi ia
Ki 3 4
ov om
p.
m y
g
Ic y
Sw d
ga
nd
an
re
xe I t a l
ar
i te lan
an
an
n
pa
an
ic
ur
en
r
a
Re
e
rk
at
rw
ee
Ch
na
Cz rla
la
st
r
to
r tu
Ne ch R
ex
Ko
nm
Sl gd
S w ng
bo
r
rm
Ne Fr a
e
la
el
al
Ja
nl
Is
lg
ov
i te EC
Tu
st
ak
e
M
er
n
d
w
d
Lu
Source: International Agency for Research on Cancer (IARC), GLOBOCAN 2012. 1 2 http://dx.doi.org/10.1787/888933280811
3.21. Breast cancer incidence, women, 2012 3.22. Prostate cancer incidence, men, 2012
Chile 35 Greece 20
Mexico 35 Mexico 27
Turkey 39 Korea 30
Greece 44 Japan 30
Japan 52 Poland 36
Estonia 52 Hungary 38
Poland 52 Turkey 41
Korea 52 Slovak Rep. 50
Hungary 55 Chile 52
Slovak Rep. 58 Portugal 64
Slovenia 67 Spain 65
Spain 67 Italy 68
Portugal 68 Czech Rep. 72
Austria 68 United Kingdom 73
Czech Rep. 70 Austria 75
Norway 73 OECD34 76
OECD34 74 Germany 77
Canada 80 Luxembourg 79
Sweden 80 Slovenia 83
Israel 81 Netherlands 83
Switzerland 83 Israel 84
New Zealand 85 Canada 89
Australia 86 Belgium 91
Luxembourg 89 Denmark 91
Finland 89 New Zealand 92
France 90 Estonia 94
Italy 91 Finland 97
Germany 92 France 98
Ireland 92 United States 98
United States 93 Iceland 107
United Kingdom 95 Switzerland 107
Iceland 96 Ireland 114
Netherlands 99 Australia 115
Denmark 105 Sweden 119
Belgium 112 Norway 130
0 25 50 75 100 125 150 0 25 50 75 100 125 150

Age-standardised rates per 100 000 women Age-standardised rates per 100 000 men
Source: International Agency for Research on Cancer (IARC), GLOBOCAN Source: International Agency for Research on Cancer (IARC), GLOBOCAN
2012. 1 2 http://dx.doi.org/10.1787/888933280811 2012. 1 2 http://dx.doi.org/10.1787/888933280811

4. NON-MEDICAL DETERMINANTS
OF HEALTH
Tobacco consumption among adults
Alcohol consumption among adults
Fruit and vegetable consumption among adults
Obesity among adults
Overweight and obesity among children
international law.

4. NON-MEDICAL DETERMINANTS OF HEALTH
Tobacco kills nearly 6 million people each year, of whom tobacco policies, new strategies such as plain packaging for
more than 5 million are from direct tobacco use and more tobacco products aimed to restrict branding have been
than 600 000 are non-smokers exposed to second-hand implemented (e.g. in Australia) and are being adopted by an
smoke (WHO, 2015). Tobacco is a major risk factor for at increasing number of countries.
least two of the leading causes of premature mortality Several studies provide strong evidence of socio-economic
cardiovascular diseases and cancer, increasing the risk of differences in smoking and mortality (Mackenbach et al.,
heart attack, stroke, lung cancer, cancers of the larynx and 2008). People in less affluent social groups have a greater
mouth, and pancreatic cancer, among others. In addition, it prevalence and intensity of smoking, a higher all-cause
is a dominant contributing factor for respiratory diseases mortality rate and lower rates of cancer survival (Woods et
such as chronic obstructive pulmonary disease (US DHHS, al., 2006). The influence of smoking as a determinant of
2014). Smoking in pregnancy can lead to low birth weight overall health inequalities is such that, if the entire popula-
and illness among infants. Smoking remains the largest tion was non-smoking, mortality differences between
avoidable risk factor for health in OECD countries and social groups would be halved (Jha et al., 2006).
worldwide.
The proportion of daily smokers in the adult population
varies greatly, even between neighboring countries
(Figure 4.1). Nineteen of 34 OECD countries had less than
20% of the adult population smoking daily in 2013. Rates The proportion of daily smokers is defined as the per-
were lowest in Sweden, Iceland, Mexico and Australia (less centage of the population aged 15 years and over who
than 13%). Rates were also less than 13% in Brazil, Colombia, report smoking every day. International comparability
and India, although the proportion of smokers among men is limited due to the lack of standardisation in the
is high, up to 23% in India. On the other hand, smoking measurement of smoking habits in health interview
rates remain high in Greece in both men and women, and surveys across OECD countries. Variations remain in
in Latvia and Indonesia where more than one in two men the age groups surveyed, the wording of questions,
smoke daily. Smoking prevalence is higher among men response categories and survey methodologies (e.g. in
than among women in all OECD countries except in Sweden a number of countries, respondents are asked if they
and Iceland. The gender gap in smoking rates is particularly smoke regularly, rather than daily). Self-reports of
large in Korea, Japan, and Turkey, as well as in the Russian behaviours may also suffer from social desirability bias
Federation, India, Indonesia, Latvia, Lithuania, South Africa that may potentially limit cross-country comparisons.
and China (Figure 4.1).
Smoking rates across most OECD countries have shown a
marked decline, although other forms of smokeless
References
tobacco use, such as snuff in Sweden, are not taken into
account. On average, smoking rates have decreased by Jha, P. et al. (2006), Social Inequalities in Male Mortality,
about one fourth since 2000, from 26% in 2000 to 20% in and in Male Mortality from Smoking: Indirect Estimation
2013. Large reductions occurred in Norway, Iceland, Swe- from National Death Rates in England and Wales,
den, Denmark and Ireland, as well as in India. Poland, and North America, The Lancet, Vol. 368,
In the period that followed World War II, smoking rates No. 9533, pp. 367-370.
were very high among men (50% or more) in most OECD Mackenbach, J.P. et al. (2008), Socio-economic Inequalities
countries through to the 1960s and 1970s, while the 1980s in Health in 22 European Countries, New England Journal
and the 1990s were characterised by a marked downturn in of Medicine, Vol. 358, pp. 2468-2481.
tobacco consumption. Non-OECD countries and emerging OECD (2015), Cardiovascular Disease and Diabetes: Policies for
economies stand at an earlier phase of the evolution of Better Health and Quality of Care, OECD Publishing, Paris,
smoking, with high rates and a wide gender gap. In OECD http://dx.doi.org/10.1787/9789264233010-en.
countries, much of the decline in tobacco use can be attrib-
US DHHS US Department of Health and Human Services
uted to policies aimed at reducing tobacco consumption
(2014), The Health Consequences of Smoking 50 Years of
through public awareness campaigns, advertising bans,
Progress: A Report of the Surgeon General, Atlanta.
increased taxation, and restriction of smoking in public
spaces and restaurants, in response to rising rates of WHO (2015), Tobacco, Fact Sheet No. 339, available at:
tobacco-related diseases. More stringent policies and www.who.int/mediacentre/factsheets/fs339/en/index.html.
higher level of taxes have led to bigger reductions in smok- Woods, L.M., B. Rachet and M.P. Coleman (2006), Origins of
ing rates between 1996 and 2011 in OECD countries (OECD, Socio-economic Inequalities in Cancer Survival: A Review,
2015). As governments continue to reinforce their anti- Annals of Oncology, Vol. 17, No. 1, pp. 5-19.

0
5
10
15
20
25
30
35
40
0
10
20
30
40
50
60
70
Sw Sw
ed ed
e 19
e
Br n
Br n 11
a 18
Ic a z il
Ic z il el
el 11
a M and
M nd 22
ex 11 Co ex i
ic lo co
o 13 m
In 12 bi
A 25 In a
Un u s di a
i te tr a 13 Un u di A
d li a 20 i te s tr a
d al
St
a 13
C a tes 19
Co S t a ia
na 14 s t tes
aR
Ne No da 22
w rw 15 Ca ic a
n

32
HEALTH AT A GLANCE 2015 OECD 2015

Lu Zea ay
15
Ne No ada
xe l a w rw
m nd 25 Lu e ay Z
bo 16
u xe a l a
Fi rg 26 m nd
nl 16 bo
an
23 F i ur g
I d 16 nl
D sr a 22
an
N e enm el 16 I d
th a D e sr a
er r k 31 Ne n el
la 17
Men
th ma
Po nd 29 er r k
r tu s 19
l

Po and
Be ga 21 r tu s
lg l 19
iu Be ga
24 lg l
Ir e m
19 i
2000
la
n 33
Ir e u m
la
Sl J a d
o p 19
Sl J a nd
So vak an 27 ov p
u t Re 19 ak an
h p. 22 R

Af 20 S o O E C e p.
OE ric a 24 ut D
Un CD 20 h 34
Af
i te 3 26
Un
i te ric
d Ko 3 20 d Ko a
Ki re 26
Women
Ki re
S w ngd a 20 Sw g a n
27
i t z om i t z dom
er 20 er
Ge l an
26 Sl l an
Note: Countries are ranked in ascending order of smoking rates for the whole population.
rm d 20 ov d
an Ge en
25 rm ia
Cz I y 21
an
ec t al
h y 24
21 Cz I y
L i Re ec t al
t h p. 27 h y
2013
ua 22 L i Re
Au ni a 32 t h p.
22
ua
st
24
Au ni a
s
Total
4.1. Daily smoking in adults, 2013 (or nearest year)
Po r ia
la 23
28
Po tr ia
Tu nd 24 la
rk
e 32
Tu nd
rk
Sp y 24 e
Ru F r a i n 32 Sp y
s s an 24 Ru F a i n
ia ce 27 s s an r
n
4.2. Change in daily smoking in adults, 2000 and 2013 (or nearest years)
24 ia ce
Fe n
d 35 Fe
Ch . 24
Ch d.
E s in a 31
t 26 E s in a
Hu oni 30 Hu toni
ng a 26 ng a
ar 30
y 27
ar
Ch Ch y
i 33 il e
In t v
L a le 30 In L a t v
do i a 33 do i a
ne 34 ne
Gr s i a 36 Gr s i a
ee 38 ee
ce 35 ce
39
69
1 2 http://dx.doi.org/10.1787/888933280827
1 2 http://dx.doi.org/10.1787/888933280827
The health burden related to harmful alcohol consump- related health problems, and indirect measures such as the
tion, both in terms of morbidity and mortality, is consider- dissemination of information on alcohol-related harm
able in most parts of the world (Rehm et al., 2009; WHO, (WHO, 2010). The OECD used this as a starting point to
2014; OECD, 2015). Alcohol use is associated with numerous identify a set of policy options to be assessed in an economic
harmful health and social consequences, including an evaluation, and showed that several policies have the
increased risk of a range of cancers, stroke, and liver cirrhosis, potential to reduce heavy drinking, regular or episodic, as
among others. Foetal exposure to alcohol increases the risk well as alcohol dependence. Governments seeking to tackle
of birth defects and intellectual impairment. Alcohol also binge drinking and other types of alcohol abuse can use a
contributes to death and disability through accidents and range of policies that have proven to be effective, including
injuries, assault, violence, homicide and suicide. The use of counselling heavy drinkers, stepping up enforcement of
alcohol is estimated to cause more than 3.3 million deaths drinking-and-driving laws, as well as raising taxes, raising
worldwide per year, and accounts for 5.1% of the global prices, and increasing the regulation of the marketing of
burden of disease (WHO, 2014). Health care costs associated alcoholic drinks (OECD, 2015).
with excessive drinking in the United States are estimated
at USD 25.6 billion (Bouchery et al., 2011). In the Russian
Federation, alcohol misuse was a major contributing factor to
the sharp rise in premature mortality and decline in life Definition and comparability
expectancy during the 1990s (OECD, 2012). The use of alco-
hol also has broader societal consequences, accounting for Alcohol consumption is defined as annual sales of
large losses in work productivity through absenteeism and pure alcohol in litres per person aged 15 years and
premature mortality, as well as injuries and death among over. The methodology to convert alcoholic drinks to
non-drinkers (e.g. because of traffic accidents caused by pure alcohol may differ across countries. Official statis-
drivers under the influence of alcohol). tics do not include unrecorded alcohol consumption,
Alcohol consumption, as measured by recorded data on such as home production. WHO produces estimates
annual sales, stands at 8.9 litres per adult, on average, for unrecorded alcohol consumption.
across OECD countries, based on the most recent data Survey-based estimates of the amount of alcohol
available (Figure 4.3). Austria, Estonia and the Czech Republic, drunk by the 20% heaviest drinkers rely on the data
as well as Lithuania, reported the highest consumption of analysis of the latest available national health surveys
alcohol with 11.5 litres or more per adult per year in 2013. for 13 OECD countries. The list of surveys is provided
Low alcohol consumption was recorded in Turkey and in Table A.1 in Annex A in the publication Tackling
Israel, as well as in Indonesia and India, where religious Harmful Alcohol Use Economics and Public Health Policy
and cultural traditions restrict the use of alcohol in some (OECD, 2015).
population groups.
Although average alcohol consumption has gradually
fallen in many OECD countries since 2000, it has risen in
Poland, Sweden and Norway, as well as in Latvia, Lithuania References
and the Russian Federation. However, national aggregate Bouchery, E.E. et al. (2011), Economic Costs of Excessive
data does not permit to identify individual drinking patterns Alcohol Consumption in the U.S., 2006, American Journal
and the populations at risk. OECD analysis based on individ- of Preventive Medicine, Vol. 41, No. 5, pp. 516-524.
ual-level data show that hazardous drinking and heavy epi-
OECD (2015), Tackling Harmful Alcohol Use Economics and
sodic drinking are on the rise in young people and women
Public Health Policy, OECD Publishing, Paris,
especially. Men of low socioeconomic status are more likely
http://dx.doi.org/10.1787/9789264181069-en.
to drink heavily than those of a higher socioeconomic status,
while the opposite is observed in women (OECD, 2015). OECD (2012), OECD Reviews of Health Systems: Russian
Alcohol consumption is highly concentrated, as the large Federation, OECD Publishing, Paris,
majority of alcohol is drunk by the 20% of the population http://dx.doi.org/10.1787/9789264168091-en.
who drink the most (Figure 4.4), with some variation across Rehm, J. et al. (2009), Global Burden of Disease and Injury
countries. The 20% heaviest drinkers in Hungary consume and Economic Cost Attributable to Alcohol Use and
about 90% of all alcohol consumed, while in France the Alcohol-use Disorder, The Lancet, Vol. 373, pp. 2223-2233.
share is about 50%. WHO (2014), Global Status Report on Alcohol and Health 2014,
In 2010, the World Health Organization endorsed a global Geneva.
strategy to combat the harmful use of alcohol, through WHO (2010), Global Strategy to Reduce the Harmful Use of
direct measures such as medical services for alcohol- Alcohol, Geneva.

4.3. Alcohol consumption among adults, 2000 and 2013 (or nearest years)
2013 2000
Liters per capita (15 years +)
16
14
12
10
L nd
Ir e g a l
C z an F r y
S o C a il e
h da
St a
CD s
Ze ds
r tu a
Po nd
Ru H u n c e
E s e p.
Au ni a
th ia
a
In y
Co Ri l
lo c a
i t e Ko a
N e F in 3 4
h .
C o Isr i a
M bia
Br d
Gr a z il
Ch n
Ch o
a
No t al y
Ic ay
Ja ce
Sw pan
N e er d
nm d
i te Slo ar k
ng ia
B e dom
S m
Sl s t in
S w a k li a
er .
L u er a nd
m ny
Fr r g
Tu sia
s t ae
i t z Rep
ec ed
OE ate
Po a t v i
ni
e
in
d re
ric
e
an
th lan
De alan
ic
d
Ki n
Li s tr
Au pa
iu
ss nga
u
rk
rw
ee
ov r a
ed
ut na
w lan
xe m a
la
la
ne
to
m
d ve
ua
R
ex
bo
a
I
l
el
lg
Af
a
do
i
G
In
Un
Un
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en; WHO for non-OECD countries.
1 2 http://dx.doi.org/10.1787/888933280835
4.4. Share of total alcohol consumed by the 20% of the population who drink the most, 2012 (or nearest year)
(%)
100
91
73
75
68 69
66 66
63 63
61
58 59
54
50
50
25
0
nd
es
ce
ain
nd
da
y
y
d
nd
an
re
ar
an
an
pa
at
an
na
la
la
Sp
Ko
ng
la
rm
al
nl
Ja
St
Ir e
er
Fr
Ca
ng
Ze
Fi
Hu
it z
Ge
d
(E
i te
w
Sw
Ne
UK
Un
Source: OECD (2015), Tackling Harmful Alcohol Use Economics and Public Health Policy.
1 2 http://dx.doi.org/10.1787/888933280835

Nutrition is an important determinant of health. Insuffi- to be higher in families where household heads have a
cient consumption of fruit and vegetables is one factor that higher level of education (Elmadfa, 2009).
can play a role in increased risk of morbidity (Bazzano et The promotion of fruit and vegetable consumption, especially
al., 2003; Riboli and Norat, 2003). Food insecurity, that is the in schools and at the workplace, features in the EU platform
inability to afford enough food for a healthy and active life, for action on diet, physical activity and health (European
is also associated with adverse health effects (Seligman et Commission, 2014).
al., 2010). Proper nutrition assists in preventing a number of
chronic conditions, including cardiovascular disease,
hypertension, type-2 diabetes, stroke, certain cancers,
musculoskeletal disorders and a range of mental health Definition and comparability
conditions.
Estimates of daily fruit and vegetable consumption
In response to a health survey question asking How often are derived from national and European Health Inter-
do you eat fruit?, the percentage of adults consuming fruit view Survey questions. Typically, respondents were
daily varied from about 30% in Finland, to 94% in Australia asked How often do you eat fruit (excluding juice)?
(Figure 4.5). Across the 29 countries providing data, on and How often do you eat vegetables or salad
average 55% of men and 66% of women reported to eat fruit (excluding juice and potatoes)?.
daily. Women reported eating fruit more often than men in
Data for Greece and Switzerland include juices as a
all countries except in Switzerland, with the largest gender
portion of fruit, and juices and soups as a portion of
differences in Germany, Slovenia, and Iceland (20 percentage
vegetable. Data for Australia, Greece, New Zealand,
points or more). In Australia, Greece, Mexico, and the United
and the United Kingdom include potatoes as vegetables.
Kingdom, gender differences were much smaller, under
Data rely on self-reporting, and are subject to errors in
5 percentage points.
recall. The same surveys also ask for information on
Persons aged 65 and over were more likely to eat fruit than age, sex and educational level. Data are not age stan-
those in younger age group; with the lowest consumption dardised, with aggregate country estimates representing
in people aged 15-24 years. Fruit consumption also varies crude rates among respondents aged 15 years and over
by education level, generally being highest among persons in all countries, except Germany and Australia which is
with higher educational levels. 18 years and over.
Daily vegetable consumption ranged from around 33% in
men in Slovenia to nearly 100% in Korea, with Australia and
New Zealand at about the same levels, but counting pota- References
toes as vegetables (Figure 4.6). The average across 29 OECD
countries was 61% for men and 70% for women. Again, Bazzano, L.A., M.K. Serdula and S. Liu (2003), Dietary
more women than men reported eating vegetables daily in Intake of Fruits and Vegetables and Risk of Cardiovascu-
all countries, except in Korea, Australia and Mexico where lar Disease, Current Atherosclerosis Reports, Vol. 5,
vegetable consumption is not significantly different pp. 492-499.
between men and women. In Sweden, Switzerland, Norway, Elmadfa, I. (ed.) (2009), European Nutrition and Health Report
Germany and Slovenia, gender differences exceeded 2009, Basel, Switzerland.
16 percentage points. European Commission (2014), EU Platform on Diet, Physical
Patterns of vegetable consumption across age groups and Activity and Health, 2014 Annual Report, European Com-
by level of education are similar to those observed for fruit. mission, Brussels.
Older persons are more likely to eat vegetables daily. Highly Riboli, E. and T. Norat (2003), Epidemiologic Evidence of the
educated persons eat vegetables more often. Protective Effect of Fruit and Vegetables on Cancer Risk,
The availability of fruit and vegetables is the most impor- American Journal of Clinical Nutrition, Vol. 78 (Suppl.),
tant determinant of consumption. Despite large variations pp. 559S569S.
between countries, vegetable, and especially fruit, availabi- Seligman, H.K., B.A. Laraia and M.B. Kushel (2010), Food
lity is higher in Southern European countries, with cereals Insecurity Is Associated with Chronic Disease among
and potatoes more available in Central and Eastern Euro- Low-income NHANES Participants, Journal of Nutrition,
pean countries. Fruit and vegetable availability also tends Vol. 140, pp. 304-310.

4.5. Daily fruit eating among adults, 2013 (or nearest year)
Men Women Total
100
80
60
40
20
0
da
nd
li a
el
m
ly
a
p.
p.
De ay
OE n
es
it z rk
ain
nd
Sw e
Ge 2 9
Es y
Sl i a
ia
ey
Be d
i te nce
d
il e
d
an
re
c
ar
an
an
an
ic
ra
n
en
It a
Re
Re
iu
do
Sw ma
rk
rw
at
ee
Ch
ra
na
ed
la
la
la
CD
Sp
to
ex
Ko
ng
rm
Un Fr a
al
el
nl
Is
lg
ov
St
Tu
Po
Ir e
er
st
ng
Gr
Ca
No
n
h
ak
M
Ze
Ic
Fi
Hu
Au
ec
d
Ki
ov
w
Cz
d
Sl
Ne
i te
Un

1 2 http://dx.doi.org/10.1787/888933280841
4.6. Daily vegetable eating among adults, 2013 (or nearest year)
Men Women Total
100
80
60
40
20
0
a
es
da
en
nd
nd
il e
li a
el
29
ay
y
ce
nd
p.
ly
ce
p.
d
ain
ia
ke
an
re
ni
ar
ar
an
an
an
ic
ra
en
It a
Re
Re
iu
do
at
rw
ee
an
Ch
ra
ed
na
la
la
la
CD
to
Sp
ex
Ko
nm
ng
r
rm
al
el
nl
Is
lg
ov
St
Tu
Ir e
Po
er
st
ng
Gr
Fr
Sw
Ca
No
Es
h
ak
M
Ic
Ze
Fi
OE
Be
Hu
Au
it z
De
Ge
ec
Sl
d
Ki
ov
i te
w
Cz
Sw
d
Sl
Ne
Un
i te
Un

Information on data for Israel: http://oe.cd/israel-disclaimer 1 2 http://dx.doi.org/10.1787/888933280841

Obesity is a known risk factor for numerous health prob- A growing number of countries have adopted policies to
lems, including hypertension, high cholesterol, diabetes, prevent obesity from spreading further. The policy mix
cardiovascular diseases, respiratory problems (asthma), includes, for instance, public awareness campaigns, health
musculoskeletal diseases (arthritis) and some forms of professionals training, advertising limits or bans on
cancer. The rise in overweight and obesity is a major public unhealthy food, taxations and restrictions on sales of certain
health concern, threatening progress in tackling cardiovas- types of food and beverages, and nutrition labelling. Better
cular diseases (OECD, 2015). informed consumers, making healthy food options available,
Estimates of obesity and overweight are derived either encouraging physical activity and focussing on vulnerable
from health examinations or self-reports, the former being groups are some of the areas in which progress has been
higher and more reliable. Based on the latest available made (European Commission, 2014).
surveys, more than half (53.8%) of the adult population in
OECD countries are overweight or obese. In countries
where height and weight are measured (as opposed to self-
reported), this proportion is even greater, at 57.5%. The
prevalence of overweight and obesity among adults Definition and comparability
exceeds 50% in no less than 22 of 34 OECD countries. In
contrast, overweight and obesity rates are much lower in Overweight and obesity are defined as excessive
Japan and Korea and in some European countries (France weight presenting health risks because of the high
and Switzerland), although even in these countries rates proportion of body fat. The most frequently used
are increasing. measure is based on the body mass index (BMI), which
The prevalence of obesity, which presents even greater is a single number that evaluates an individuals weight
health risks than overweight, varies about six fold across in relation to height (weight/height2, with weight in
OECD countries, from a low of 5% in Japan and Korea, to kilograms and height in metres). Based on the WHO
over 32% in Mexico and the United States (Figure 4.7). classification (WHO, 2000), adults with a BMI from 25
Across OECD countries, 19% of the adult population are to 30 are defined as overweight, and those with a BMI
obese. Obesity rates in men and women are similar in most of 30 or over as obese. This classification may not be
countries. However, in Chile, Mexico and Turkey, as well as suitable for all ethnic groups, who may have equiva-
Colombia, the Russian Federation and South Africa, a lent levels of risk at lower or higher BMI. The thresholds
greater proportion of women are obese, while the reverse is for adults are not suitable to measure overweight and
true in Slovenia. obesity among children.
The prevalence of obesity has increased over the past For most countries, overweight and obesity rates are
decade in all OECD countries (Figure 4.8). In 2013, at least self-reported through estimates of height and weight
one in five adults was obese in twelve OECD countries, from population-based health interview surveys.
compared to one in eight a decade ago. Since 2000, obesity However, around one-third of OECD countries derive
rates have increased by a third or more in 14 countries. The their estimates from health examinations. These dif-
rapid rise occurred regardless of where levels stood a ferences limit data comparability. Estimates from
decade ago. Obesity increased by around 45% in both health examinations are generally higher, and more
Denmark and Australia, even though the current rate in reliable than estimates from health interviews. Note
Denmark is only half that of Australia. that the OECD average is based on both types of esti-
mates (self-reported and measured) and, thus, may be
The rise in obesity has affected all population groups,
underestimated.
regardless of sex, age, race, income or education level, but to
varying degrees. Evidence from Canada, the United Kingdom,
France, Italy, Mexico, Spain, Switzerland and the United
States shows that obesity tends to be more common in
lower educated groups, especially in women (OECD, 2014). References
Rates of overweight and obesity vary by education level and
Devaux, M. and F. Sassi (2013), Social Inequalities in
socioeconomic status, and these disparities are significant
Obesity and Overweight in 11 OECD Countries, European
in women while less clear-cut in men (Devaux and Sassi,
Journal of Public Health, Vol. 23, No. 3, pp. 464-469, June.
2013).
European Commission (2014), EU Platform on Diet, Physical
A number of behavioural and environmental factors have
Activity and Health, 2014 Annual Report, Brussels.
contributed to the long-term rise in overweight and obesity
rates in industrialised countries, including the widespread OECD (2015), Cardiovascular Disease and Diabetes: Policies for
availability of energy dense foods and more time spent being Better Health and Quality of Care, OECD Publishing, Paris,
physically inactive. These factors have created obesogenic http://dx.doi.org/10.1787/9789264233010-en.
environments, putting people, and especially those socially OECD (2014), Obesity Update, OECD Publishing, Paris, June
vulnerable, more at risk of obesity. 2014, www.oecd.org/health/Obesity-Update-2014.pdf.

0
10
20
30
40
0
10
20
30
40
Ja Ja
pa pa
n1 3 Ko n 3.7
4 re
Ko a 4.7
re 3
In In d
a1 do i a
No 5 ne 5.0
rw s
ay 8 Ch i a 5.7
10 No in a
It a rw 7.0
Sw
ly 9 ay
it z 10
Sw 10.0
er it z It al
la N e er y 10.3
8
Ne nd
th th l an
10 er d
er
la la 10.3
nd 9 S w nds
s 11.1
11

Sw ed

11.7
ed Au en
en 9 B e s tr i
Au 12 l a 12.4
st D e giu
ria 9 nm m 13.7
Be 12
lg Fr ar k 14.2
iu 12
m Po anc
De
nm 14 r tu e 14.5
g
ar
k 10 Is a l 15.4
Fr 14 r
an Po ael 15.7
9 la

ce
Po 15 Sl S nd
ov p a 15.8
r tu a k in
ga 13 16.6
l 15
Re
B p. 16.9
Self-reported data
Is
2000
ra 13 Sl r a z
el ov il 17.5
Po 16
O E eni
la 18.3
nd 11 CD a

16 Es 3 4
Sp to 19.0
ain 13 Ru G n i a
19.0
OE 17 ss ree
ia ce
CD
15
n 19.6
1. Data are based on measurements rather than self-reported height and weight.
28 Co F e
Es 18 C z l o m d. 19.6
to ec bi
Cz ni 14 h a 20.9
ec a 19 R
h I c e p.
Re el 21.0
p. 1 14
Ic 21 Lu T and
xe ur k 22.2
Lu el m ey
xe an
d 12 bo 22.3
m 22
2013
bo Ir e u r g 22.7
ur
g1 16 la
Ge
rm 23 L nd 23.0
Ge at v
4.7. Obesity among adults, 2013 (or nearest year)
an 23.6
y1 20 Co rm ia
Un Fi 24 s t any
i te nl Un a 23.6
d an
d 23 i te F Ric a
Measured data
Ki 25 d inl 24.4
ng K i an
do
m 21
ng d
do 24.8
25 m
Ch 24.9
L i Ch
il e 1 25 t h il e
Ca 25 ua 25.1
na
25.7
S o C a ni a
Au da 1 22 ut na
st 26 h da
A 25.8
Ne ra Au fric
w 20
4.8. Increasing obesity among adults in OECD countries, 2000 and 2013 (or nearest years)
li a 1 st a 26.5
Ze 28 r
al N e Hu a li a
an 28.3
d1 25 w nga
M 31
Ze r y
a 28.5
Un ex
ic 24
Un M l a nd
i te
d o1 i t e ex 30.6
St 32 d ico
at St 32.4
es 1 31 at
es
35 35.3
75
1 2 http://dx.doi.org/10.1787/888933280857
1 2 http://dx.doi.org/10.1787/888933280857
Children who are overweight or obese are at greater risk of

poor health in adolescence, as well as in adulthood. Among Definition and comparability
young people, orthopaedic problems and psychosocial
problems such as low self-image, depression and impaired Estimates of overweight and obesity are based on body
quality of life can result from being overweight. Excess mass index (BMI) calculations using either measured or
weight problems in childhood are associated with an self-reported height and weight, the latter possibly
increased risk of being an obese adult, at which point under-estimating obesity and overweight. Overweight
cardiovascular disease, diabetes, certain forms of cancer, and obese children are those whose BMI is above a set
osteoarthritis, a reduced quality of life and premature of age- and sex-specific cut-off points (Cole et al.,
death become health concerns (Lobstein, 2010; Currie et al., 2000).
2012). Measured data are gathered by the World Obesity
Overweight (including obesity) rates based on measured Federation (WOF, former IASO) from different national
(rather than self-reported) height and weight are about 24% studies, except for Germany (data come from the
for boys and 22% for girls, on average, in OECD countries, 2003-06 KIGGS survey) and Korea (based on the 2013
although rates are measured in different age groups in dif- KNHANES survey). The estimates are based on
ferent countries (Figure 4.9). Boys tend to carry excess national surveys of measured height and weight among
weight more often than girls, with the largest gender differ- children at various ages. Caution is therefore needed
ences observed in China, Denmark, Iceland, Korea and in comparing rates across countries. Definitions of over-
Poland. In contrast, Ireland and South Africa show larger weight and obesity among children may sometimes
overweight rates among girls. More than one in three chil- vary among countries, although whenever possible
dren are overweight in Brazil, Chile, Greece, Italy, Mexico, the IOTF BMI cut-off points are used.
New Zealand, United Kingdom (England) and the United Self-reported data are from the Health Behaviour in
States, and about one in three boys in Spain, and one in School-aged Children (HBSC) surveys undertaken
three girls in Portugal. between 2001-02 and 2009-10. Data are drawn from
Child obesity has increased in the past few decades world- school-based samples of 1 500 in each age group (11-,
wide and seems to be stabilising in high-income countries 13-and 15-year-olds) in most countries. Self-reported
(Ng et al., 2014; Lobstein et al., 2015). Self-reported over- height and weight are subject to under-reporting,
weight rates (including obesity) across OECD countries missing data and error, and require cautious inter-
slightly increased between 2001-02 and 2009-10 from 13% pretation.
to 15% in 15-year-olds (Figure 4.10). The largest increases
during this period were in the Czech Republic, Estonia,
Poland and Slovenia, all greater than 5%. Significant reduc- References
tions in the proportion of overweight or obese children at
age 15 were only observed in Denmark and the United Bemelmans, W. et al. (2011), Overview of 71 European
Kingdom between 2001-02 and 2009-10, although non- Community-based initiatives against Childhood Obesity
response rates to questions about self-reported height and Starting between 2005 and 2011: General Characteristics
weight may bias the results downward. and Reported Effects, BMC Public Health, Vol. 14, No. 758.
Childhood is an important period for forming healthy Cole, T.J. et al. (2000), Establishing a Standard Definition for
behaviours, and the increased focus on obesity has stimu- Child Overweight and Obesity Worldwide: International
lated the implementation of many community-based Survey, British Medical Journal, Vol. 320, pp. 1-6.
initiatives in OECD countries in recent years. Studies show Currie, C. et al. (eds.) (2012), Social Determinants of Health and
that locally focussed interventions, targeting children up to Well-being Among Young People. Health Behaviour in School-
12 years of age can be effective in changing behaviours. aged Children (HBSC) Study: International Report from the
Schools provide opportunities to ensure that children 2009/2010 Survey, WHO Regional Office for Europe,
understand the importance of good nutrition and physical Copenhagen.
activity, and can benefit from both. Teachers and health Lobstein, T. (2010), The Size and Risks of the International
professionals are often involved as providers of health and Epidemic of Child Obesity, in F. Sassi (eds.), Obesity and the
nutrition education, and the most frequent community- Economics of Prevention: Fit Not Fat, OECD Publishing, Paris,
based initiatives target professional training, the social or pp. 107-114, http://dx.doi.org/10.1787/9789264084865-en.
physical environment, and actions for parents (Bemelmans
Lobstein, T. et al. (2015), Child and Adolescent Obesity: Part
et al., 2011).
of a Bigger Picture, The Lancet, Vol. 385, pp. 2510-2520.
Ng, M. et al. (2014), Global, Regional, and National Preva-
lence of Overweight and Obesity in Children and Adults
during 19802013: A Systematic Analysis for the Global
Burden of Disease Study 2013, The Lancet, Vol. 384,
No. 9945, pp. 766-781.

4.9. Measured overweight (including obesity) among children, 2013 (or nearest year)
Boys Girls
% of children at various ages
50
40
30
20
10
M gal
N o um
ak s
Au nd
F i en
Ir e e l
L u nm d
S a
il e
il
Sl D 3 3
Un Z e l y
UK d S nd
ng s
ce
Ic nd
B e and
Is a
m rk
O E ur g
S o un i a
h ry
st a
C a li a
Po p ain
Gr d )
Po ia
K y
Ne ch R y
Sl er l a .
E s e p.
Ch a
Ru J n a
ia an
it z nce
Sw ria
d
Sw ra .
th ep
d
ov n d
( E t ate
a
ni
di
G e or e
Au fric
az
De l an
an
ic
ra
H en
s
w It a
Fe
xe a
ut ga
n
rw
ee
Ch
ra
i
ed
na
C z rma
ss ap
la
la
i te ala
st
ne
r tu
to
In
R
ex
i
bo
Br
la
el
nl
lg
ov
C
er
A
n
do
F
e
In
Ne
Source: World Obesity Federation (2015), KIGGS (2003-06) for Germany and KNHANES (2013) for Korea.
1 2 http://dx.doi.org/10.1787/888933280866
4.10. Change in self-reported overweight among 15-year-olds, 2001-02, 2005-06 and 2009-10
2001-02 2005-06 2009-10
% of 15-year-olds
35
30
25
20
15
10
0
l
s
nd
ce
nd
p.
en
da
es
d.
nd
26
ria
y
ain
ly
g
p.
ia
ce
ga
nd
an
ke
ni
ar
ar
an
an
ur
en
It a
Re
Re
iu
Fe
do
at
an
ee
ed
na
la
la
la
CD
st
to
r tu
Sp
nm
ng
bo
r
rm
la
nl
el
lg
ov
St
Tu
Po
Ir e
er
ng
Au
Gr
Fr
n
Sw
Ca
Es
ak
Ic
er
Fi
OE
Be
Hu
Po
ia
it z
De
Ge
ec
Sl
d
Ki
ov
th
ss
xe
i te
Cz
Sw
Ne
d
Sl
Ru
Lu
Un
i te
Un
Source: Currie et al. (2004); Currie et al. (2008); Currie et al. (2012).
1 2 http://dx.doi.org/10.1787/888933280866

5. HEALTH WORKFORCE
Doctors (overall number)
Doctors by age, sex and category
Medical graduates
International migration of doctors
Remuneration of doctors (general practitioners and specialists)
Nurses
Nursing graduates
International migration of nurses
Remuneration of nurses
The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of
the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the
terms of international law.

5. HEALTH WORKFORCE
The number of doctors per capita varies widely across be enough new doctors to replace those who will retire. In
OECD countries. In 2013, Greece had the highest number some countries where domestic training efforts increased
(with 6.3 doctors per 1 000 population), followed by Austria. (e.g., the United Kingdom and the Netherlands), there have
Turkey and Chile had the lowest number among OECD been recent concerns of possible surpluses of certain catego-
countries, with slightly less than two doctors per ries of doctors in the years ahead. This has led to recom-
1 000 population. The OECD average was just over three mendations to reduce slightly student intakes in medical
doctors per 1 000 population. The number of doctors per schools and post-graduate training programmes for certain
capita is much lower in some partner countries. There was specialties (CfWI, 2012; ACMMP, 2014).
less than one doctor per 1 000 population in Indonesia, In many countries, current concerns focusses more specif-
India and South Africa. In China, the number of doctors per ically on shortages of general practitioners (see the indica-
capita is still about half the OECD average, but it has grown tor related to doctors by age, sex and category) or the
significantly since 2000 (Figure 5.1). undersupply of doctors in rural and remote regions (see the
Since 2000, the number of doctors has grown in nearly all indicator on the geographic distribution of doctors in
OECD countries, both in absolute number and on a per cap- Chapter 7).
ita basis. The growth rate was particularly rapid in some
countries which started with lower levels in 2000 (Turkey,
Korea and Mexico), but also in countries which already had
a large number such as Greece and Austria. In Greece, the
number of doctors per capita increased strongly between The data for most countries refer to practising doctors,
2000 and 2008, but has stabilised since then. The number of defined as the number of doctors who are providing care
doctors has also increased strongly in Australia and the directly to patients. In many countries, the numbers
United Kingdom (Figure 5.2), driven mainly by a strong rise include interns and residents (doctors in training).
in the number of graduates from domestic medical educa- The numbers are based on head counts. The data for
tion programmes (see indicator on medical graduates). Ireland are based on estimations. Several countries
On the other hand, the number of physicians per capita also include doctors who are active in the health sector
remained fairly stable between 2000 and 2013 in Estonia, even though they may not provide direct care to
France, Israel and the Slovak Republic. In France, the number patients, adding another 5-10% of doctors. Portugal
of doctors increased by 10%, more or less at the same pace reports the number of physicians entitled to practice,
as the population size. resulting in a larger over-estimation of the number
The number of doctors has continued to grow in most of practicing doctors of about 30%. Belgium and
OECD countries following the 2008-09 recession, although Luxembourg set a minimum threshold of activities
the growth slowed down in some countries such as Greece. for doctors to be considered to be practising, thereby
In the United Kingdom, the growth did not slow down resulting in an under-estimation compared with other
much; there were 15% more employed doctors in 2013 than countries which do not set such minimum thresholds.
in 2008 (Figure 5.2). Data for India are likely over-estimated as they are
based on medical registers which are not updated to
Projecting the future supply and demand of doctors is
account for migration, retirement or death, nor do
challenging given the high levels of uncertainty concerning
they take into account doctors registered in multiple
their retirement and migration patterns and their demand
states.
(Ono, Lafortune and Schoenstein, 2013). In Australia, a
recent projection exercise based on a status quo policy
scenario estimated that there may be an over-supply of
doctors by 2017 before moving to an under-supply from References
2020 to 2030. This projection exercise explored different
ACMMP (2014), The 2013 Recommendations for Medical Special-
scenarios that may either mitigate or exacerbate these
ist Training, Utrecht.
imbalances. If the demand for doctors is growing at a
slightly slower pace than projected because of slower GDP CfWI Centre for Workforce Intelligence (2012), A Strategic
growth, the projected shortage in the next decade may Review of the Future Healthcare Workforce: Informing Medical
disappear and there may be a slight over-supply of doctors and Dental Student Intakes, London.
by 2030. On the other hand, if there is a sharp reduction in Health Workforce Australia (2014), Australias Future Health
the number of immigrant doctors, a growing number of Workforce Doctors, Canberra.
domestic medical graduates would be required to close any Ono, T., G. Lafortune and M. Schoenstein (2013), Health
projected gap (Health Workforce Australia, 2014). Workforce Planning in OECD Countries: A Review of 26
Many countries have anticipated the upcoming retirement Projection Models from 18 Countries, OECD Health Work-
of a significant number of doctors by increasing their training Papers, No. 62, OECD Publishing, Paris,
ing efforts over the past decade to ensure that there would http://dx.doi.org/10.1787/5k44t787zcwb-en.

5. HEALTH WORKFORCE
5.1. Practising doctors per 1 000 population, 2000 and 2013 (or nearest year)
2013 2000
Per 1 000 population
7
6.3
5.0 4.9
5
4.3 4.3 4.3
4.1 4.0 4.0
4 3.9 3.8
3.7 3.6 3.6
3.4
3.4 3.4 3.3 3.3
3.3 3.3 3.2 3.2
3.0 3.0
3 2.8 2.8 2.8
2.7 2.6
2.6 2.6
2.3 2.2
2.2 2.2
2 1.9 1.8
1.8 1.8
1.7
1 0.8 0.7
0.3
0
S w er m al
S w and
en
ia ria
r t ia
Es ds
N o e d.
th y
er y
C z Sp l y
h n
n m p.
Ic ar k
Sl s el
C a
Un Slo l a nd
a
I
a k li a
Ne Fr ep.
la
ng 4
N e mb m
La r y
Fi via
Lu B el nd
i t e Z e ur g
ng d
Ja a
Ir e o m
Po n
M rea
d ni a
Ca ates
Ko d
Ch o

lo z il
Tu bia
So C ey
h na
do i a
a
Ru Au ce
er c e
d
il e
Li r wa
i t z an
O E oni
si
ric
e c ai
K i an
n
pa
ic
Hu D 3
Po uan
A u sr a
In In d
It a
d
De Re
xe giu
an
Co Br a
ov tr a
ed
u t hi
G ug
la
ss st
ne
t
rk
th an
m
i te ve
ex
n
na
F
ee
d
Un w o
R
l
nl
d al
Af
St
t
el
n
Gr
1. Data include not only doctors providing direct care to patients, but also those working in the health sector as managers, educators, researchers, etc.
(adding another 5-10% of doctors).
2. Data refer to all doctors licensed to practice (resulting in a large over-estimation of the number of practising doctors in Portugal, of around 30%).
1 2 http://dx.doi.org/10.1787/888933280876
5.2. Evolution in the number of doctors, selected OECD countries, 2000 to 2013 (or nearest year)
Non-European countries European countries
Australia Canada France Germany
Japan United States Spain United Kingdom
Index (2000 = 100) Index (2000 = 100)
170 170
160 160
150 150
140 140
130 130
120 120
110 110
100 100
2000 2003 2006 2009 2012 2000 2003 2006 2009 2012

1 2 http://dx.doi.org/10.1787/888933280876

5. HEALTH WORKFORCE
Beyond the overall number of doctors, the age and gender maintain a more equal balance between specialists and
composition of the medical workforce and the mix between generalists. In Ireland and Portugal, most generalists are
different categories of doctors also have important implica- not really general practitioners, but rather non-specialist
tions on the supply of medical services. The ageing of doc- doctors working in hospitals or other settings. In some
tors in OECD countries has, for many years, raised concerns countries such as the United States, general internal medi-
that there may not be sufficient new recruits to replace cine doctors are categorised as specialists although their
them, although there is evidence that the retirement of practice is often very similar to that of general practitio-
doctors often only occurs gradually and that their retire- ners, resulting in some underestimation of the capacity to
ment age is increasing (Pong, 2011). The rising share of provide generalist care.
female doctors may affect the overall supply of medical In most OECD countries, specialists earn more than general
services, as women tend to work fewer hours than men, practitioners, providing financial incentives for doctors to
although it appears that working time preferences are specialise (see indicator on the remuneration of doctors). In
becoming more similar among new generations of men response to concerns about shortages of general practitio-
and women doctors. The growing imbalance in favour of ners, many countries have taken steps to improve the num-
greater specialisation over general medicine raises con- ber of training places in general medicine. For example, in
cerns in many countries about access to primary care for all France, about 50% of all post-graduate medical training
the population. places are reserved for general medicine (DREES, 2014). In
In 2013, on average across OECD countries, one-third of all Canada, the number of post-graduate training places in
doctors were over 55 years of age, up from one-fifth in 2000 family medicine more than doubled between 2000 and
(Figure 5.3). While these doctors might be expected to retire 2013, as part of a national effort to strengthen access to pri-
over the next ten years, a growing number of doctors con- mary care for the whole population (CAPER, 2015). How-
tinue to practice after 65 years. In Israel and Italy, almost ever, for these training policies to have lasting effects on
half (49%) of all doctors were over 55 years of age in 2013, the composition of the medical workforce, they need to be
but in Israel this high share may be due partly to the fact complemented by other measures to improve the employ-
that these numbers relate to all doctors licensed to practice ment and remuneration conditions of general practitioners
(and some of them may no longer be practicing). By con- in order to attract and retain a sufficient number of new
trast, only about 15% of doctors in the United Kingdom and doctors.
Korea were aged over 55 due to large numbers of new grad-
uates entering medical practice over the past decade (see
the indicator on medical graduates).
Pension reforms, as well as a possible greater willingness
and capacity of many doctors to work longer, are likely to The definition of doctors is provided under the previ-
have a significant impact on future replacement needs. ous indicator. In some countries, the data are based
Several OECD countries have reformed their pension sys- on all doctors licensed to practice, not only those
tems and increased the retirement age to take into account practising (e.g., Ireland and Portugal). Not all countries
longer life expectancy. While few studies have examined are able to report all their physicians in the two broad
the impact of these pension reforms specifically on doc- categories of specialists and generalists. This may be
tors, it is likely that they will prolong working lives after age due to the fact that specialty-specific data are not
65 in the coming years. available for doctors in training or for those working in
In 2013, 45% of doctors on average across OECD countries private practice.
were women, up from 38% in 2000 and 29% in 1990
(Figure 5.4). At least half of all doctors now are women in
10 countries. By contrast, only about one-in-five doctors in
References
Japan and Korea were women in 2013.
CAPER Canadian Post-M.D. Education Registry (2015),
The balance between generalists and specialists has
Field of Post-M.D. Training by Faculty of Medicine Providing
changed over the past few decades, with the number of
Post-M.D. Training 2013-2014, database available at
specialists increasing much more rapidly, raising concerns
www.caper.ca.
in many countries about shortages of general practitioners.
On average across OECD countries, generalists made up DREES (2014), Les affectations des tudiants en mdecine
only about 30% of all physicians in 2013; there were more lissue des preuves classantes nationales en 2013
than two specialists for every generalist (Figure 5.5). Medi- [The allocations of medical students following national
cal specialists greatly outnumber generalists in Central and ranking exams in 2013], tudes et Rsultats, No. 894.
Eastern European countries and in Greece. Some countries Pong, R.W. (2011), Putting Up the Stethoscope for Good?, CIHI,
such as France, Canada and Australia have been able to available at www.cihi.ca.

5. HEALTH WORKFORCE
5.3. Share of doctors aged 55 years and over, 2000 and 2013 (or nearest year)
% 2013 2000
50 49 49
44 45
42 43
40 41
40 37 38
36
33 33 33 34 34 34
33
30 26 27 27 28
25 26 26 26
25
21
20
15
13
10
0
m
29
nd
m
a
nd
il e
es
ly
el
d
li a
p.
p.
ce
a in
ia
ay
ria
da
en
nd
an
ni
re
ar
ar
an
an
an
pa
ur
ra
en
It a
do
Re
Re
iu
at
rw
an
Ch
ra
ed
na
la
la
CD
st
to
Sp
Ko
nm
ng
bo
rm
la
el
al
nl
Ja
Is
lg
ov
St
ng
Ir e
er
st
Au
Fr
Sw
Ca
No
Es
ak
Ic
Ze
er
Fi
OE
Be
Hu
m
Au
it z
De
Ge
ec
Sl
Ki
d
ov
th
xe
i te
w
Cz
Sw
d
Ne
Sl
Ne
Lu
i te
Un
Un

1 2 http://dx.doi.org/10.1787/888933280883
5.4. Share of female doctors, 2000 and 2013 (or nearest year)
% 2013 2000
80 74
62
60 56 57 57
55 55
53
50 50
46 46 47 48
42 43 45 45 45
40 40 41 41 41 41
40 38 38 39 39
34 35
32
20 22
20
0
ec gal
No 3 3
nd
P y
th ain
Po ds
Es ia
a
es
n
a
g
B e il e
A u um
Tu y
De den
H u e p.
ov nd
F i p.
Sl nd
i t z li a
Ca y
ce
el
Ze d
Fr d
da
Ge c e
OE ny
i te Au y
ng i a
Sw m
a
ni
e
re
ar
l
ar
w an
an
an
pa
ur
ra
en
It a
d s tr
Re
do
at
rk
rw
ee
an
Ch
S w tr a
n
na
a
la
Sl ol a
a
CD
r tu
to
Ne Sp
R
Ko
nm
i
ng
bo
rm
la
e
nl
N e Ir e l
al
Ja
el
Is
lg
ov
St
er
Gr
ak
s
Ic
er
m
d
Ki
xe
i te
Cz
Lu
Un
Un

1 2 http://dx.doi.org/10.1787/888933280883
5.5. Generalists and specialists as a share of all doctors, 2013 (or nearest year)
% Generalists Specialists Medical doctors not further defined

100
80
40
52
51
53
53
56
58
47
61
65
69
70
69
71
73
73
77
60
62
75
52
80
45
85
86
58
88
63
45
61
54
40
58
40
42
61
60
51
49
47
47
45
44
20
42
38
36
35
33
31
31
31
29
29
28
27
27
27
23
22
20
20
19
19
16
16
15
14
12
12
5
0
Ir e l 3
3
ce
m 33
ov a te s
K i and
Au nds
Po hil e
y
Ze a
ey
Au el
Fi o
Po .
Sw d
Be d
G e ium
th ny
Fr a
Ca e
da
Ic n
D e e p.
k
a in
Sl ay
Es y
ia
Un w i t and
O m
Tu g
M ia
p
nd
li
ni
re
c
ar
l
ar
C z lan
e
an
ic
ur
ga
ra
en
It a
r
Re
do
rk
ee
rw
an
ra
ed
Ne rma
na
la
Lu ECD
st
to
Sp
R
ex
N e Ko
nm
ng
la
bo
C
la
S al
nl
Is
lg
r tu
ov
St
i t e z er
st
ng
Gr
No
ak
er
Hu
ec
d
xe
i te
d
Sl
Un
1. Generalists include general practitioners/family doctors and other generalist (non-specialist) medical practitioners.
2. Specialists include paediatricians, obstetricians/gynaecologists, psychiatrists, medical, surgical and other specialists.
3. In Ireland and Portugal, most generalists are not GPs (family doctors), but rather non-specialist doctors working in hospitals or other settings. In
Portugal, there is some double-counting of doctors with more than one specialty.

5. HEALTH WORKFORCE
Medical graduates
The number of new medical graduates in a given year countries, mainly after 2005, so the number of medical
reflects to a large extent government decisions taken a few graduates has only started to go up recently. In addition to
years earlier on the number of students admitted in medi- the growing number of medical graduates from American
cal schools (so-called numerus clausus policies). Since 2000, universities, there has also been a growing number of
most OECD countries have increased the number of stu- American students who have gone to study abroad (notably
dents admitted to medical education in response to con- in Caribbean countries), with the intention of coming back
cerns about current or possible future shortages of doctors to complete their post-graduate training and practice in the
(OECD, forthcoming), but large variations remain across United States. This is expected to create additional pres-
countries. sures to increase the number of residency posts to allow
In 2013, there were on average about 12 new medical grad- both domestic graduates and US foreign-trained graduates
uates per 100 000 population across OECD countries to complete their post-graduate training.
(Figure 5.6). This proportion was highest in Ireland, In Nordic countries, there has been a fairly steady rise in
whereas Israel and Japan had the lowest number of new the number of medical graduates, with the number of grad-
medical graduates relative to their population. In Ireland, uates in Finland and Norway rising by about 50% between
the number of medical graduates increased strongly in 2000 and 2013. Many Norwegian students also go to study
2013 due at least partly to the opening of new Graduate medicine abroad, notably in Germany, Poland and Hungary,
Entry Programmes a few years earlier, allowing students with the intention of coming back to practice in their
with an undergraduate degree in another discipline to country.
obtain a medical degree in four years only. In Israel, the low There has also been a strong rise in the number of medical
number of domestic medical graduates is compensated by graduates in the Czech Republic, Hungary and Poland. This
the high number of foreign-trained doctors. About one- sharp increase can be explained partly by the growing
third of foreign-trained doctors in Israel are in fact people number of international students choosing these countries
who were born in the country but have pursued their study to purse their medical studies. International students
abroad before coming back. The situation is quite different accounted for about 30% of all medical graduates in the
in Japan, where there are very few foreign-trained doctors. Czech Republic in recent years.
Since 2008, the Japanese government decided to increase
This growing internationalisation of medical education,
intakes in medical education in response to current and
combined with the international migration of already
projected shortages of doctors, which should lead to a
trained doctors, makes it more difficult for national gov-
growing number of medical graduates in the coming years.
ernments to set their own domestic numerus clausus poli-
Following the expansion of the numerus clausus in most cies, given that these policies may be affected by policies
countries over the past fifteen years, the number of medi- and actions taken by actors in other countries (OECD, forth-
cal graduates has increased, though at different paces coming).
(Figure 5.7). In Australia, the number of medical graduates
increased by two-and-a-half times between 2000 and 2013.
Most of this growth reflects an increase in the number of
domestic students, but there has also been a growing num-
ber of international students in medical schools in Australia.
In the United Kingdom, the number of medical graduates Medical graduates are defined as the number of stu-
doubled between 2000 and 2013, reflecting an effort to dents who have graduated from medical schools in a
increase the domestic supply and rely less on foreign- given year. The data for Austria and the United Kingdom
trained doctors. Most of the increase in admission in med- exclude foreign graduates, while other countries
ical schools occurred between 2000 and 2004. In 2013, the include them. In Denmark, the data refer to the num-
number of graduates decreased slightly for the first time, ber of new doctors receiving an authorisation to prac-
and so did the number of students admitted in medical tice, which may result in an over-estimation if these
schools following a 2% reduction in medical school intakes include a certain number of foreign-trained doctors.
based on a projected oversupply of doctors in the coming
years (Department of Health, 2012).
In France, the number of medical graduates has increased
steadily since 2006 following a large increase in the numerus References
clausus between 2000 and 2006. However, the number of Department of Health (2012), The Health and Education
graduates should stabilize in the coming years, as admis- National Strategic Exchange Review of Medical and
sion quotas have remained fairly stable over the past few Dental School Intakes in England, UK Government.
years.
OECD (forthcoming), Health Workforce Policies in OECD Coun-
In the United States, the increase in admission intakes to tries: Right Jobs, Right Skills, Right Places (preliminary title),
medical schools occurred a bit later than in several other OECD Publishing, Paris.

5. HEALTH WORKFORCE
Medical graduates
5.6. Medical graduates, 2013 (or nearest year)

25
20.3
20 19.7
15.5 15.3 15.1

15 14.4 14.3
13.9 13.6
13.2
12.8 12.7 12.7
12.2 11.9
11.5 11.4 11.4 11.2
10.9
10.3 10.2 9.9 9.9 9.7
10 9.0 8.9 8.6
8.2
7.5 7.3
6.5
6.0
5.1
5
0
nd
Be 3 3
K i gal
k
li a
Hu nd
th ar y
Gr s
No m
Au e
Un P tr i a
e c p.
F i p.
Ge and
Sl ny
ay
OE ia
Es ly
nd
Sw ia
ak m
en
a in
Po o
i t z nd
il e
Ze ce
d
Un C an a
es
a
ey
n
el
nd
re
i te ad
ic
an
pa
en
ra
It a
C z Re
Re
ar
iu
Sl gdo
rw
rk
ee
at
n
Ch
ra
ed
la
la
Sw la
CD
u
to
Sp
ex
Ko
Ne ng
rm
Ne Fr a
la
el
nl
al
Ja
lg
Is
nm
ov
St
Ir e
i te or t
Tu
er
st
M
Ic
er
n
Au
d
De
ov
w
d
1. In Denmark, the number refers to new doctors receiving an authorisation to practice, which may result in an over-estimation if these include
foreign-trained doctors.
1 2 http://dx.doi.org/10.1787/888933280892
5.7. Evolution in the number of medical graduates, selected OECD countries, 2000 to 2013 (or nearest year)
Non-European countries European countries
Australia Canada France Germany
Japan United States Netherlands United Kingdom
Index (2000 = 100) Index (2000 = 100)
250 250
200 200
150 150
100 100
50 50
2000 2003 2006 2009 2012 2000 2003 2006 2009 2012
Nordic countries Central and Eastern European countries
Finland Norway Czech Republic Hungary
Sweden Poland Slovak Republic
Index (2000 = 100) Index (2000 = 100)
250 250
200 200
150 150
100 100
50 50
2000 2003 2006 2009 2012 2000 2003 2006 2009 2012

1 2 http://dx.doi.org/10.1787/888933280892

5. HEALTH WORKFORCE
The international migration of doctors and other health Nearly 50% of foreign-trained doctors in the United States
workers is not a new phenomenon, but has drawn a lot of come from Asian countries, with doctors coming from
attention in recent years because of concerns that it might India representing by far the largest number, followed by
exacerbate shortages of skilled health workers in certain the Philippines and Pakistan. More than 10% of doctors
countries, particularly in some developing countries that were trained in the Caribbean Islands, but in many cases
are already suffering from critical workforce shortages. The these were American students who went to study abroad
Global Code of Practice on the International Recruitment of and then came back to the United States to complete their
Health Personnel, adopted by the World Health Assembly post-graduate training and practice. Most foreign-trained
in May 2010, was designed to respond to these concerns. It doctors in the United Kingdom also came from Asian coun-
provides an instrument for countries to promote a more tries, with India also leading by a wide margin, although a
ethical recruitment of health personnel, encouraging coun- growing number of foreign-trained doctors in the United
tries to achieve greater self-sufficiency in the training of Kingdom come from other EU countries.
health workers, while recognising the basic human right of Even if smaller countries in Africa, Asia or Central and Eastern
every person to migrate. Europe lose a small number of doctors in absolute terms,
There are significant differences across OECD countries in this may nonetheless have a large impact on their health
the proportion of doctors trained abroad. In 2013, the share systems. There is growing recognition that OECD countries
of foreign-trained doctors ranged from less than 3% in Turkey, should avoid actively recruiting from countries that are suf-
Poland, Estonia, the Netherlands and the Czech Republic to fering from acute shortages of doctors.
more than 40% in Israel and New Zealand (Figure 5.8). The
very high proportion of foreign-trained doctors in Israel
reflects not only the importance of immigration in this
country, but also that a large number of new licenses are
issued to people born in Israel but trained abroad (one-third The data relate to foreign-trained doctors working in
in 2013). Norway, Ireland and Australia also have a high OECD countries measured in terms of total stocks.
share of foreign-trained doctors, although in Norway The OECD health database also includes data on the
roughly half of foreign-trained doctors are people who were annual flows for most of the countries shown here, as
born in the country but went to pursue their medical stud- well as by country of origin. The data sources in most
ies in another country. The share of foreign-trained doctors countries are professional registries or other adminis-
in the United Kingdom, Switzerland, the United States, trative sources.
Sweden and Canada varies between 23% and 30%.
The main comparability limitation relates to differ-
Since 2000, the number and share of foreign-trained doc- ences in the activity status of doctors. Some registries
tors has increased in many OECD countries (Figure 5.9), are regularly updated, making it possible to distin-
contributing to the overall rise in the number and density guish doctors who are still actively working in health
of doctors. In the United States and the United Kingdom, systems, while other sources include all doctors
the share has remained relatively stable over time, but the licensed to practice, regardless of whether they are
absolute number of doctors trained abroad has continued still active or not. The latter will tend to over-estimate
to increase more or less at the same pace as the number of not only the number of foreign-trained doctors, but
domestically-trained doctors (OECD, forthcoming). Sweden also the total number of doctors (including the
has experienced a strong rise in the number and share of domestically-trained), making the impact on the
foreign-trained doctors, with most of these foreign-trained share unclear. The data source in some countries
doctors coming from Germany, Poland and Iraq. The num- includes interns and residents, while these physicians
ber and share of foreign-trained doctors has also increased in training are not included in other countries. Because
in France and Germany, though at a slower pace. In France, foreign-trained doctors are often over-represented in the
the rise is partly due to a fuller recognition of the qualifica- categories of interns and residents, this may result in an
tions of foreign-trained doctors who were already working under-estimation of the share of foreign-trained doctors
in the country, as well as the inflow of doctors from new EU in countries where they are not included (e.g., France,
member states, notably Romania. Hungary, Poland and Switzerland).
In absolute numbers, the United States has by far the high- The data for Germany and Spain is based on national-
est number of foreign-trained doctors, with more than ity (or place of birth in the case of Spain), not on the
200 000 doctors trained abroad in 2013. Following the place of training.
United States is the United Kingdom with more than
48 000 foreign-trained doctors in 2014. The composition of
migration flows by country of origin depends on several
factors, including: i) the importance of migratory ties; References
ii) languag e; and iii) recognition of qualifications. OECD (forthcoming), Health Workforce Policies in OECD Coun-
Figure 5.10 provides an illustration of the distribution of the tries: Right Jobs, Right Skills, Right Places (preliminary title),
countries of training for the two main OECD receiving Chapter on Changing patterns in the international migra-
countries, the United States and the United Kingdom. tion of doctors and nurses, OECD Publishing, Paris.

5. HEALTH WORKFORCE
5.8. Share of foreign-trained doctors in OECD countries, 2013 (or nearest year)
%
60 58.5
50
43.5
40
35.8
34.2
30.5
30 28.7
27.0
25.0 24.3 23.5
19.9
20
17.3
15.2 14.4
10.7
10 9.4 9.2 8.8
7.6
5.6
4.4
3.0 2.7 2.6 2.4 1.8
0.2
0
26
nd
el
nd
es
en
il e
m
d
s
ay
li a
da
Ge c e
y
Hu
ov r i a
p.
nd
ey
ia
m
N e Rep
a in
nd
ni
ar
ar
an
an
ra
en
an
Re
iu
do
rk
at
rw
an
Ch
ra
ed
na
la
la
la
CD
st
to
nm
ng
Sp
la
al
nl
Is
lg
ov
St
rm
Tu
Ir e
Po
er
st
ng
Au
Fr
Sw
Ca
No
Es
ak
h
Ze
er
Fi
OE
Be
Au
it z
De
ec
Sl
d
Ki
th
i te
w
Cz
Sw
d
Sl
Ne
Un
i te
Un
1. In Germany and Spain, the data is based on nationality (or place of birth in Spain), not on the place of training.
1 2 http://dx.doi.org/10.1787/888933280906
5.9. Evolution in the share of foreign-trained doctors, selected OECD countries, 2000 to 2013 (or nearest year)
Canada New Zealand United States Sweden France Germany
% %
50 50
40 40
30 30
20 20
10 10
0 0
2000 2003 2006 2009 2012 2000 2003 2006 2009 2012

1 2 http://dx.doi.org/10.1787/888933280906
5.10. Main countries of training of foreign-trained doctors, United States and United Kingdom
United States, 2013 United Kingdom, 2014
Other, 16%
Other, 16%
Africa, 6% India, 34%

India, 22%
Africa, 8%
Canada, 4%
Asia 48% Philippines, 6%
Mexico, 5%
Pakistan, 5%
China, 3%
Other EU countries, 18%
Caribbean Isl., 13% Other Asia, 12%
Pakistan, 11%
Ireland, 4%
EU countries, 11% Other Asia, 9%

1 2 http://dx.doi.org/10.1787/888933280906

5. HEALTH WORKFORCE
The remuneration level for different categories of doctors tion. This has been the case in countries such as Estonia,
has an impact on the financial attractiveness of different France, Ireland, Italy and Slovenia, where doctors saw their
medical specialties. In many countries, governments influ- remuneration decrease in nominal terms in certain years
ence the level and structure of physician remuneration by after the crisis. However, in more recent years, the remu-
being one of the main employers of physicians or pur- neration of doctors and other health workers have started
chaser of their services, or by regulating their fees. to rise again (OECD, forthcoming).
OECD data on physician remuneration distinguishes
between salaried and self-employed physicians, although
in some countries this distinction is increasingly blurred,
as some salaried physicians are allowed to have a private Definition and comparability
practice and some self-employed doctors may receive part
The remuneration of doctors refers to average gross
of their remuneration through salaries. A distinction is also
annual income, including social security contribu-
made between general practitioners and all other medical
tions and income taxes payable by the employee. It
specialists combined, though there may be wide differ-
should normally exclude practice expenses for self-
ences in the income of different medical specialties.
employed doctors.
As expected, the remuneration of doctors (both generalists
A number of data limitations contribute to an under-
and specialists) is much higher than that of the average
estimation of remuneration levels in some countries:
worker in all OECD countries (Figure 5.11). Self-employed
1) payments for overtime work, bonuses, other sup-
general practitioners in Australia earned about two times
plementary income or social security contributions
the average wage in 2013 (although this is an under-
are excluded in some countries (Austria, Ireland for
estimation as it includes the remuneration of physicians in
salaried specialists and Italy); 2) incomes from private
training), whereas in Austria, Canada, Denmark, the
practices for salaried doctors are not included in some
Netherlands, Luxembourg and the United Kingdom, self-
countries (e.g. Czech Republic, Hungary, Iceland, Ireland
employed GPs earned about three times the average wage
and Slovenia); 3) informal payments, which may be
in the country.
common in certain countries (e.g. Greece and Hungary),
In most countries, GPs earn less than specialists, and in are not included; 4) data relate only to public sector
many cases much less. In Canada and the Netherlands, employees who tend to earn less than those work-
self-employed specialists earned about 4.5 times the aver- ing in the private sector in Chile, Denmark, Greece,
age wage in 2013, in Germany, it was over five times, while Hungary, Iceland, Ireland, Norway, the Slovak Republic
in Belgium and Luxembourg, they earned more than and the United Kingdom; and 5) physicians in training
six times the average wage (although in Belgium their are included in Australia, the Czech Republic and the
remuneration include practice expenses, thereby resulting United Kingdom for specialists.
in an over-estimation). In France, self-employed specialists
The data for some countries include part-time workers,
earned almost four times the average wage, compared with
while in other countries the data refer only to doctors
just over two times for salaried specialists and self-
working full time.
employed GPs. The income gap between GPs and special-
ists is particularly large in Belgium and the Netherlands. In Belgium, the data for self-employed doctors include
practice expenses, resulting in an over-estimation.
In many OECD countries, the income gap between general
practitioners and specialists has continued to widen over The income of doctors is compared to the average
the past decade, reducing the financial attractiveness of wage of full-time employees in all sectors in the
general practice (Figure 5.12). Since 2005, the remuneration country. The source for the average wage of workers
of specialists has risen faster than that of general practitio- in the economy is the OECD Labour Force Statistics
ners in Canada, Finland, France, Hungary, Iceland, Israel, Database.
Luxembourg and Mexico. On the other hand, in Austria,
Belgium and the Netherlands, the gap has narrowed
slightly, as the income of GPs grew faster than that of spe-
cialists. Reference
In many OECD countries, the economic crisis which started OECD (forthcoming), Health Workforce Policies in OECD Coun-
in 2008-09 had a significant impact on the remuneration of tries: Right Jobs, Right Skills, Right Places (preliminary title),
doctors and other health workers. Several European coun- Chapter on Trends in health labour markets following
tries hard hit by the recession either froze or cut down, at the economic crisis and current policy priorities to
least temporarily, the wages or fees of doctors in efforts to address health workforce issues, OECD Publishing,
reduce cost while protecting access to care for the popula- Paris.

5. HEALTH WORKFORCE
5.11. Remuneration of doctors, ratio to average wage, 2013 (or nearest year)
Specialists General practitioners (GPs)
Salaried Self-employed Salaried Self-employed
3.9 Australia 1.8

4.1 Austria 2.7
6.1 Belgium 2 2.3
4.6 Canada 2.9
2.2 n.a.
Czech Rep.
2.5
Denmark 2.7
2.1 1.5
Estonia
2.6 1.8
Finland
2.2
3.9 France 2.4
3.7
5.3 Germany 4.0
2.3
Greece n.a.
2.0 1.5
Hungary
3.7
Ireland 2.5
3.8 2.3
Israel
2.5 n.a.
Italy
4.2 4.6
6.2 Luxembourg 2.8
3.7 Mexico 2.9
2.9 Netherlands 1.7
4.6 2.8
1.7 Norway n.a.
1.6 Poland 2.1
n.a. Slovak Rep. 2.2

2.3 Slovenia 2.3
2.4 Spain 2.0
2.4 1.8
United Kingdom 3 3.2
6 4 2 0 0 2 4 6
Ratio to average wage in each country Ratio to average wage in each country
1. Physicians in training included (resulting in an underestimation).

2. Practice expenses included (resulting in an over-estimation).
3. Specialists in training included (resulting in an underestimation).
1 2 http://dx.doi.org/10.1787/888933280919
5.12. Growth in the remuneration of GPs and specialists, 2005-2013 (or nearest year)
GPs Specialists
Average annual growth rate (%, in nominal terms)
7
6.3
6.2
6
5.5
5 4.8 4.8
4.6
4.1 4.2
4.1
4 3.7
3.4
2.9 3.0 2.9
3 2.8 2.9
2.6
2.4 2.3
2
1.5
1.2
1
0.3
0
g1
s1
m
da
ria
ce
el
o
ar
an
an
ic
ra
nd
iu
ur
an
na
st
ex
ng
nl
el
Is
lg
bo
la
Au
Fr
Ca
M
Ic
Fi
Be
Hu
er
m
th
xe
Ne
Lu
1. The growth rate for the Netherlands and for Luxembourg is for self-employed GPs and specialists.
1 2 http://dx.doi.org/10.1787/888933280919

5. HEALTH WORKFORCE
Nurses
Nurses greatly outnumber physicians in most OECD coun- Ireland and Denmark (with at least 4.5 nurses per doctor). It
tries. Nurses play a critical role in providing health care not was lowest in Greece (with only about half a nurse per doc-
only in traditional settings such as hospitals and long-term tor) and in Turkey and Mexico (with only about one nurse
care institutions but increasingly in primary care (espe- per doctor).
cially in offering care to the chronically ill) and in home In response to shortages of doctors and to ensure proper
care settings. access to care, some countries have developed more
There are concerns in many countries about current and advanced roles for nurses. Evaluations of nurse practitio-
possible future shortages of nurses, given that the demand ners from the United States, Canada, and the United King-
for nurses is expected to rise in a context of population age- dom show that advanced practice nurses can improve
ing and the retirement of the current baby-boom genera- access to services and reduce waiting times, while deliver-
tion of nurses. These concerns have prompted actions in ing the same quality of care as doctors for a range of
many countries to increase the training of new nurses (see patients, including those with minor illnesses and those
the indicator on nursing graduates), combined with efforts requiring routine follow-up. Existing evaluations find a
to increase the retention rate of nurses in the profession. high patient satisfaction rate, while the impact on cost is
The latter has increased in recent years in many countries either cost-reducing or cost-neutral. The implementation
either because of the impact of the economic crisis that of new advanced practice nursing roles may require
have prompted more nurses to stay or come back in the changes to legislation and regulation to remove any barrier
profession, or following deliberate efforts to improve their to extensions in their scope of practice (Delamaire and
working conditions (OECD, forthcoming). Lafortune, 2010).
On average across OECD countries, there were around
nine nurses per 1 000 population in 2013, up from less than
eight nurses in 2000, so the number of nurses has gone up
both in absolute terms and on a per capita basis (Figure 5.13). Definition and comparability
In 2013, the number of nurses per capita was highest in
The number of nurses includes those employed in
Switzerland, Norway, Denmark, Iceland and Finland, with
public and private settings providing services directly
more than 14 nurses per 1 000 population. The number of
to patients (practising) and in some cases also those
nurses per capita in OECD countries was lowest in Turkey
working as managers, educators or researchers.
(with less than 2 nurses per 1 000 population), and Mexico
and Greece (with between 2 and 4 nurses per 1 000 popula- In those countries where there are different levels of
tion). With regards to partner countries, the number of nurses, the data include both professional nurses
nurses per capita was generally low compared with the who have a higher level of education and perform
OECD average. In 2013, Colombia, Indonesia, South Africa, higher level tasks and associate professional nurses
India and Brazil had fewer than 1.5 nurse per 1 000 popula- who have a lower level of education but are nonethe-
tion, although numbers have been growing quite rapidly in less recognised and registered as nurses. Midwives, as
Brazil in recent years. well as nursing aids who are not recognised as nurses,
should normally be excluded. However, about half of
The number of nurses per capita increased in almost all
OECD countries include midwives because they are
OECD countries since 2000. This was the case in countries
considered as specialist nurses.
that already had a high density of nurses in 2000 such as
Switzerland, Norway and Denmark, but also in Korea, Austria reports only nurses working in hospital,
Portugal and France which used to have a relatively low resulting in an under-estimation.
density of nurses but have converged towards the OECD
average (in the case of Korea and Portugal) or have now
moved beyond the OECD average (in the case of France).
The number of nurses per capita declined between 2000 References
and 2013 in Israel, as the size of the population grew more Delamaire, M.-L. and G. Lafortune (2010), Nurses in
rapidly than the number of nurses. It also declined in the Advanced Roles: A Description and Evaluation of Experi-
Slovak Republic, in both absolute numbers and on a per ences in 12 Developed Countries, OECD Health Working
capita basis. Paper, No. 54, OECD Publishing, Paris,
In 2013, there were about three nurses per doctor on aver- http://dx.doi.org/10.1787/5kmbrcfms5g7-en.
age across OECD countries, with about half of the countries OECD (forthcoming), Health Workforce Policies in OECD
reporting between two to four nurses per doctor (Figure 5.14). Countries: Right Jobs, Right Skills, Right Places (prelimi-
The nurse-to-doctor ratio was highest in Finland, Japan, nary title), OECD Publishing, Paris.

5. HEALTH WORKFORCE
Nurses
5.13. Practising nurses per 1 000 population, 2000 and 2013 (or nearest year)
2013 2000
20
18 17.4
16.7
16.3
16 15.5
14.1
14
13.0
12.4
12.1 11.9
12 11.5
11.2 11.1
10.5
10.0
10 9.5 9.5 9.4
9.1
8.3 8.2
8.0 7.9 7.6
8 7.4
6.4 6.2
6.1 6.1
6 5.8 5.6 4.9
5.3 5.2 5.1
4.9
4 3.6
2.6
2.0 1.8
2 1.5 1.3 1.2
1.2 1.0
Re
No and
Ne J tes
do c a
nm y
Ic ar k
F i n d
Ze an
C a um
Fr da
rm d
N e Ir e a n y
L u e r l n d
m s
st g
i te we a
St n
Be land
i te Slo 3 4
CD
C z ngd ia
e c om
.
Ru t h a
ia ia
E s ar y
a
Sl r t y
Ch .
Po il e
Sp a
Ko d
L a in
Is i a
Gr r a e l
ng .
M ce
Ch o
Tu ina
Br y
u t In d l
h ia
lo ia
a
zi
Au ep
Hu F e d
OE nce
ak al
De r wa
Un S r a li
bi
ni
re
p
d de
Ge lan
ic
A u our
ss an
Ki n
Co nes
tv
xe a n d
e
Po It al
a
Li s tr i
In A f r i
na
w ap
ov ug
la
rk
ee
m
d ve
to
R
ex
i
a
th la
a
l
lg
u
a
n
er
a
el
n
h
it z
Sw
So
Un
1. Data include not only nurses providing direct care to patients, but also those working in the health sector as managers, educators, researchers, etc.
2. Data in Chile refer to all nurses who are licensed to practice (less than one-third are professional nurses with a university degree).
3. Austria reports only nurses employed in hospital.
1 2 http://dx.doi.org/10.1787/888933280929
5.14. Ratio of nurses to physicians, 2013 (or nearest year)

5
4.7
4.6
4.5
4.3 4.3 4.2
4.2 4.1
4.0
4 3.9
3.7 3.7
3.6
3.4
3.2 3.2
3.1
3.0 3.0
3 2.8 2.8 2.8
2.4 2.4
2.2
2.0
2 1.9
1.8 1.7 1.7
1.6 1.6 1.5
1.5 1.5
1.4 1.4 1.3
1.2 1.2
1.0
1 0.8
0.6 0.6
0
L u Ic nd
Ir e t e s
Ne o nd
N e do ds
rm m
er ay
st d
B e r a li a
ng l e
Is a l
Ja d
S w nm n
er k
m nd
U n C ur g
d ada
Z e si a
ov y
C z ol a
E s ar y
Au ic a
d C ia
C e
Sw D3 4
Ko n
h d
ng .
th ia
Sl I i a
So ak ia
Af .
Fr m
Ru L t r i a
ia via
r tu y
Sp el
Ch in
M na
Tu co
ey
lo z il
Gr b i a
d.
ce
Hu Rep
h p
Sl an
OE anc
re
i t z ar
Po It al
Au lan
an
De pa
ec an
K i hi
en
L i ton
n
ov nd
g
ra
a
u t Re
G e giu
Fe
do
C o Br a
th r w
rk
ee
In l a n
ed
i
i
la
xe e l a
la
w ne
ss at
m
ua
a
ex
bo
s
i te an
nl
St
n
l
Fi
P
N
i te
Un
Note: For those countries which have not provided data for practising nurses and/or practising physicians, the numbers relate to the same concept
(professionally active or licensed to practice) for both nurses and physicians, for the sake of consistency. The ratio for Portugal is understimated
because the number of doctors includes all licensed to practise.

5. HEALTH WORKFORCE
Nursing graduates
Many OECD countries have taken steps over the past In Germany, there has been a big increase in the number of
decade or so to increase the number of students admitted nurse graduates in recent years, related at least partly to a
in nursing schools in response to concerns about current or greater offer of registered nurse training programmes in
possible future shortages of nurses (OECD, forthcoming). several universities, in addition to the programmes tradi-
Nonetheless, there are wide variations across countries in tionally offered in vocational nursing schools (Cassier-
training efforts of new nurses, which may be explained by Woidasky, 2013).
differences in the current number and age structure of the The increase in the number of nursing graduates has been
nursing workforce (and hence the replacement needs), in much more modest in Japan and Norway. In Japan, the
the capacity of nursing schools to take on more students, number of nursing graduates rose by only 13% between
as well as the future employment prospects of nurses. 2003 and 2013, but this number has gone up further in 2014.
In 2013, there were on average nearly 50 new nurse gradu- In Norway, this slow increase might be explained by a sig-
ates per 100 000 population across OECD countries, up nificant proportion of Norwegian students who choose to
from about 40 in 2003. Korea and Denmark had the highest go abroad to pursue nursing studies, and then come back to
number of new nurse graduates relative to their popula- their home country to work (see the indicator on interna-
tion, with these two countries graduating more than tional migration of nurses).
90 new nurses per 100 000 population in 2013. Mexico,
Luxembourg and the Czech Republic had the lowest number,
with less than 15 nurse graduates per 100 000 population
(Figure 5.15). Definition and comparability
Over the past decade, the number of nursing graduates has
increased in all OECD countries, but at different rates Nursing graduates refer to the number of students
(Figure 5.16). In the United States, following a marked who have obtained a recognised qualification
decrease in student intakes during the 1990s, the number required to become a licensed or registered nurse.
of students admitted to nursing schools started to increase They include graduates from both higher level and
strongly in the early 2000s, in response to concerns about a lower level nursing programmes. They exclude gradu-
potential significant shortage of nurses in the coming ates from Masters or PhD degrees in nursing to avoid
years. Between 2003 and 2013, the number of nursing grad- double-counting nurses acquiring further qualifica-
uates increased by 70% (from 119 000 to over 200 000 per tions.
year since 2010). Given this strong rise in admission and The data for Denmark and the United Kingdom are
graduation numbers, the most recent projections from the based on the number of new nurses receiving an
US Department of Health and Human Services estimate authorisation to practice.
that there may be an over-supply of registered nurses and
licensed practical nurses in the United States by 2025, if
student admissions and nurse retention rates remain at
their current level (Health and Human Services, 2014).
References
In France, the number of graduates from nursing schools
also increased strongly over the past decade, by 50% overall Cassier-Woidasky, A.-K. (2013), Nursing Education in Germany
between 2003 and 2013. The numerus clausus set by the French Challenges and Obstacles in Professionalisation, DHBW,
Ministry of Health to control entry in nursing education Stuttgart.
programmes increased substantially since 1999, with the Health and Human Services (2014), The Future of the
number of places growing by nearly 70% (rising from Nursing Workforce: National- and State-level Projec-
around 18 400 places in 1999 to over 31 000 in 2013). Most of tions, 2012-2025, US Department of Health and Human
the growth occurred in the academic year of 2000/2001 Resources, Rockville, Maryland, United States.
when the annual quota was increased by 43%, driven by a
OECD (forthcoming), Health Workforce Policies in OECD Coun-
projected reduction in the supply of nurses resulting from
tries: Right Jobs, Right Skills, Right Places (preliminary title),
the reduction of working time to 35 hours per week, as well
Chapter on Changes in education and training capaci-
as a more general concern about the anticipated retirement
ties for doctors and nurses: Whats happening with
of a large number of nurses.
numerus clausus policies?, OECD Publishing, Paris.

5. HEALTH WORKFORCE
Nursing graduates
5.15. Nursing graduates, 2013 (or nearest year)

100 97
92
84
78
75
75 72
70
67
63 63
55 55 54 53
50 47 47 47
43 42
40 39
38 36 36
34 33
25 24
25 23 22
20 19
15
11 11
0
it z rk 1
Gr a l
n a
Sl nd
ak s
Ze e
OE m
Es s
Au ni a
No li a
Ic y
Un F in d
d
G e e p.
Au y
Ca ia
da
Be nd
i te J 4
ng n
Po i a
Ir e y
Po nd
ce
ly
Tu l
Sw m
en
th nc e
Hu nd
ey
ec in
bo .
M rg
o
e
xe e p
a
nd
ov te
an
w Chil
D e or e
ar
an
i te lan
d apa
ic
3
g
n
ra
r
It a
C z Sp a
iu
do
u
rk
rw
ee
ra
ed
na
la
la
la
a
st
CD
Sw ma
to
r tu
e
Sl S t a
Lu h R
ex
ng
rm
Ne Fr a
la
el
al
Is
lg
ov
er
st
K
er
m
d
Ki
Ne
Un
1. In Denmark, the number refers to new nurses receiving an authorisation to practice, which may result in an over-estimation if these include
foreign-trained nurses.
1 2 http://dx.doi.org/10.1787/888933280939
5.16. Evolution in the number of nursing graduates, selected OECD countries, 2003 to 2013 (or nearest year)
Denmark Finland France Germany
Norway Switzerland Japan United States
Index (2003 = 100) Index (2003 = 100)
200 200
150 150
100 100
50 50
2003 2006 2009 2012 2003 2006 2009 2012

1 2 http://dx.doi.org/10.1787/888933280939

5. HEALTH WORKFORCE
In nearly all OECD countries, the proportion of foreign- there has also been a strong growth in the number of
trained nurses is much lower than that of foreign-trained nurses trained in Portugal (with the number rising to over
doctors. However, given that the overall number of nurses 1 100 in 2013, up from less than 100 in 2008) and in Spain
is usually much greater than the number of doctors, the abso- (rising to over 1 600 in 2013, up from 1 100 in 2008). In
lute number of foreign-trained nurses tends to be greater Belgium, there has been a strong rise in the number of
than that of foreign-trained doctors (OECD, forthcoming). nurses trained in Romania (exceeding 1 000 in 2014, up
OECD countries vary widely in the number and share of from 150 in 2008), Portugal (with the number reaching 500
foreign-trained nurses working in their health system in 2014, up from 10 only in 2008) and to a lesser extent
(Figure 5.17). While there are almost no foreign-trained Spain (with the number reaching 300 in 2014, up from
nurses working in countries such as Estonia, Turkey, Slovenia about 50 in 2008).
and the Netherlands, these make up nearly 25% of the In 2014, more than 6 500 nurses trained in Portugal and
nursing workforce in New Zealand, and between 10% and more than 9 200 nurses trained in Spain were working in
20% of the nursing workforce in Switzerland, Australia, the other EU countries, with a majority of them working in the
United Kingdom and Israel. The share of foreign-trained United Kingdom.
nurses also exceeds 5% in Norway, Canada, the United States,
Germany and Italy. In absolute numbers, the United States
has by far the highest number of foreign-trained nurses
(with almost 250 000 nurses trained abroad in 2013), followed Definition and comparability
by the United Kingdom (86 000 foreign-trained nurses in
2014) and Germany (70 000 foreign-trained nurses in 2010, The data relate to foreign-trained nurses working in
latest year available). OECD countries measured in terms of total stocks.
The number and share of foreign-trained nurses has The OECD health database also includes data on the
increased over the past ten years in several OECD coun- annual flows for most of the countries shown here, as
tries, including New Zealand, Australia, Canada and Italy well as by country of origin. The data sources in most
(Figure 5.18). In Italy, the increase in the immigration of for- countries are professional registries or other adminis-
eign-trained nurses since 2000 was primarily driven by the trative sources.
arrival of a large number of nurses trained in Romania, who The main comparability limitation relates to differ-
are now accounting for nearly half of all foreign-trained ences in the activity status of nurses. Some registries
nurses (Figure 5.19). The movement of Romanian nurses to are regularly updated, making it possible to distin-
Italy preceded Romanias entry in the European Union in guish nurses who are still actively working in health
2007, but has continued since then. systems, while other sources include all nurses
In the United Kingdom, in 2014, nearly half of all foreign- licensed to practice, regardless of whether they are
trained nurses came from Asian countries, mainly from the still active or not. The latter will tend to over-estimate
Philippines (26%) and India (19%). But a growing number of the number of foreign-trained nurses, although it will
foreign-trained nurses also come from other EU countries, also over-estimate the total number of nurses (includ-
such as Spain, Portugal, Romania and Poland. In 2014, more ing the domestically-trained), so the impact on the
than 5 600 nurses trained in Spain were working in the share is not clear.
U n i t e d K i n g d o m , a n d t h e re we re a l s o m o re t h a n The data for some regions in Spain is based on nation-
4 000 nurses trained in Portugal and Romania, and over ality or country of birth, not the place of training.
2 500 nurses trained in Poland.
In other EU countries such as France and Belgium, the per-
centage of nurses trained abroad remains low compared
with the United Kingdom, but their numbers have
References
increased rapidly. The number of foreign-trained nurses OECD (forthcoming), Health Workforce Policies in OECD Coun-
more than doubled in France between 2000 and 2013. About tries: Right Jobs, Right Skills, Right Places (preliminary title),
half of these foreign-trained nurses received their diploma Chapter on Changing patterns in the international
from Belgium (in many cases, these were French citizens migration of doctors and nurses, OECD Publishing,
who went to study to Belgium before coming back), but Paris.

5. HEALTH WORKFORCE
5.17. Share of foreign-trained nurses in OECD countries, 2013 (or nearest year)
%
25 24.3
20
18.7
16.0
15
12.7
10.3
10
8.8
7.5
6.0 5.9 5.8

5.1
5
3.0 2.7 2.7 2.6
2.1 2.0 1.8
1.3 1.2
0.7 0.4 0.2 0.0
0
23
l
nd
es
il e
d
li a
el
ly
en
s
d
a
m
ay
da
ce
ia
ga
a in
nd
an
ke
ni
ar
ar
an
an
ra
en
It a
iu
do
at
rw
an
Ch
ra
ed
na
la
CD
r tu
to
nm
ng
r
rm
Sp
la
al
nl
Is
lg
ov
St
Tu
er
st
ng
Fr
Sw
Ca
No
Es
Ze
er
Fi
OE
Be
Hu
Po
Au
it z
Ge
De
Sl
d
Ki
th
i te
w
Sw
Ne
d
Ne
Un
i te
Un
1. Data for some regions in Spain relate to foreign nationality or country of birth, not the place of training.
1 2 http://dx.doi.org/10.1787/888933280940
5.18. Evolution in the share of foreign-trained nurses, selected OECD countries, 2000 to 2013 (or nearest year)
Canada Australia New Zealand Belgium France Italy
% %
25 25
20 20
15 15
10 10
5 5
0 0
2000 2003 2006 2009 2012 2000 2003 2006 2009 2012

1 2 http://dx.doi.org/10.1787/888933280940
5.19. Main countries of training of foreign-trained nurses, United Kingdom and Italy
United Kingdom, 2014 Italy, 2013
Other Africa, 8%
Nigeria, 4% Other, 10%
Other, 15%
South Africa, 4%
Spain, 6%
Latin America, 11%
Portugal, 5%
India, 19% EU countries, Romania, 5%
29% Romania, 49%
Albania, 4%
Poland, 3%
Other EU, 10% Other EU, 11%
Philippines, 26% Poland, 11%

1 2 http://dx.doi.org/10.1787/888933280940

5. HEALTH WORKFORCE
The remuneration level of nurses is one of the factors introduced in 2012, phased over a three-year period. In the
affecting their job satisfaction and the attractiveness of the Czech Republic, nurses also benefitted from a pay increase
profession. It also has a direct impact on costs, as wages following protests of hospital workers in 2011 (although
represent one of the main spending items in health their pay raise was lower than that for doctors), accompa-
systems. nied by some improvement in other aspects of their working
The data presented in this section generally focus on the conditions (OECD, forthcoming).
remuneration of nurses working in hospitals, although the
data coverage differs for some countries (see the box on
Definition and comparability).
The data are presented in two ways. First, it is compared Definition and comparability
with the average wage of all workers in each country, pro-
The remuneration of nurses refers to average gross
viding some indication of the relative financial attractive-
annual income, including social security contribu-
ness of nursing compared to other occupations. Second,
tions and income taxes payable by the employee. It
the remuneration level in each country is converted into a
should normally include all extra formal payments,
common currency, the US dollar, and adjusted for purchas-
such as bonuses and payments for night shifts and
ing power parity, to provide an indication of the relative
overtime. In most countries, the data relate specifi-
economic well-being of nurses compared with their coun-
cally to nurses working in hospitals, although in Canada
terparts in other countries.
the data also cover nurses working in other settings.
In most countries, the remuneration of hospital nurses was In some federal states, such as Australia, Canada and
at least slightly above the average wage of all workers in the United States, the level and structure of nurse
2013 (Figure 5.20). In Israel and Luxembourg, the income of remuneration is determined at the sub-national level,
nurses was respectively 60% and 40% greater than the aver- which may contribute to variations across jurisdic-
age wage. In Spain and the United States, it was about 30% tions.
greater than the average wage, while in Greece, Australia
Data refer only to registered (professional) nurses in
and Germany it was 20% higher. In other countries, the sal-
Australia, Canada, Chile, Ireland and the United
ary of hospital nurses is roughly equal to the average wage
States, resulting in an overestimation compared to
in the economy. In the Slovak Republic, Hungary and
other countries where lower-level nurses (associate
France, it is about 10% lower.
professional) are also included. Data for New Zealand
When converted to a common currency (and adjusted for relate to nurses employed by publically funded district
purchasing power parity), the remuneration of nurses was health boards, and includes registered nurses, health
at least four times higher in Luxembourg than in Hungary, assistants, nurse assistants, and enrolled nurses.
the Slovak Republic and Estonia (Figure 5.21). Nurses in the These latter three categories have a different and
United States also had relatively high earnings compared significantly lower salary structure than registered
with their counterparts in other countries, which explains, nurses.
at least partly, the ability of the United States to attract
The data relate to nurses working full time, with the
many nurses from other countries.
exception of Belgium where part-time nurses are also
In many countries, the remuneration of nurses has been included (resulting in an under-estimation). The data
affected by the economic crisis in 2008, but to varying for some countries do not include additional income
degrees (Figure 5.22). Outside Europe, the growth in the such as overtime payments and bonuses (e.g., Italy
remuneration of nurses in countries such as the United and Slovenia). Informal payments, which in some
States, Australia and New Zealand slowed down temporar- countries represent a significant part of total income,
ily following the economic crisis, while the crisis did not are not reported.
appear to have any effect on the growth rate in nurse remu-
The income of nurses is compared to the average
neration level in Mexico. In Europe, following the economic
wage of full-time employees in all sectors in the coun-
crisis, the remuneration of nurses was cut down in some
try. The source for the average wage of workers in the
countries, such as in Hungary and Italy, and has been fro-
economy is the OECD Labour Force Statistics Database.
zen in Italy over the past few years. In Greece, the remuner-
ation of nurses has been reduced on average by 20%
between 2009 and 2013.
Some Central and Eastern European countries have intro-
duced a series of measures in recent years to increase the
References
retention of nurses and other health workers, including pay OECD (forthcoming), Health Workforce Policies in OECD Coun-
raise despite tight budget constraints. In Hungary, a staged tries: Right Jobs, Right Skills, Right Places (preliminary title),
increase of 20% in the salaries of nurses and doctors was OECD Publishing, Paris.

5. HEALTH WORKFORCE
5.20. Remuneration of hospital nurses, ratio to average 5.21. Remuneration of hospital nurses, USD PPP, 2013
wage, 2013 (or nearest year) (or nearest year)
Israel 1.6 Luxembourg 88
Luxembourg 1.4
United States 71
Ireland 62
Spain 1.3
Australia 59
United States 1.3 Denmark 56
Greece 1.2 Belgium 54
Australia 1.2 Canada 54
Germany 1.2
Norway 53
Chile 52
Japan 1.1
New Zealand 51
Canada 1.1 Spain 50
OECD24 1.1 Netherlands 50
Belgium 1.1 Iceland 49
Ireland 1.1
Israel 49
United Kingdom 49
Czech Rep. 1.1
Germany 48
Denmark 1.1
OECD29 45
Italy 1.1 Japan 43
United Kingdom 1.1 Finland 41
Poland 1.0
Italy 40
France 37
Estonia 1.0
Turkey 37
Netherlands 1.0 Greece 36
Norway 1.0 Slovenia 35
Slovenia 1.0 Mexico 31
Finland 1.0
Poland 24
Czech Rep. 24
Slovak Rep. 0.9
Estonia 22
France 0.9
Slovak Rep. 21
Hungary 0.9 Hungary 20
0 0.5 1.0 1.5 2.0 0 40 80

Ratio to average wage in each country USD PPP, thousands
1. Data refer to registered (professional) nurses in the United States, 1. Data refer to registered (professional) nurses in the United States,
Australia, Canada and Ireland (resulting in an over-estimation). Ireland, Australia, Canada and Chile (resulting in an over-estimation).
5.22. Evolution in the remuneration of hospital nurses, selected OECD countries, 2005-13 (or nearest year)
Belgium Czech Rep. France 1 Australia Canada Mexico

Hungary Italy Greece 2 New Zealand United States
Index (2005 = 100) Index (2005 = 100)
150 150
140 140
130 130
120 120
110 110
100 100
90 90
80 80
2005 2007 2009 2011 2013 2005 2007 2009 2011 2013
1. Index for France, 2006 = 100.

2. Index for Greece, 2009 = 100.

6. HEALTH CARE ACTIVITIES
Consultations with doctors
Medical technologies
Hospital beds
Hospital discharges
Average length of stay in hospitals
Cardiac procedures
Hip and knee replacement
Caesarean sections
Ambulatory surgery

Consultations with doctors can take place in doctors The number and type of doctor consultations can vary
offices or clinics, in hospital outpatient departments or, in among different population groups in each country. An
some cases, in patients own homes. In many countries OECD study found that the probability of a visit to the GP
(e.g., Denmark, Italy, Netherlands, Norway, Portugal, Slovak tends to be equally distributed in most countries, but in
Republic, Spain and United Kingdom), patients are required nearly all countries, higher income people are more likely
or given incentives to consult a general practitioner (GP) to see a specialist than those with low income, and also
about any new episode of illness. The GP may then refer more frequently (Devaux and de Looper, 2012).
them to a specialist, if indicated. In other countries,
patients may approach specialists directly.
In 2013, the number of doctor consultations per person Definition and comparability
ranged from over 12 in Korea and Japan, to less than three
in Mexico, Finland and Sweden, as well as in South Africa Consultations with doctors refer to the number of
and Brazil (Figure 6.1). The OECD average was about contacts with physicians, including both generalists
6.5 consultations per person per year, with most countries and specialists. There are variations across countries
reporting between five and eight consultations. Cultural in the coverage of these consultations, notably in out-
factors appear to play a role in variations across countries, patient departments of hospitals. The data come
although certain health system characteristics may also be mainly from administrative sources, although in
important. Some countries where doctors receive fee-for- some countries (Ireland, Israel, Italy, Netherlands,
service tend to have above-average consultation rates (e.g. New Zealand, Spain, Switzerland and United Kingdom)
Japan and Korea), while countries with mostly salaried doc- the data come from health interview surveys. Esti-
tors tend to have below-average rates (e.g. Mexico, Finland mates from administrative sources tend to be higher
and Sweden). However, there are examples of countries than those from surveys because of problems with
such as Switzerland and the United States where doctors recall and non-response rates.
are paid mainly by fee-for-service and where consultation In Hungary, the figures include consultations for diag-
rates are below average, suggesting that other factors are nostic exams such as CT and MRI scans (resulting in
also important. an over-estimation). The figures for the Netherlands
In Sweden and Finland, the low number of doctor consulta- exclude contacts for maternal and child care. The data
tions may also be explained partly by the fact that nurses for Portugal exclude visits to private practitioners,
and other health professionals play an important role in while those for the United Kingdom exclude consulta-
providing primary care to patients in health centres, less- tions with specialists outside hospital outpatient
ening the need for consultations with doctors (Delamaire departments (resulting in an under-estimation). In
and Lafortune, 2010). Germany, the data include only the number of cases of
The average number of doctor consultations per person has physicians treatment according to reimbursement
increased in many OECD countries since 2000. This was regulations under the Social Health Insurance Scheme
particularly the case in Korea, partly explained by the rapid (a case only counts the first contact over a three-
increase in the number of physicians during that period. In month period, even if the patient consults a doctor
some other countries, the number of consultations with more often, leading to an under-estimation). Tele-
doctors per person fell. This was the case in Japan, the phone contacts are included in some countries (e.g.
Czech Republic and the Slovak Republic, although the num- Ireland, Spain and United Kingdom). In Turkey, a
ber remains well above average in these three countries. majority of consultations with doctors occur in outpa-
tient departments in hospitals.
Information on the number of doctor consultations per
person can be used to estimate the annual numbers of con-
sultations per doctor. This indicator should not be taken as
a measure of doctors productivity, since consultations can References
vary in length and effectiveness, and because it excludes
Delamaire, M.-L. and G. Lafortune (2010), Nurses in
the work doctors do on hospital inpatients, administration
Advanced Roles: A Description and Evaluation of Experi-
and research. Keeping these reservations in mind, the esti-
ences in 12 Developed Countries, OECD Health Working
mated number of consultations per doctor is highest in
Paper, No. 54, OECD Publishing, Paris,
Korea and Japan, followed by Turkey and Hungary
http://dx.doi.org/10.1787/5kmbrcfms5g7-en.
(Figure 6.2). On the other hand, the estimated number of
consultations per doctor was lowest in Sweden and Finland, Devaux, M. and M. de Looper (2012), Income-related
where consultations with doctors in both primary care set- Inequalities in Health Service Utilisation in 19 OECD
tings and hospitals tend to be concentrated more for Countries, OECD Health Working Papers, No. 58, OECD
patients with more severe and complex cases. Publishing, Paris,
http://dx.doi.org/10.1787/5k95xd6stnxt-en.

0
2
4
6
8
10
12
14
16
0
1 000
2 000
3 000
4 000
5 000
6 000
7 000
8 000
Ko
Ko re
re Ja a 14.6
Ja a 6732
Hu p a n
pa 12.9
Tu n 5633 C z ng
r e ar
Hu ke y Sl ch R y 11.7
So n 4668 ov e
u t g ar Ru a k p. 11.1
h 3646 s s Re
Sl A f y
ov r i c i a p.
n 11.0
ak a 3407 Ge F ed
Re rm . 10.5
Annual consultations per doctor

P o p. 3244 a
Annual consultations per person
la Tu ny 9.9
C nd 3168 L i r ke
C z ana th y 8.2
ec da ua
3076

h C a ni a 8.1
Co Rep
lo . 3010
n
m Be ada 7.7
lg
B e bi a
lg 2727 iu
m
Sp 7.4
Sl ium 2540
ov Au ai
st n 7.4
Ge eni ra
rm a 2472
Po li a 7.1
O an 2442
la
Lu EC y Au nd 7.1
xe D 3 st
2277

Ru m b 2 ria
s s ou 6.8
ia rg 2168 O
It a
n Lu EC l y 6.8
Au F ed
2145
xe D 3
st . m 3 6.6
ra bo
E s li a 2092 S l ur g
to ov 6.5
n e
L a ia 1949 E s ni a 6.5

tv to
i 1943 n
Sp a Fr i a 6.4
1. In Chile, data for the denominator include all doctors licensed to practice.
ain an
Fr 1930 c 6.4
an N e Is e
ce th r a
1923 er el 6.2
Un L Chil la
i t 1894
nd
ite h e
d ua La s 6.2
K n Un tv
N e in gd i a 1894 i t e Ic i a 6.2
t h om d ela
Ki n
er 1888 ng d 6.0
la
nd De dom
Is s 1885 n 5.0
ra Co mar
Un Ic el 1794 lo k 4.7
i te ela m
d nd 1656 No bia 4.7
St rw
at
es 1644 Un Por ay 4.2
It a i te tu
d ga 4.1
Ne B ly 1622 Sw St l
w ra it z ates
Z e z il 1470 er 4.0
al la
a
1370 3.9
6.1. Number of doctor consultations per person, 2013 (or nearest year)
Au nd N e Ir e n d
6.2. Estimated number of consultations per doctor, 2013 (or nearest year)
w lan
3.8
De s tr i Ze d
nm a 1363 al
ar an
M 1298 d 3.7
S w ex k Ch
it z ico
1294
S w il e 3.3
er ed
la
No nd 979
M en 2.9
rw ex
ic
Ir e a y 975 Br o 2.8
la az
2.7
F i nd
nl 921 S o F in il
a ut lan
h 2.6
S w nd 862 Af d
ed ric
en a 2.5
724
101
1 2 http://dx.doi.org/10.1787/888933280967
1 2 http://dx.doi.org/10.1787/888933280967
A corrigendum has been issued for this page. See http://www.oecd.org/about/publishing/Corrigendum-HealthataGlance2015.pdf

New medical technologies are improving diagnosis and Clinical guidelines have been developed in several OECD
treatment, but they are also increasing health spending. countries to promote a more rational use of MRI and CT
This section presents data on the availability and use of exams. In the United Kingdom, the National Institute for
two diagnostic technologies: computed tomography (CT) Health and Clinical Excellence (NICE) has issued a number
scanners and magnetic resonance imaging (MRI) units. of guidelines on the appropriate use of MRI and CT exams
CT and MRI exams help physicians diagnose a range of (NICE, 2012). In the United States, a Choosing Wisely cam-
conditions. U nlike conventional radiography and paign was launched in 2012, led by professional medical
CT scanning, MRI exams do not expose patients to ionising associations, to develop clear guidelines for doctors and
radiation. patients to reduce the use of unnecessary diagnostic tests
The availability of CT scanners and MRI units has and procedures. The guidelines include, for instance,
increased rapidly in most OECD countries over the past two avoiding imaging studies such as MRI, CT or X-rays for
decades. Japan has, by far, the highest number of MRI and patients with acute low back pain without specific indica-
CT scanners per capita, followed by the United States for MRI tions (Choosing Wisely, 2015). A similar Choosing Wisely
units and by Australia for CT scanners (Figures 6.3 and 6.4). campaign was launched in Canada in 2014, and work has
Greece, Iceland, Italy, Korea and Switzerland also has also started in several other OECD countries to produce
significantly more MRI and CT scanners per capita than the similar clear guidelines and recommendations to promote
OECD average. The number of MRI units and CT scanners a more proper use of diagnostic tests and other procedures.
per population is the lowest in Mexico, Hungary, Israel and It is still too early to tell to what extent these campaigns
the United Kingdom. will succeed in reducing the overuse of MRI and CT exams.
There is no general guideline or benchmark regarding the

ideal number of CT scanners or MRI units per population.
However, if there are too few units, this may lead to access
problems in terms of geographic proximity or waiting Definition and comparability
times. If there are too many, this may result in an overuse
The data in most countries cover MRI units and CT
of these costly diagnostic procedures, with little if any
scanners installed both in hospitals and the ambula-
benefits for patients.
tory sector, but the coverage is more limited in some
Data on the use of these diagnostic scanners are available countries. MRI units and CT scanners outside hospi-
for a smaller group of countries, excluding Japan. Based on tals are not included in Belgium, Germany, Portugal
this more limited country coverage, the number of MRI and Switzerland (for MRI units). For Australia and
exams per capita is highest in Turkey and the United Hungary, the number of MRI units and CT scanners
States, followed by France, Luxembourg and Belgium includes only those eligible for public reimbursement.
(Figure 6.5). In the United States, the (absolute) number of Similarly, MRI and CT exams performed outside
MRI exams more than doubled between 2000 and 2013. In hospitals are not included in Austria, Germany, Ireland,
Turkey, it has grown even faster, by two-and-a-half times Portugal, Switzerland and the United Kingdom. Further-
between 2008 and 2013. In this country, there is growing more, MRI and CT exams for Ireland only cover public
evidence that MRI exams are being systematically hospitals. In Australia, the data only include exams
prescribed for patients with various health problems, for private patients (in or out of hospitals), while in
resulting in overuse of these tests. The number of CT Korea and the Netherlands, they only include publicly
exams per capita is highest in the United States, followed financed exams.
by Luxembourg, France and Greece (Figure 6.6). However, in
Greece, the number of CT exams decreased by over 40%
between 2008 and 2012, while the number of MRI exams
also came down by about 30%.
There are large variations in the use of CT and MRI scan- References
ners not only across countries, but also within countries.
Choosing Wisely (2015), Recommendations from the American
For example, in Belgium, there was almost a two-fold vari-
Society of Anesthesiologists, available at: www.choosing-
ation in MRI and CT exams between provinces with the
wisely.org/clinician-lists/american-society-anesthesiologists-
highest and lowest rates in 2010. In the United Kingdom
imaging-studies-for-acute-low-back-pain/ .
(England), the utilisation of both types of diagnostic exams
is generally much lower, but the variation across regions is NICE National Institute for Health and Care Excellence (2012),
greater, with almost a four-fold difference between the Published Diagnostics Guidance, available at
Primary Care Trusts that had the highest rates and lowest http://guidance.nice.org.uk/DT/Published.
rates of MRI and CT exams in 2010/11. In Canada, there has OECD (2014), Geographic Variations in Health Care: What Do We
been a strong rise in the use of both MRI and CT exams in Know and What Can Be Done to Improve Health System
all parts of the country over the past decade, but there con- Performance?, OECD Publishing, Paris,
tinues to be wide variations across provinces (OECD, 2014). http://dx.doi.org/10.1787/9789264216594-en.

6.3. MRI units, 2013 (or nearest year) 6.4. CT scanners, 2013 (or nearest year)
Japan 46.9 Japan 101.3
United States 35.5 Australia 53.7
Italy 24.6 United States 43.5
Korea 24.5 Iceland 40.5
Greece 24.3 Denmark 37.8
Finland 22.1 Korea 37.7
Iceland 21.8 Switzerland 36.6
Switzerland 19.9 Greece 35.2
Austria 19.2 Italy 33.3
Denmark 15.4 Austria 29.6
Spain 15.3 OECD32 24.4
OECD32 14.1 Belgium 22.2
Australia 13.4 Luxembourg 22.1
Ireland 13.3 Finland 21.7
Luxembourg 12.9 Portugal 20.3
Germany 11.6 Estonia 19.0
Netherlands 11.5 Germany 18.7
Estonia 11.4 Ireland 17.8
New Zealand 11.2 Spain 17.6
Belgium 10.8 Poland 17.2
Turkey 10.5 New Zealand 16.6
France 9.4 Slovak Rep. 15.3
Canada 8.8 Czech Rep. 15.0
Slovenia 8.7 Canada 14.7
Czech Rep. 7.4 France 14.5
Slovak Rep. 6.7 Turkey 14.2
Chile 6.6 Chile 12.3
Portugal 6.5 Slovenia 12.1
Poland 6.4 Netherlands 11.5
United Kingdom 6.1 Israel 8.9
Israel 3.1 United Kingdom 7.9
Hungary 3.0 Hungary 7.9
Mexico 2.1 Mexico 5.3
0 10 20 30 40 50 0 25 50 75 100 125
Per million population Per million population
1. Equipment outside hospital not included. 1. Equipment outside hospital not included.
2. Only equipment eligible for public reimbursement. 2. Only equipment eligible for public reimbursement.
6.5. MRI exams, 2013 (or nearest year) 6.6. CT exams, 2013 (or nearest year)
Turkey 119 United States 240
United States 107 Luxembourg 202
France 91 France 193
Luxembourg 81 Greece 181
Belgium 77 Belgium 179
Iceland 75 Iceland 173
Spain 70 Korea 145
Greece 68 Turkey 145
Switzerland 61 Denmark 142
Denmark 60 Israel 141
Canada 53 Portugal 141
OECD28 52 Austria 134
Estonia 51
Canada 132
Austria 50
Slovak Rep. 123
Netherlands 50
Slovak Rep. OECD27 120
46
Finland 45
Australia 110
Czech Rep. 45
Spain 96
United Kingdom 40 Czech Rep. 96
Slovenia 36 Hungary 92
Hungary 35 Switzerland 90
Israel 31 United Kingdom 76
Portugal 30 Netherlands 71
Australia 28 Chile 71
Korea 26 Germany 62
Poland 23 Ireland 59
Germany 22 Poland 55
Ireland 16 Slovenia 55
Chile 13 Finland 32
0 25 50 75 100 125 0 25 50 75 100 125

Per 1 000 population Per 1 000 population
1. Exams outside hospital not included (in Ireland, exams in private 1. Exams outside hospital not included (in Ireland, exams in private
hospital also not included). hospital also not included).
2. Exams on public patients not included. 2. Exams on public patients not included.
3. Exams privately-funded not included. 3. Exams privately-funded not included.

Hospital beds
The number of hospital beds provides a measure of the In several countries, the reduction in the number of hospital
resources available for delivering services to inpatients in beds has been accompanied by an increase in their occu-
hospitals. This section presents data on the number of hos- pancy rates. The occupancy rate of curative care beds stood
pital beds overall and for different types of care (curative at 77% on average across OECD countries in 2013, slightly
care, psychiatric care, long-term care and other functions). above the 2000 level (Figure 6.9). Israel and Ireland had the
It also includes an indicator of bed occupancy rates focus- highest rate of hospital bed occupancy at approximately
sing on curative care beds. 94%, followed by Norway and Canada at around 90%. In the
Among OECD countries, the number of hospital beds per United Kingdom, Belgium and France, the bed occupancy
capita is highest in Japan and Korea, with 11 beds or more rate remained relatively stable during that period.
per 1 000 population in 2013 (Figure 6.7). In Japan and Korea
hospitals have so-called social admissions, that is, a signifi-
cant part of hospital beds are devoted to long-term care. Definition and comparability
The number of hospital beds is also well above the OECD
average in the Russian Federation, and in Germany and Hospital beds are defined as all beds that are regularly
Austria. On the other hand, some of the large partner coun- maintained and staffed and are immediately available
tries in Asia (India and Indonesia) have very few hospital for use. They include beds in general hospitals, mental
beds compared with the OECD average. This is also the case health hospitals, and other specialty hospitals. Beds
for countries in Latin America (Colombia, Mexico, Chile and in residential long-term care facilities are excluded
Brazil). (OECD, 2015).
The number of hospital beds per capita has decreased over Curative care beds are accommodating patients
the past decade in most OECD countries, falling on average where the principal intent is to do one or more of the
from 5.5 per 1 000 population in 2000 to 4.8 in 2013. This following: manage labour (obstetric), treat non-mental
reduction has been driven partly by progress in medical illness or injury, and perform surgery, diagnostic or
technology which has enabled a move to day surgery and a therapeutic procedures.
reduced need for hospitalisation. In many European coun- Psychiatric care beds are accommodating patients
tries, the financial and economic crisis which started in with mental health problems. They include beds in
2008 also provided a further stimulus to reduce hospital psychiatric departments of general hospitals, and all
capacity as part of policies to reduce public spending on beds in mental health hospitals.
health. Only in Korea and Turkey has the number of hospital
Long-term care beds are accommodating patients
beds per capita grown since 2000.
requiring long-term care due to chronic impairments
More than two-thirds of hospital beds (69%) are allocated and a reduced degree of independence in activities of
for curative care on average across OECD countries daily living. They include beds in long-term care
(Figure 6.8). The rest of the beds are allocated for psychiatric departments of general hospitals, beds for long-term
care (14%), long-term care (13%) and other types of care in specialty hospitals, and beds for palliative care.
care (4%). However, in some countries, the share of beds
The occupancy rate for curative (acute) care beds is
allocated for psychiatric care and long-term care is much
calculated as the number of hospital bed-days related
greater than the average. In Korea, 35% of hospital beds are
to curative care divided by the number of available
allocated for long-term care. In Finland, this share is also
curative care beds (multiplied by 365).
relatively high (27%) as local governments (municipalities)
use beds in health care centres (which are defined as
hospitals) for at least some of the needed long-term care in
institutions. In Belgium and Norway, about 30% of hospital References
beds are devoted to psychiatric care.
OECD (2015), OECD Health Statistics 2015, OECD Publishing,
Paris, http://dx.doi.org/10.1787/health-data-en.

0
3
6
9
12
15
0
20
40
60
80
100
%
40
60
80
100
%
Ja
Is Ja pa
ra pa Ru n
el n s s Kor 13.3
Ko ia ea
Ir e re n 11.0
la Ge Ge F ed
nd rm a rm .
an a 9.1
No Au ny
rw Au y 8.3
ay st Li s tr
th ia
Un Ca Hu r i a u 7.7
na ng Hu a ni
i te ar
d da ng a 7.3
Ki Po y ar
ng C z lan C z Pol y 7.0
do ec d ec an
Sw m h h d
it z Re Re 6.6
er p. F r p.
la 6.5

Fr an
nd an B
Curative care beds

6.3
Au Be ce Sl el g c e
l ov iu
st Sl giu ak m
ria ov m R 6.3
Ge Lu ak R L u L e p.
rm xe ep xe a t v 5.8
an
y m m ia
bo . bo 5.8
Be u
lg E s ur g
iu Es rg to 5.1
m to
n F i ni a
Fi ia nl
a 5.0
Ch nl

Gr n d
il e an 4.9
Gr d OE eec
4.8
ee Sw C e
It a it z D2
ly OE ce er 5
OE Sw CD 4.8
CD Sl l an
it z 3 3 ov d 4.7
Psychiatric care beds

er e
2000
24
2000
la No ni a
Ja Sl nd 4.6

ov Au r wa
pa en
n st y
ra 3.9
No ia li a
Sp rw 3.8
a in Au ay Po It al
st r tu y 3.4
M ra g
ex li a
ic Ch al 3.4
o It a Ic in a
Note: Countries ranked from highest to lowest total number of hospital beds per capita.
Fr Po l y el 3.3
an r tu an
ce g Is d 3.2
2013
Po Ic al De r a
2013
r tu el
an
nm e l 3.1
ga d Un ar
Cz
ec l Is i te Sp k 3.1
h d ain
3.0
Re De r ael St
p. nm a
Long-term care beds
N I te
Gr ar Un ew r el a s 2.9
Lu ee Un S k i te Ze nd
xe ce i te pa d ala 2.8
m d in Ki n
ng d
2.8
bo St
ur at d
g es C a om
Es N e Ir e l na 2.8
to w a
6.8. Hospital beds by function of health care, 2013 (or nearest year)
Un nd Tu da 2.7
ni
a i te Zea r
6.7. Hospital beds per 1 000 population, 2000 and 2013 (or nearest year)
Hu d la
n
S w ke y
2.7
ng Ki ed
ar ng d So B en
y do 2.6
6.9. Occupancy rate of curative (acute) care beds, 2000 and 2013 (or nearest year)
Sl ut r a
ov Ca m h z il
en na Af 2.3
Other hospital beds
Sl
ov ia d ric
ak Tu a Ch a 2.3
Re rk M il e
p. Sw ey C o ex i c 2.2
ed lo o
Tu en 1.6
Un r ke
In m b
do i a
i te y Ch ne 1.5
d i si
St M le 1.0
at ex In a
es ic di
a
o 0.5
105
1 2 http://dx.doi.org/10.1787/888933280981
1 2 http://dx.doi.org/10.1787/888933280981
1 2 http://dx.doi.org/10.1787/888933280981
Hospital beds
Hospital discharges
Hospital discharge rates measure the number of patients changes in medical technologies and clinical practices. The
who leave a hospital after staying at least one night. diffusion of new medical interventions often gradually
Together with the average length of stay, they are impor- extends to older population groups, as interventions
tant indicators of hospital activities. Hospital activities are become safer and more effective for people at older ages.
affected by a number of factors, including the capacity of But the diffusion of new medical technologies may also
hospitals to treat patients, the ability of the primary care involve a reduction in hospitalisation if it involves a shift
sector to prevent avoidable hospital admissions, and the from procedures requiring overnight stays in hospitals to
availability of post-acute care settings to provide rehabilita- same-day procedures. In the group of countries where dis-
tive and long-term care services. charge rates have decreased since 2000, there has been a
In 2013, hospital discharge rates were highest in Austria strong rise in the number of day surgeries (see indicator on
and Germany, followed by Lithuania, the Russian Federation, Ambulatory surgery).
the Czech Republic and Hungary (Figure 6.10). They were Hospital discharge rates vary not only across countries, but
the lowest in Colombia, Mexico, South Africa, Brazil and also within countries. In several OECD countries (e.g., Canada,
Canada. In general, those countries that have more hospi- Finland, Germany, Italy, Portugal, Spain and the United
tal beds tend to have higher discharge rates. For example, the Kingdom), hospital medical admissions (excluding admis-
number of hospital beds per capita in Austria and Germany is sions for surgical interventions) vary by more than two-
more than two-times greater than in Canada and Spain, times across different regions in the country (OECD, 2014).
and discharge rates are also more than two-times larger
(see indicator on Hospital beds).
Across OECD countries, the main conditions leading to hos-
pitalisation in 2013 were circulatory diseases, pregnancy
and childbirth, injuries and other external causes, diseases Definition and comparability
of the digestive system, cancers, and respiratory diseases.
Discharge is defined as the release of a patient who
Austria and Germany have the highest discharge rates for
has stayed at least one night in hospital. It includes
both circulatory diseases and cancers, followed by Hungary
deaths in hospital following inpatient care. Same-day
and Estonia for circulatory diseases (Figure 6.11), and
discharges are usually excluded, with the exceptions
Greece and Hungary for cancers (Figure 6.12). While the
of Chile, the Slovak Republic, Turkey and the United
high rates of hospital discharges for circulatory diseases in
States which include some same-day separations.
Hungary and Estonia are associated with lots of people
having heart and other circulatory diseases (see indicator Healthy babies born in hospitals are excluded from
on Mortality from cardiovascular diseases in Chapter 3), hospital d ischarg e rates in several countries
this is not the case for Germany and Austria. Similarly, can- (e.g. Australia, Austria, Canada, Chile, Estonia, Finland,
cer incidence is not higher in Austria, Germany or Greece Greece, Ireland, Luxembourg, Mexico, Spain). These
than in most other OECD countries (see indicator on Cancer comprise some 3-10% of all discharges. The data for
incidence in Chapter 3). In Austria, the high discharge rate Canada also exclude unhealthy babies born in hospitals.
is associated with a high rate of hospital readmissions for Data for some countries do not cover all hospitals. For
further investigation and treatment of cancer patients instance, data for Denmark, Ireland, Mexico, New
(European Commission, 2008). Zealand and the United Kingdom are restricted to public
Trends in hospital discharge rates vary widely across OECD or publicly-funded hospitals only. Data for Portugal
countries. Since 2000, discharge rates have increased in relate only to public hospitals on the mainland
some countries where discharge rates were low in 2000 and (excluding the Islands of Azores and Madeira). Data
have increased rapidly since then (e.g. Korea and Turkey) as for Canada, Ireland and the Netherlands include only
well as in other countries such as Germany where it was acute care/short-stay hospitals. Data for France and
already above-average. In other countries (e.g. Belgium, Japan refer to acute care hospitalisations.
Czech Republic and Japan), they have remained relatively
stable, while in other countries (including Canada, Finland,
France, Italy and Spain), discharge rates fell between 2000
and 2013.
References
Trends in hospital discharges reflect the interaction of several European Commission (2008), Hospital Data Project Phase 2,
factors. Demand for hospitalisation may grow as popula- Final Report, European Commission, Luxembourg.
tions age, given that older population groups account for a OECD (2014), Geographic Variations in Health Care: What Do We
disproportionately high percentage of hospital discharges. Know and What Can Be Done to Improve Health System Per-
However, population ageing alone may be a less important formance?, OECD Publishing, Paris,
factor in explaining trends in hospitalisation rates than http://dx.doi.org/10.1787/9789264216594-en.

Hospital discharges
6.10. Hospital discharges, 2013 (or nearest year)

300
266
252
235
250
213
204
202
199
196
200 182
175
173
173
172
172
171
170
167
166
166
163
162
161
159
155
146
141
140
139
135
129
150
125
124
119
113
111
99
95
100
83
56
53
48
34
50
0
Cz n F a
S l Gr e r y
Fr nd
rm
Ru i t h n y
Ko n
Ja al
h .
ng .
ov
No ni a
F i way
s t d
L a
nm a
E s ar k
Is
lg
S w Po m
er d
Sw ce
Tu rea
Ne EC el
Ze 3 4
m nd
Ch
i t e Ir e i n a
L u Ic a n d
i te gd
r tu s
er l y
Sp n
Ch in
Ca
So B a
h z il
d om
M ric a
lo o
a
Hu Rep
Ge tr ia
ec ed
B e ni a
g
ak ce
Un K in a nd
S l e p.
es
ey
ia i
De a t v i
Po nd
bi
e
i t z lan
pa
ss uan
ra
g
th It a
li
d
iu
ur
Co x ic
Au lan
il e
a
ut r a
an
L a
ed
la
a
xe e l a
w D
ov e
ra
m
e
rk
at
na
to
la
al
Af
s
bo
d l
e
St
n
Au
Ne
Un
1. Excludes discharges of healthy babies born in hospital (between 3-10% of all discharges).
2. Includes same-day discharges.
1 2 http://dx.doi.org/10.1787/888933280996
6.11. Hospital discharges for circulatory diseases, 2013 6.12. Hospital discharges for cancers, 2013
(or nearest year) (or nearest year)
Germany 36.9 Austria 29.3
Austria 35.8 Germany 24.5
Hungary 35.6 Greece 24.0
Estonia 30.1 Hungary 23.3
Slovak Rep. 30.0 Japan 22.2
Czech Rep. 27.7 Estonia 19.8
Greece 27.1 Korea 17.3
Poland 26.6 Slovenia 17.1
Finland 26.3 Slovak Rep. 16.8
Norway 23.8 Finland 16.0
Sweden 23.6 Norway 15.9
Slovenia 20.1 Czech Rep. 15.1
Denmark 20.1 Switzerland 13.6
Belgium 19.8 Denmark 13.5
Italy 19.6 OECD34 13.3
OECD34 19.3 Luxembourg 13.2
France 19.2 Sweden 13.0
Switzerland 18.2 Poland 12.8
United States 18.2 Iceland 11.8
Luxembourg 17.8 Italy 11.6
Australia 16.5 Australia 11.5
Netherlands 16.4 France 11.4
Iceland 14.4 Belgium 11.1
Japan 13.8 Netherlands 11.0
New Zealand 13.7 Portugal 10.8
Portugal 13.3 Spain 9.7
Spain 12.9 United Kingdom 8.2
Turkey 12.7 New Zealand 7.8
United Kingdom 12.2 Ireland 7.6
Israel 11.8 Chile 6.7
Ireland 11.3 Turkey 6.2
Korea 11.3 Israel 6.2
Canada 10.3 Canada 5.8
Chile 7.2 United States 5.1
Mexico 2.4 Mexico 3.0
0 10 20 30 40 0 10 20 30 40

The average length of stay in hospitals (ALOS) is often cost of each episode of care. In Switzerland, the cantons
regarded as an indicator of efficiency. All other things being which moved from per diem payments to diagnosis-related
equal, a shorter stay will reduce the cost per discharge and groups (DRG) based payments, have experienced a reduc-
shift care from inpatient to less expensive post-acute set- tion in their hospital lengths of stay (OECD and WHO, 2011).
tings. However, shorter stays tend to be more service inten- Most countries are seeking to reduce average length of stay
sive and more costly per day. Too short a length of stay whilst maintaining or improving the quality of care. A
could also cause adverse effects on health outcomes, or diverse set of policy options at clinical, service and system
reduce the comfort and recovery of the patient. If this leads level are available to achieve these twin aims. Strategic
to a greater readmission rate, costs per episode of illness reductions in hospital bed numbers alongside development
may fall only slightly, or even rise. of community care services can be expected to shorten
In 2013, the average length of stay in hospitals for all causes average length of stay, such as seen in Denmarks quality-
across OECD countries was about eight days (Figure 6.13). driven reforms of the hospital sector (OECD, 2013). Other
Turkey and Mexico had the shortest stays, with about four options include promoting the uptake of less invasive sur-
days (half the OECD average), whereas Japan and Korea had gical procedures, changes in hospital payment methods,
the longest stays, with over 16 days (more than double the the expansion of early discharge programmes which enable
OECD average). Across OECD countries, the average length patients to return to their home to receive follow-up care,
of stay has fallen from an average of almost 10 days in 2000 and support for hospitals to improve the co-ordination of
to 8 days in 2013. But there are a few exceptions to this gen- care across diagnostic and treatment pathways.
eral pattern, with the average length of stay increasing in
Korea, but also in Hungary and Luxembourg, where it is
now above the OECD average.
Focusing on average length of stay for specific diseases or
conditions can remove some of the effect of different case Average length of stay refers to the average number of
mix and severity. Figure 6.14 shows that average length of days that patients spend in hospital. It is generally
stay following a normal delivery was slightly less than measured by dividing the total number of days stayed
three days on average in 2013, down from more than three- by all inpatients during a year by the number of
and-a-half days in 2000. This ranged from less than admissions or discharges. Day cases are excluded. The
two days in Mexico, Turkey, the United Kingdom, Iceland, data cover all inpatient cases (including not only cura-
Canada, New Zealand and the Netherlands, to five days or tive/acute care cases) for most countries, with the
more in the Slovak Republic and Hungary. exceptions of Canada, Japan and the Netherlands
The average length of stay following acute myocardial where the data still refer to curative/acute care only
infarction was around seven days on average in 2013. It was (resulting in an under-estimation).
shortest in some of the Nordic countries (Denmark, Norway Discharges and average length of stay of healthy
and Sweden), Turkey and the Slovak Republic, at fewer than babies born in hospitals are excluded in several coun-
five days, and highest in Korea and Germany, at more than tries (e.g. Australia, Austria, Canada, Chile, Estonia,
ten days (Figure 6.15). Finland, Greece, Ireland, Luxembourg, Mexico, Spain),
Several factors can explain these cross-country variations. resulting in a slight over-estimation (e.g., the inclu-
Differences in the clinical need of the patient may obvi- sion of healthy newborns would reduce the ALOS by
ously play a role, but these variations also likely reflect dif- 0.5 day in Canada).
ferences in clinical practices and payments systems. The
combination of an abundant supply of beds with the struc-
ture of hospital payments may provide hospitals with
References
incentives to keep patients longer. A growing number of
countries (France, Germany, Poland) have moved to pro- OECD (2013), OECD Reviews of Health Care Quality: Denmark
spective payment methods often based on diagnosis- 2013: Raising Standards, OECD Publishing, Paris,
related groups (DRGs) to set payments based on the esti- http://dx.doi.org/10.1787/9789264191136-en.
mated cost of hospital care for different patient groups in OECD and WHO (2011), OECD Reviews of Health Systems:
advance of service provision. These payment methods Switzerland 2011, OECD Publishing, Paris,
have the advantage of encouraging providers to reduce the http://dx.doi.org/10.1787/9789264120914-en.

6.13. Average length of stay in hospital, 2000 and 2013 (or nearest year)
2000 2013
Days
25
20
17.2
16.5
15
10.8
10.1
9.5
9.4
9.1
8.9
8.9
8.7
10 8.1
8.1
7.9
7.9
7.9
7.6
7.6
7.5
7.5
7.1
7.0
6.6
6.5
6.5
6.4
6.1
6.1
6.0
5.9
5.8
5.7
5.6
4.3
4.0
3.9
5
0
i t z gal
OE nd
Ir e s
m
Un No
St y
Ic d
en
Au hil e
Sw d
De a li a
M rk
Tu o
ey
n
P o ur g
Au 4
Be ria
Ze y
C a nd
a in
Un lov oni a
K i p.
Gr d
ce
Ne lov l
er a
d
Hu c e
ec r y
G e e p.
m y
s
da
i te r wa
e
xe a n
th eni
re
w It al
n
n
an
an
ic
3
ra
nd
d Re
iu
pa
do
a
C z nga
rk
at
an
ee
ed
la
la
a
la
st
CD
Sw r tu
Sp
R
ex
na
Ko
nm
bo
r
Lu rm
C
el
al
nl
Is
lg
S st
Po
er
la
st
Ja
ng
Fr
i te ak
h
Fi
d
S
Ne
1. Data refer to average length of stay for curative (acute) care (resulting in an under-estimation).
1 2 http://dx.doi.org/10.1787/888933281004
6.14. Average length of stay for normal delivery, 2013 6.15. Average length of stay for acute myocardial
(or nearest year) infarction (AMI), 2013 (or nearest year)
Slovak Rep. 5.1 Korea 13.1
Hungary 5.0 Germany 10.3
Czech Rep. 4.3 Estonia 9.1
France 4.1 Chile 8.9
Belgium 4.0 Finland 8.3
Luxembourg 4.0 Austria 8.1
Poland 3.9 Hungary 7.9
Austria 3.8 Portugal 7.9
Greece 3.6 Italy 7.8
Slovenia 3.6 New Zealand 7.8
Switzerland 3.6
Slovenia 7.4
Spain 7.3
Italy 3.4
Switzerland 7.3
Norway 3.1
Ireland 7.1
Finland 3.0
United Kingdom 7.1
OECD32 2.9
Belgium 6.9
Germany 2.9
Luxembourg 6.9
Israel 2.9
OECD33 6.8
Chile 2.8
Mexico 6.6
Australia 2.7
Poland 6.2
Denmark 2.7
Czech Rep. 6.1
Portugal 2.7 Greece 6.1
Korea 2.5 France 6.0
Spain 2.4 Israel 6.0
Sweden 2.3 Netherlands 5.6
Ireland 2.0 Canada 5.5
United States 2.0 Iceland 5.5
Netherlands 1.9 Australia 5.4
New Zealand 1.7 United States 5.4
Canada 1.6 Slovak Rep. 4.9
Iceland 1.6 Turkey 4.9
Turkey 1.5 Sweden 4.7
United Kingdom 1.5 Norway 4.0
1.3 Denmark
Mexico 3.9
0 2 4 6 0 5 10 15
Days Days

Cardiac procedures
Heart diseases are a leading cause of hospitalisation and 2006 than afterwards. In the United States, the share of
death in OECD countries (see indicator on Mortality from angioplasty increased quickly between 2000 and 2006, but
cardiovascular diseases in Chapter 3). Coronary artery has fallen slightly since then. Part of the explanation for
bypass graft and angioplasty have revolutionised the treat- this slight reduction may be due to the fact that the data
ment of ischemic heart diseases in the past few decades. A reported by the United States do not cover the growing
coronary bypass is an open-chest surgery involving the number of angioplasties carried out as day cases (without
grafting of veins and/or arteries to bypass one or multiple any overnight stay in hospital). In addition, the greater use
obstructed arteries. A coronary angioplasty is a much less of drug-eluting stents in the United States as well as in
invasive procedure involving the threading of a catheter other countries reduces the likelihood that the same
with a balloon attached to the tip through the arterial sys- patient will need another angioplasty (Epstein et al., 2011).
tem to distend the coronary artery at the point of obstruc-
tion; the placement of a stent to keep the artery open
accompanies the majority of angioplasties.
In 2013, Germany, Hungary and Austria had the highest
rates of coronary revascularisation procedures, while the The data for most countries cover both inpatient and
rates were lowest in Mexico and Chile (Figure 6.16). day cases, with the exception of Chile, Denmark,
A number of reasons can explain cross-country variations Iceland, Norway, Portugal, Switzerland and the United
in the rate of coronary bypass and angioplasty, including: States, where they only include inpatient cases
1) differences in the capacity to deliver and pay for these (resulting in some under-estimation in the number of
procedures; 2) differences in clinical treatment guidelines coronary angioplasties; this limitation in data cover-
and practices; and 3) differences in coding and reporting age does not affect the number of coronary bypasses
practices. However, the large variations in the number of since nearly all patients are staying at least one night
revascularisation procedures across countries do not seem in hospital after such an operation). Some of the vari-
to be closely related to the incidence of ischemic heart ations across countries may also be due to the use of
disease (IHD), as measured by IHD mortality (see Figure 3.6 different classification systems and different codes
in Chapter 3). For example, IHD mortality in Germany is for reporting these two procedures.
slightly below the OECD average, but Germany has the In Ireland, Mexico, New Zealand and the United
highest rate of revascularisation procedures. Kingdom, the data only include activities in publicly-
National averages can hide important variations in utilisa- funded hospitals, resulting in an under-estimation (it
tion rates within countries. For example, in Germany, the is estimated that approximately 15% of all hospital
rate of coronary bypass surgery and angioplasty is nearly activity in Ireland is undertaken in private hospitals).
three times higher in certain regions compared with oth- Data for Portugal relate only to public hospitals on the
ers. There are also wide variations in the use of these revas- mainland. Data for Spain only partially include activi-
cularisation procedures across regions in other countries ties in private hospitals.
such as Finland, France and Italy (OECD, 2014).
The use of angioplasty has increased rapidly over the past
20 years in most OECD countries, overtaking coronary References
bypass surgery as the preferred method of revascularisa-
Epstein, A. et al. (2011), Coronary Revascularization Trends
tion around the mid-1990s about the same time that the
in the United States, 2001-2008, Journal of the American
first published trials of the efficacy of coronary stenting
Medical Association, Vol. 305, No. 17, pp. 1769-1775, May 4.
began to appear. On average across OECD countries, angio-
plasty now accounts for 78% of all revascularisation proce- OECD (2014), Geographic Variations in Health Care: What Do We
dures (Figure 6.17), and is equal or exceeds 88% in Korea, Know and What Can Be Done to Improve Health System Per-
Estonia, France and Spain. In many OECD countries, the formance?, OECD Publishing, Paris,
growth in angioplasty was more rapid between 2000 and http://dx.doi.org/10.1787/9789264216594-en.

Cardiac procedures
6.16. Coronary revascularisation procedures, 2013 (or nearest year)

Coronary angioplasty Coronary bypass
500
435
400
325
304
296
300 288
273 267
256 256 252 245 244 243 240 240 236
225 220 219 214 210 207
196
200 181
168
159 153
144
132
124
100
62
6
0
B e ds
m
De nd
K i gal
a
No el
es
Ca 1
F a
Hu ny
A y
th r ia
Ic y
d
i t z p.
k
ce
ly
nd
Au en
O E li a
m d
ia
w ke y
Un Po n d
a in
a
il e
o
m
a
d
ni
ar
re
ar
C z lan
i te an
n
ur
ic
ra
en
It a
S w Re
iu
do
rw
at
an
Ch
ra
n
a
ed
na
la
L u inl a
la
a
CD
Ne us t
to
i te r tu
Sp
ex
Ko
nm
ng
bo
Ne Tur
rm
la
Un Pol
al
Is
lg
ov
St
e
Ir e
er
st
ng
Fr
Sw
Es
M
er
Ze
Ge
ec
Sl
d
xe
d
Note: Some of the variations across countries are due to different classification systems and recording practices.
1 2 http://dx.doi.org/10.1787/888933281011
6.17. Coronary angioplasty as a share of total revascularisation procedures, 2000 to 2013 (or nearest years)
2000 2006 2013
%
100
95
90 88 88 88
86 86 85 85 84 84
83 82 82 81 80 80
80 79 79 78 78 78
76 76 75 75 74
72 71
70 66
64 63
60
50
41
40
30
20
10
0
a
a
ce
a in
S w el
ly
No d
K i ds
Fi d
C a al
Au n
ria
Ir e y
Sl ay
Un the ni a
Ic 1
ec m
i t z p.
d
nd
OE rg
um
Au nd
Po a
Po nd
De ar y
es
il e
M y
Hu d a
o
an
re
ni
li
e
i te ar
n
e
an
3
ic
ra
g
It a
S w Re
o
u
rw
rk
an
at
Ch
ra
ed
na
la
la
L u nl a
la
CD
st
to
Sp
Ne ve
r tu
ex
Ko
Un nm
N e el gi
d
ng
bo
rm
i te r la
el
al
Is
St
Tu
er
st
ng
Fr
Es
h
o
Ze
m
Ge
d
xe
w
Cz
d
Note: Revascularisation procedures include coronary bypass and angioplasty.

1 2 http://dx.doi.org/10.1787/888933281011

Significant advances in surgical treatment have provided and 2013. In Germany, these surgical activity rates appear
effective options to reduce the pain and disability associ- to have stabilised in recent years and even come down
ated with certain musculoskeletal conditions. Joint slightly in 2013.
replacement surgery (hip and knee replacement) is consid-
ered the most effective intervention for severe osteoarthri-
tis, reducing pain and disability and restoring some
patients to near normal function.
Osteoarthritis is one of the ten most disabling diseases in Hip replacement is a surgical procedure in which the
developed countries. Worldwide estimates show that 10% hip joint is replaced by a prosthetic implant. It is gen-
of men and 18% of women aged over 60 years have symp- erally conducted to relieve arthritis pain or treat
tomatic osteoarthritis, including moderate and severe severe physical joint damage following hip fracture.
forms (WHO, 2014). Age is the strongest predictor of the Knee replacement is a surgical procedure to replace
development and progression of osteoarthritis. It is more the weight-bearing surfaces of the knee joint in order
common in women, increasing after the age of 50 espe- to relieve the pain and disability of osteoarthritis. It
cially in the hand and knee. Other risk factors include obe- may also be performed for other knee diseases such
sity, physical inactivity, smoking, excess alcohol and as rheumatoid arthritis.
injuries. While joint replacement surgery is mainly carried
Classification systems and registration practices vary
out among people aged 60 and over, it can also be per-
across countries, which may affect the comparability
formed among people at younger ages.
of the data. Some countries only include total hip
In 2013, Switzerland, Germany and Austria had the highest replacement (e.g. Estonia), while most countries also
rates of hip replacement, while the United States had the include partial replacement. In Ireland, Mexico, New
highest rate of knee replacement, followed by Austria, Zealand and the United Kingdom, the data only
Finland and Germany (Figures 6.18 and 6.19). Differences in include activities in publicly-funded hospitals (it is
population structure may explain part of these variations estimated that approximately 15% of all hospital
across countries, and age standardisation reduces to some activity is undertaken in private hospitals). Data for
extent the cross-country variations. Still, large differences Portugal relate only to public hospitals on the main-
persist and the country ranking does not change signifi- land. Data for Spain only partially include activities in
cantly after age standardisation (McPherson et al., 2013; private hospitals.
OECD, 2014).
National averages can mask important variations in hip
and knee replacement rates within countries. In Australia,
References
Canada, Germany, France and Italy, the rate of knee
replacement is more than two times higher in certain McPherson, K., G. Gon and M. Scott (2013), International
regions compared with others, even after age-standardisa- Variations in a Selected Number of Surgical Procedures,
tion (OECD, 2014). OECD Health Working Papers, No. 61, OECD Publishing,
The number of hip and knee replacements has increased rap- Paris, http://dx.doi.org/10.1787/5k49h4p5g9mw-en.
idly since 2000 in most OECD countries (Figures 6.20 OECD (2014), Geographic Variations in Health Care: What Do We
and 6.21). On average, the rate of hip replacement Know and What Can Be Done to Improve Health System Per-
increased by about 35% between 2000 and 2013 and the rate formance?, OECD Publishing, Paris,
of knee replacement nearly doubled. In France, the growth http://dx.doi.org/10.1787/9789264216594-en.
rate for both interventions was slightly lower, but still the WHO (2014), Chronic Rheumatic Conditions, Fact Sheet,
hip replacement rate increased by about 15% while the Geneva, available at: www.who.int/chp/topics/rheumatic/en/.
knee replacement rate rose by nearly 90% between 2000

6.18. Hip replacement surgery, 2013 (or nearest year) 6.19. Knee replacement surgery, 2013 (or nearest year)
Switzerland 292
United States 226
Germany 283 Austria 215
Austria 276 Finland 202
Belgium 246 Germany 190
Norway 243 Belgium 187
Finland 242 Australia 180
Sweden 238 Switzerland 176
France 236 Luxembourg 171
Denmark 227 Denmark 167
Netherlands 216
Canada 166
Luxembourg 216
United States 204
France 145
Iceland 185 United Kingdom 141
United Kingdom 183 Sweden 135
Australia 171 Iceland 132
Czech Rep. 170 OECD30 121
Italy 166 Netherlands 118
OECD33 161 Czech Rep. 115
Slovenia 161 Korea 115
Greece 152 Spain 112
New Zealand 150 Slovenia 108
Hungary 137
Italy 104
Canada 136
Ireland 127
New Zealand 94
Estonia 112 Norway 89
Spain 107 Turkey 67
Slovak Rep. 105 Portugal 62
Portugal 88 Hungary 59
Poland 85 Israel 53
Israel 63 Ireland 50
Turkey 44 Poland 26
Chile 33 Chile 11
Korea 20
Mexico 3
Mexico 8
0 100 200 300
0 100 200 300 400
1 2 http://dx.doi.org/10.1787/888933281026
1 2 http://dx.doi.org/10.1787/888933281026
6.20. Trend in hip replacement surgery, selected OECD 6.21. Trend in knee replacement surgery, selected OECD
countries, 2000 to 2013 (or nearest years) countries, 2000 to 2013 (or nearest years)
France Germany Italy France Germany Italy

United States OECD33 United States OECD30
350 250
300 200
250 150
200 100
150 50
100 0
2000 2002 2004 2006 2008 2010 2012 2000 2002 2004 2006 2008 2010 2012

Caesarean sections
Rates of caesarean delivery have increased in nearly all A number of countries have taken different measures to
OECD countries, although in a few countries this trend has reduce unnecessary caesarean sections. Public reporting,
reversed at least slightly in the past few years. Reasons for provider feedback, the development of clearer clinical
the increase include the rise in first births among older guidelines, and adjustments to financial incentives have
women and in multiple births resulting from assisted been used to try to reduce the inappropriate use of caesar-
reproduction, malpractice liability concerns, scheduling eans. In Australia, where caesarean section rates are high
convenience for both physicians and patients, and the pref- relative to most OECD countries, a number of States have
erences of some women to have a caesarean section. None- developed clinical guidelines and required reporting of hos-
theless, caesarean delivery continues to result in increased pital caesarean section rates, including investigation of
maternal mortality, maternal and infant morbidity, and performance against the guidelines. These measures have
increased complications for subsequent deliveries, raising discouraged variations in practice and contributed to slow-
questions about the appropriateness of caesarean delivery ing down the rise in caesarean sections. Other countries
that may not be medically required. have reduced the gap in hospital payment rates between a
In 2013, caesarean section rates were lowest in Nordic caesarean section and a normal delivery, with the aim to
countries (Iceland, Finland, Sweden and Norway), Israel discourage the inappropriate use of caesareans (OECD,
and the Netherlands, with rates ranging from 15% to 16.5% 2014).
of all live births (Figure 6.22). They were highest in Turkey,
Mexico and Chile, with rates ranging from 45% to 50%.
Caesarean rates have increased since 2000 in most OECD
countries, with the average rate going up from 20% in 2000 Definition and comparability
to 28% in 2013 (Figure 6.23). The growth rate has been par-
The caesarean section rate is the number of caesarean
ticularly rapid in those countries that have the highest
deliveries performed per 100 live births.
rates now (Turkey, Mexico and Chile), as well as in Poland,
the Slovak Republic and the Czech Republic which used to In Mexico, the number of caesarean sections is esti-
have relatively low rates. In some countries, however, the mated based on public hospital reports and data
growth rate has slowed down since the mid-2000s and it obtained from National Health Surveys. Estimation is
has even come down slightly in Israel, Finland and Sweden. required to correct for under-reporting of caesarean
In Italy also, caesarean rates have come down significantly deliveries in private facilities. The combined number
in recent years, although they remain very high. The rates of caesarean deliveries is then divided by the total
have also come down in Spain. number of live births as estimated by the National
Population Council.
There can be substantial variations in caesarean rates
across regions and hospitals within the same country. In
Italy, there continues to be huge variations in caesarean
rates, driven by very large rates in the south of the country.
In Spain also, there are large variations across regions References
(OECD, 2014).
FHF Fdration hospitalire de France (2008), tude sur les
In several countries, there is evidence that private hospitals csariennes [Study on caesareans], Paris.
tend to perform more caesarean sections than public hos-
OECD (2014), Geographic Variations in Health Care: What Do We
pitals. In France, private for-profit hospitals authorised to
Know and What Can Be Done to Improve Health System Per-
provide maternity care for pregnancies without complica-
formance?, OECD Publishing, Paris,
tions have caesarean rates as high as public hospitals
http://dx.doi.org/10.1787/9789264216594-en.
which have to deal with more complicated cases (FHF,
2008). In Switzerland, caesarean sections have been found OFSP Office fdral de la sant publique (2013), Accouche-
to be substantially higher in private clinics (41%) than in ments par csarienne en Suisse [Births by Caesareans in
public hospitals (30.5%) (OFSP, 2013). Switzerland], Bern.

Caesarean sections
6.22. Caesarean section rates, 2013 (or nearest year)

Per 100 live births
50
50.4
44.7 45.2
40
36.0 36.1
34.6 35.0 35.3
32.1 32.5 32.5
30.7 30.9
30 28.5 28.8
27.6
26.1 26.3 26.8
25.2 25.8
23.0
22.2
20.2 20.8
20 19.5 20.1
16.4 16.5
15.2 15.4 15.6 15.8
10
Hu g al
s
nd
a
ly
il e
o
ey
A d
Po nd
s
Be ia
m
a
d
el
d
en
ay
Es ia
e
K i ar k
m
a in
ov r i a
G e e p.
d
C p.
m a
OE rg
Ir e 2
e
nd
an
i t e a li
re
xe a d
c
ar
n
ic
an
an
Cz alan
3
ra
n
en
It a
Re
iu
do
rk
Sw St at
rw
an
Ch
ed
la
la
la
CD
Sl us t
r tu
to
N e Sp
ex
Ko
i t e nm
ng
bo
Un us tr
rm
Lu an
la
el
nl
Is
lg
ov
Tu
Po
er
ng
Fr
Sw
No
ak
h
M
Ic
er
Ze
Fi
it z
Un De
ec
Sl
d
th
A
w
Ne

1 2 http://dx.doi.org/10.1787/888933281033
6.23. Changes in caesarean section rates, 2000 to 2013 (or nearest years)
2000 2006 2013
Per 100 live births
50
50.4
45.2
44.7
40
36.0
36.1
35.3
35.0
34.6
32.5
32.5
32.1
30.9
30.7
28.8
28.5
30
27.6
26.8
26.2
26.1
25.8
25.2
23.0
22.2
20.8
20.2
20.1
19.5
20
16.4
16.5
15.8
15.4
15.6
15.2
10
0
m
Hu g al
d
el
ay
Po d
d
en
Es ia
Be ia
Un De c e
K i ar k
ain
m a
d
C p.
OE rg
Ir e 2
A d
ov r i a
G e e p.
it z es
m
y
ia
y
Po d
a
ly
M le
o
ey
nd
xe a d
an
re
ar
an
n
an
Cz alan
ic
3
ra
en
i te r al
i
It a
Re
iu
do
rk
rw
Sw t at
an
Ch
ed
la
la
la
CD
Sl us t
to
N e Sp
r tu
R
ex
i t e nm
Ko
ng
bo
Lu an
rm
la
el
nl
Is
lg
ov
Tu
er
Un us t
ng
Fr
Sw
No
S
h
ak
Ic
er
Ze
Fi
ec
Sl
d
th
A
w
Ne

1 2 http://dx.doi.org/10.1787/888933281033

Ambulatory surgery
In the past few decades, the number of surgical procedures due to large tonsils. Although the operation is performed
carried out on a same-day basis has increased markedly under general anaesthesia, it is now carried out mainly as a
in OECD countries. Advances in medical technologies, same-day surgery in many countries, with children returning
particularly the diffusion of less invasive surgical home the same day (Figure 6.25). This is the case in Finland
interventions and better anaesthetics, have made this (where the share of same-day surgery has increased greatly
development possible. These innovations have improved since 2000), Canada, Belgium, the Netherlands, Sweden and
patient safety and health outcomes, and have also in many Norway, where more than half of all tonsillectomy are
cases reduced the unit cost per intervention by shortening now performed on a same-day basis. This proportion is
the length of stay in hospitals. However, the impact of the much lower in Austria (where virtually no tonsillectomy
rise in same-day surgeries on health spending depends not is performed on a same-day basis), Luxembourg, Ireland
only on changes in their unit cost, but also on the growth in and Germany. These large differences in the share of same-
the volume of procedures performed. There is also a need day surgery may reflect variations in the perceived risks of
to take into account any additional cost related to post- postoperative complications, or simply clinical traditions of
acute care and community health services following the keeping children for at least one night in hospital after the
interventions. operation.
Cataract surgery and tonsillectomy (the removal of tonsils, In some countries, there has been a strong rise however in
glands at the back of the throat, mainly performed on the share of tonsillectomy performed as day surgery since
children) provide good examples of high-volume surgeries 2000. Beyond Finland which is now leading the way, the
which are now carried out mainly on a same-day basis in share of same-day surgery has increased rapidly over the
many OECD countries. past decade in the United Kingdom, Denmark, Portugal and
Day surgery now accounts for over 90% of all cataract Italy. In France, there has virtually been no increase in the
surgeries in a majority of OECD countries (Figure 6.24). In share of day surgery for tonsillectomy since 2000, while this
several countries, nearly all cataract surgeries are performed share has decreased slightly in Israel and Switzerland.
as day cases. However, the use of day surgery is still relatively There appears to be ample room for further growth in day
low in Poland, Hungary and the Slovak Republic, where surgery for tonsillectomy in these countries to reduce cost
they still account for less than half of all cataract surgeries. without affecting patient outcomes.
While this may be partly explained by limitations in the
data coverage of outpatient activities in hospital or outside
hospital, this may also reflect more advantageous reim-
bursement for inpatient stays or constraints on the develop-
ment of day surgery. In Hungary, the government recently Cataract surgery consists of removing the lens of the
abolished the budget cap on the number of same-day surgery eye because of the presence of cataracts which are
that can be performed in hospital; this is expected to lead to a partially or completely clouding the lens, and replacing
steady increase in the number of cataract and other surgeries it with an artificial lens. It is mainly performed on
performed as day cases. elderly people. Tonsillectomy consists of removing the
The number of cataract surgeries performed on a same-day tonsils, glands at the back of the throat. It is mainly
basis has grown very rapidly since 2000 in many countries, performed on children.
such as Portugal and Austria (Figure 6.24). Whereas fewer The data for several countries do not include outpa-
than 10% of cataract surgeries in Portugal were performed tient cases in hospital or outside hospital (i.e., patients
on a same-day basis in 2000, this proportion has increased who are not formally admitted and discharge), leading
to 92%. In Austria, the share of cataract surgeries per- to some under-estimation. In Ireland, Mexico, New
formed as day cases increased from 1% only in 2000 to 67% Zealand and the United Kingdom, the data only include
in 2013. The number of cataract surgeries carried out as day cataract surgeries carried out in public or publicly-
cases has also risen rapidly in France, Ireland, Switzerland funded hospitals, excluding any procedures performed
and Luxembourg, although there is still room for further in private hospitals (in Ireland, it is estimated that
development. approximately 15% of all hospital activity is under-
Tonsillectomy is one of the most frequent surgical proce- taken in private hospitals). Data for Portugal relate only
dures on children, usually performed on children suffering to public hospitals on the mainland. Data for Spain only
from repeated or chronic infections of the tonsils or suffer- partially include activities in private hospitals.
ing from breathing problems or obstructive sleep apnea

Ambulatory surgery
6.24. Share of cataract surgeries carried out as ambulatory cases, 2000 and 2013 (or nearest years)
2000 2013
%
100 100 99
100 99 99 98 98 98 98 96 96 96 96 95
93 92 91 89
87
83
81
80 78 77
69
67
64
60
53
47
40 37
27
20
nd
da
nd
ly
ria
p.
y
nd
a
en
el
g
m
k
a in
ia
ay
li a
ce
29
y
d
p.
ga
ke
nd
an
ni
re
ar
ar
ic
an
an
ur
ra
en
It a
Re
Re
iu
do
rw
an
ra
ed
na
la
la
la
CD
st
to
r tu
Sp
ex
Ko
nm
ng
bo
r
rm
la
nl
al
Is
lg
ov
Tu
Po
Ir e
er
st
ng
Au
Fr
Sw
Ca
No
Es
ak
h
M
er
Ze
Fi
OE
Be
Hu
m
Po
Au
it z
De
Ge
ec
Sl
Ki
ov
th
xe
w
Cz
Sw
Ne
d
Sl
Ne
Lu
i te
Un

1 2 http://dx.doi.org/10.1787/888933281044
6.25. Share of tonsillectomy carried out as ambulatory cases, 2000 and 2013 (or nearest years)
2000 2013
%
100
84
80
75
71
68
63
60
51 50
47
45
42
40
35 34
31 30
22
20
20 18
11
9
7
5
4 3 3
0.1
0
da
en
nd
d
ay
24
ly
a in
el
y
ce
li a
nd
nd
ria
ga
nd
ke
an
ar
re
an
ic
an
ur
ra
It a
iu
do
rw
an
ra
na
ed
la
la
la
CD
st
r tu
Sp
ex
nm
Ko
bo
r
rm
nl
la
al
Is
lg
Tu
Po
Ir e
er
st
ng
Au
Fr
Sw
Ca
No
M
er
Ze
Fi
OE
Be
m
Po
Au
it z
De
Ge
Ki
th
xe
w
Sw
Ne
Ne
Lu
i te
Un

1 2 http://dx.doi.org/10.1787/888933281044

7. ACCESS TO CARE
Coverage for health care
Unmet needs for medical care and dental care
Out-of-pocket medical expenditure
Geographic distribution of doctors
Waiting times for elective surgery
international law.

7. ACCESS TO CARE
Health care coverage through public or private health security system. The Netherlands has the largest supple-
insurance promotes access to medical goods and services, mentary market (86% of the population), followed by
and provides financial security against unexpected or serious Israel (83%), whereby private insurance pays for prescrip-
illness. However, the percentage of the population covered by tion drugs and dental care that are not publicly reimbursed.
such insurance does not provide a complete indicator of Duplicate markets, providing faster private-sector access to
accessibility, since the range of services covered and the medical services where there are waiting times in public
degree of cost-sharing applied to those services also affects systems, are largest in Ireland (45%) and Australia (47%).
access to care. The population covered by private health insurance has
Most OECD countries have achieved universal (or near- increased in some OECD countries over the past decade,
universal) coverage of health care costs for a core set of whereas it has decreased in others. It increased in some
services, which usually include consultations with doctors Nordic countries such as Denmark where one-third of the
and specialists, tests and examinations, and surgical and population now has a private health insurance (up from
therapeutic procedures (Figure 7.1). Generally, dental care less 10% in 2005) and in Finland where the growth has been
and pharmaceutical drugs are partially covered, although more modest, but remains almost non-existent in other
there are a number of countries where these services must Nordic countries. Private health insurance coverage has
be purchased separately (OECD, 2015). also increased in Australia and Korea, but it has come
Three OECD countries do not have universal (or near-uni- down in Ireland, New Zealand and the United Kingdom
versal) health coverage: Greece, the United States and (Figure 7.3).
Poland. In Greece, the economic crisis has reduced health The importance of private health insurance is linked to sev-
insurance coverage among people who have become long- eral factors, including gaps in access to publicly financed ser-
term unemployed, and many self-employed workers have vices, government interventions directed at private health
also decided not to renew their health insurance plan insurance markets, and historical development.
because of reduced disposable income. However, since
June 2014, uninsured people are covered for prescribed
pharmaceuticals and for services in emergency depart-
ments in public hospitals, as well as for non-emergency Definition and comparability
hospital care under certain conditions (Eurofound, 2014). In
Coverage for health care is defined here as the share
the United States, coverage is provided mainly through
of the population receiving a core set of health care
private health insurance, and 54% of the population had this
goods and services under public programmes and
for their basic coverage in 2014. Publicly financed coverage
through private health insurance. It includes those
insured 34.5% of the population (the elderly, people with
covered in their own name and their dependents. Public
low income or with disabilities), leaving 11.5% of the popu-
coverage refers both to government programmes,
lation without insurance. The percentage of the population
generally financed by taxation, and social health
uninsured decreased from 14.4% in 2013 to 11.5% in 2014,
insurance, generally financed by payroll taxes. Take-
following the implementation of the Affordable Care Act
up of private health insurance is often voluntary,
which is designed to expand health insurance coverage
although it may be mandatory by law or compulsory
(Cohen and Martinez, 2015). In Poland, a tightening of the
for employees as part of their working conditions.
law in 2012 made people lose their social health insurance
Premiums are generally non-income-related, although
coverage if they fail to pay their contribution. However, it is
the purchase of private coverage can be subsidised by
common for uninsured people who need medical care to go
government.
to emergency services in hospital, where they will be
encouraged to get an insurance.
Basic primary health coverage, whether provided through
public or private insurance, generally covers a defined
References
basket of benefits, in many cases with cost-sharing. In
some countries, additional health coverage can be purchased Cohen, R.A. and M.E. Martinez, M.E. (2015), Health Insurance
through private insurance to cover any cost-sharing left Coverage: Early Release of Estimates from the National Health
after basic coverage (complementary insurance), add Interview Surve, 2014, National Center for Health Statis-
additional services (supplementary insurance) or provide tics, June.
faster access or larger choice to providers (duplicate insur- Eurofound (2014), Access to Healthcare in Times of Crisis,
ance). Among the 34 OECD countries, nine have private Dublin.
coverage for over half of the population (Figure 7.2). OECD (2015), Measuring Health Coverage, OECD, Paris,
Private health insurance offers 95% of the French population available at: www.oecd.org/els/health-systems/measuring-
complementary insurance to cover cost-sharing in the social health-coverage.htm.

7. ACCESS TO CARE
7.1. Health insurance coverage for a core set of services, 7.2. Private health insurance coverage, by type, 2013
2013 (or nearest year)
Total public coverage Primary Complementary

Primary private health coverage Supplementary Duplicate
Australia 100.0 France 95.0

Canada 100.0
Czech Rep. 100.0 Netherlands 86.0
Denmark 100.0 Israel 82.9
Finland 100.0
Belgium 81.3
Iceland 99.8 0.2
Ireland 100.0 Slovenia 72.8
Israel 100.0 Canada 67.0
Italy 100.0
Japan 100.0 Korea 61.0
Korea 100.0 United States 60.1
New Zealand 100.0
Australia 54.9
Norway 100.0
Portugal 100.0 Luxembourg 49.7
Slovenia 100.0 Ireland 44.6
Sweden 100.0
Switzerland 100.0 Austria 35.2
United Kingdom 100.0 Denmark 33.0
Austria 99.9
Germany 33.0
France 99.9
Spain 99.0 0.9 New Zealand 29.7
Germany 88.8 11.0 Switzerland 27.9
Netherlands 99.8
Portugal 21.1
Belgium 99.0
Mexico 91.6 7.3 Chile 18.3
Chile 79.9 18.3
Finland 14.7
Turkey 98.1
Luxembourg 96.4 Spain 13.4
Hungary 96.0 Greece 12.5
Colombia 91.7 4.0
United Kingdom 10.6
Slovak Republic 94.6
Estonia 93.6 Mexico 7.3
Poland 91.6
Turkey 5.6
United States (2014) 34.5 54.0
Greece 79.0 Iceland 0.2
0 20 40 60 80 100 0 20 40 60 80 100
Percentage of total population Percentage of total population
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en. Note: Private health insurance can be both duplicate and supplementary
1 2 http://dx.doi.org/10.1787/888933281052 in Australia; both complementary and supplementary in Denmark and
Korea; and duplicate, complementary and supplementary in Israel and
Slovenia.
1 2 http://dx.doi.org/10.1787/888933281052
7.3. Evolution in private health insurance coverage, 2005 to 2013

2005 2010 2013
Percentage of total population
100 95.0
86.0
81.3 82.9
80
72.8
67.0
60.1 61.0
60 54.9
44.6
40 33.0 35.2
29.7 33.0
27.9
25.1
21.1
20 14.7
10.6 12.5
5.6 7.3
0
y
il
nd
nd
es
el
ce
o
ce
ria
li a
da
ia
m
ga
ke
nd
an
re
ar
az
an
an
ic
ra
en
iu
do
at
an
ee
ra
na
la
la
st
r tu
ex
Ko
nm
r
Br
rm
la
nl
al
Is
lg
ov
St
Tu
Ir e
er
st
ng
Au
Gr
Fr
Ca
M
er
Ze
Fi
Be
Po
Au
it z
De
Ge
Sl
d
Ki
th
i te
w
Sw
Ne
d
Ne
Un
i te
Un

1 2 http://dx.doi.org/10.1787/888933281052

7. ACCESS TO CARE
Access to health care may be prevented for a number of Sweden. There were large gaps in unmet care needs
reasons related either to the functioning of the health care between high and low income people in the Czech Republic,
system itself (like the cost of a doctor visit or medical France and the United States.
treatment, the distance to the closest health care facility, or It is important to consider self-reported unmet care needs
waiting lists) or to personal reasons (like fear of not being in conjunction with other indicators of potential barriers to
understood by the doctor or not having the time to seek access, such as the extent of health insurance coverage and
care). People who forgo health care when they need it may the amount of out-of-pocket payments. For instance, the
jeopardize their health status. Any inequalities in unmet rate of unmet care needs decreased in Germany, following
care needs may result in poorer health status and increase the abolition of a quarterly fee of EUR 10 charged to
health inequalities. patients.
Around 3% of the population on average across Europe Strategies to improve access to care for disadvantaged or
reported unmet needs for medical care due to cost, travel- underserved populations need to tackle both financial and
ling distance and waiting lists in 2013, according to the non-financial barriers, as well as promoting an adequate
European Union Statistics on Income and Living Conditions supply and proper distribution of doctors, dentists and
survey. But there are large variations across countries other medical practitioners (see the indicator on Geographic
(Figure 7.4). Larger shares of the population report unmet distribution of doctors).
needs in Latvia, Greece, Poland and Estonia, while less
than 1% of the population reported unmet needs in the
Netherlands, Austria, Spain, Luxembourg and the Czech
Republic. Unmet needs for medical examination are consis-
tently higher among people in low income groups compared
with those in high income groups (Figure 7.4). The gap was Data on unmet care needs come from two main
particularly large in 2013 in Latvia, Italy and Greece. sources. First, the European Union Statistics on
A higher proportion of the population in European coun- Income and Living Conditions survey (EU-SILC) ask
tries reports some unmet needs for dental care than for survey respondents whether there was a time in the
medical care, reflecting the fact that public coverage for previous 12 months when they felt they needed a
dental care is generally lower in most countries. Latvia medical or dental examination but did not receive it,
(18.9%), Portugal (14.3%), Iceland (11.1%) and Italy (10%), followed by a question as to why the need for care was
reported the highest rates of unmet needs for dental care unmet (with the reasons including that care was too
among European countries in 2013 (Figure 7.5). In these expensive, the waiting time was too long, the travel-
countries, there were large inequalities in unmet dental ling distance to receive care was too far, a lack of time,
care needs between low and high income groups. On aver- or that they wanted to wait and see if problem got
age across European countries covered under this survey, better on its own). The data presented in Figures 7.4
nearly 10% of low income people reported having some and 7.5 cover unmet care needs due to cost, waiting
unmet needs for dental care, compared with 1.6% for high time and travelling distance.
income people. The second source is the 2013 Commonwealth Fund
Countries participating in the Commonwealth Fund Inter- International Health Policy Survey which asks
national Health Policy Survey, and other countries using whether people did not visit a doctor when they had a
the same survey module, collect data on unmet care needs medical problem, skipped a medical test, treatment,
for doctor visits, medical care and prescribed pharmaceutical or follow-up that was recommended by a doctor, or
drugs due to cost. As expected, the results from these surveys did not fill prescription for medicines or skipped
show consistently higher unmet care needs for financial doses because of cost in the past year. This survey was
reasons among low income people compared with high carried out in eleven countries. Similar questions
income people (Figure 7.6). The largest proportions of were also asked in the national survey in the Czech
unmet care needs in 2013 were found in the United States, Republic a few years earlier (2010).
while the United Kingdom had the lowest rates, followed by

7. ACCESS TO CARE
7.4. Unmet care needs for medical examination, 7.5. Unmet care needs for dental examination,
by income level, 2013 by income level, 2013
High income Average Low income High income Average Low income
Netherlands Netherlands
Austria Austria
Spain Luxembourg
Luxembourg Czech Rep.
Czech Rep. Germany
Switzerland Slovak Rep.
Denmark United Kingdom
Norway Belgium
Germany Hungary
United Kingdom Denmark
Belgium Lithuania
Slovak Rep. Finland
Sweden Switzerland
Hungary Poland
France OECD23
Portugal Sweden
OECD23 France
Lithuania Norway
Ireland Ireland
Iceland Spain
Finland Estonia
Italy Greece
Estonia Italy
Poland Iceland
Greece Portugal
Latvia Latvia
0 5 10 15 20 25 0 10 20 30 40
% %
Note: Unmet care needs for following reasons: too expensive, too far to Note: Unmet care needs for following reasons: too expensive, too far to
travel, or waiting time. travel, or waiting time.
Source: EU-SILC 2013. Source: EU-SILC 2013.
7.6. Unmet care needs due to cost, by income level, 2013
Below average income Above average income

%
50 49
40
30 29 29
27
24
23
21 21
20 20
20 19
16 16
14
12
11
10
10 9
8
7
6
5
4
3
0
nd
es
m
en
ay
da
li a
ce
d
p.
nd
an
an
do
Re
rw
at
an
ra
ed
na
la
rm
la
al
St
er
st
ng
Fr
Sw
Ca
No
er
Ze
Au
it z
Ge
ec
d
Ki
th
i te
w
Cz
Sw
Ne
d
Ne
Un
i te
Un
Note: Either did not visit doctor when they had a medical problem, did not get recommended care or did not fill/skipped prescription.
Source: 2013 Commonwealth Fund International Health Policy Survey, complemented with data from the national survey for the Czech Republic (2010).
1 2 http://dx.doi.org/10.1787/888933281066

7. ACCESS TO CARE
Financial protection through public or private health insur- Poland, the Czech Republic, Hungary and Canada, half of
ance substantially reduces the amount that people pay out-of-pocket payments or more are for pharmaceuticals.
directly for medical care, yet in some countries the burden In some of these countries, in addition to co-payments for
of out-of-pocket spending can still create barriers to health prescribed pharmaceuticals, spending on over-the-counter
care access and use. Households that face difficulties pay- medicines for self-medication has been historically high.
ing medical bills may delay or even forgo needed health Payments for dental treatment also play a significant part
care. On average across OECD countries, 19% of health in household medical spending, accounting for 20% of all
spending is paid directly by patients (see indicator out-of-pocket expenditure across OECD countries. In
Financing of health care in Chapter 9 on Health expen- Estonia, Norway, Denmark and Spain, this figure reaches
diture). 30% or more. This can at least partly be explained by the
In contrast to publicly-funded care, out-of-pocket pay- limited public coverage for dental care in these countries
ments rely on peoples ability to pay. If the financing of compared with a more comprehensive coverage for other
health care becomes more dependent on out-of-pocket categories of care. The significance of therapeutic appliances
payments, the burden shifts, in theory, towards those who (eye-glasses, hearing aids, etc.) in households total medical
use services more, and possibly from high to low income spending differs widely, but is as much as 33% in the
households, where health care needs are higher. In practice, Netherlands. The average across OECD countries was 13%.
many countries have policies in place to protect certain More than half of this relates to eye-care products. In many
population groups from excessive out-of-pocket payments. countries, public coverage is limited to a contribution to the
These consist in partial or total exemptions for social assis- cost of lenses. Frames are often exempt from public coverage,
tance beneficiaries, seniors, or people with chronic diseases leaving private households to bear the full cost if they are
or disabilities by capping direct payments, either in absolute not covered by complementary private insurance.
terms or as a share of income (Paris et al., 2010; OECD,
2015).
The burden of out-of-pocket medical spending can be mea-
sured either by its share of total household income or its Definition and comparability
share of total household consumption. The share of house-
hold consumption allocated to medical spending varied Out-of-pocket payments are expenditures borne
considerably across OECD countries in 2013, ranging from directly by a patient where neither public nor private
less than 1.5% of total household consumption in countries insurance cover the full cost of the health good or
such as Turkey, the Netherlands, France and the United service. They include cost-sharing and other expen-
Kingdom, to more than 4% in Korea, Switzerland and diture paid directly by private households and should
Greece (Figure 7.7). On average across OECD countries, 2.8% also include estimations of informal payments to
of household spending went towards medical goods and health care providers. Only expenditure for medical
services. spending (i.e. current health spending less expenditure
for the health part of long-term care) is presented
Health systems in OECD countries differ in the degree of
here, because the capacity of countries to estimate
coverage for different health services and goods. In most
private long-term care expenditure varies widely.
countries, the degree of coverage is higher for hospital care
and doctor consultations than for pharmaceuticals, dental Household final consumption expenditure covers all
care and eye care (Paris et al., 2010; OECD, 2015). Taking into purchases made by resident households to meet their
account these differences and also the relative importance everyday needs such as food, clothing, rent or health
of these different spending categories, there are significant services.
variations between OECD countries in the breakdown of
the medical costs that households have to bear themselves.
In most OECD countries, curative care (including both
References
inpatient and outpatient care) and pharmaceuticals are the
two main spending items for out-of-pocket expenditure OECD (2015), Measuring Health Coverage, OECD, Paris,
(Figure 7.8). On average, these two components account for available at: www.oecd.org/els/health-systems/measuring-
two-thirds of all medical spending by households, but the health-coverage.htm.
importance varies between countries. In Luxembourg, Paris, V., M. Devaux and L. Wei (2010), Health Systems
Belgium and Switzerland, household payments for inpa- Institutional Characteristics: A Survey of 29 OECD Coun-
tient and outpatient curative care account for close to 50% tries, OECD Health Working Paper, No. 50, OECD Publish-
of total household outlays. In other countries such as ing, Paris, http://dx.doi.org/10.1787/5kmfxfq9qbnr-en.

7. ACCESS TO CARE
7.7. Out-of-pocket medical spending as a share of final household consumption, 2013 (or nearest year)
%
5
4.7
4.5
4.1 4.0 4.0

4 3.9 3.8
3.4 3.4
3.2 3.2 3.2 3.2 3.1
3.0 2.9
3 2.9 2.8 2.8 2.8
2.6 2.6
2.5
2.4
2.3
2.2 2.2 2.1
2.0
2 1.8
1.8
1.4 1.4 1.3

1.2
Tu 1
l
Gr d
Fi m
OE el
De tes
Un No 4
Ca ia
it z rea
Au and
St y
Po k
Es d
da
Ze n
ec nd
S l e p.
G ni a
Un xem a n y
Ki rg
er e
M e
Hu i c o
Po ar y
S w il e
en
a in
ak a
p.
Ic l y
B e nd
Au d
ria
ey
ga
s
i te r wa
ov li
th nc
c
ar
n
an
w pa
3
ra
n
It a
nd
Re
iu
do
d ou
rk
ee
Ch
Sl s tr a
ed
na
la
la
Cz ala
st
CD
to
r tu
e
Sp
R
ex
S w Ko
nm
ng
L u er m
Ne Fr a
l
el
nl
Ne Ja
Is
lg
ov
Ir e
er
la
ng
i te b
h
d
Note: This indicator relates to current health spending excluding long-term care (health) expenditure.
1. The value for the Netherlands is underestimated as it excludes compulsory co-payments by patients to health insurers (if these were taken into
account this would double the share).
1 2 http://dx.doi.org/10.1787/888933281072
7.8. Shares of out-of-pocket medical spending by services and goods, 2013 (or nearest year)
% Curative care 1 Dental care Pharmaceuticals Therapeutic appliances 2 Other

100 3
1
8 5 8 7 6
5 10 10 12
13 14 13 14 13 14
17 12
90 21 18
28 27
18
80 15 28 33
32 44
26 49 38 35 28 42
70 45 36 36 26
38
62 34
26 51 49
60 22 41 28
11 20
38
50 19 28
8 10 19 30
40 19 20 26 29 31
18
9 21 31
32
30 8 13 17 21
49 49 48 15
46
20 41 39 37
33 31 30 30 30 28 28 26 25 25 24 22
10 19 19 17 17 15
0
g
nd
en
ria
li a
ce
23
ia
p.
ay
s
y
p.
nd
ain
da
ur
nd
an
ni
re
ar
ar
pa
an
an
en
Re
Re
iu
rw
an
ra
ed
na
la
la
CD
st
to
Sp
bo
Ko
nm
ng
rm
la
el
Ja
nl
lg
ov
Po
er
st
Au
Fr
Sw
Ca
No
Es
ak
h
Ic
er
Fi
OE
m
Be
Hu
Au
it z
De
Ge
ec
Sl
ov
th
xe
Cz
Sw
Ne
Sl
Lu
Note: This indicator relates to current health spending excluding long-term care (health) expenditure.
1. Including rehabilitative and ancillary services.
2. Including eye care products, hearing aids, wheelchairs, etc.
1 2 http://dx.doi.org/10.1787/888933281072

7. ACCESS TO CARE
Access to medical care requires an adequate number and their practice and recurrent payments such as income
proper distribution of doctors in all parts of the country. guarantees and bonus payments (Ono et al., 2014).
Any shortage of doctors in certain regions can increase In France, the Ministry of Health launched at the end of
travel times or waiting times for patients, and result in 2012 a Health Territory Pact to promote the recruitment
unmet care needs. The uneven distribution of doctors is an and retention of doctors and other health workers in
important policy issue in most OECD countries, especially underserved regions. This Pact includes a series of measures
in those countries with remote and sparsely populated to facilitate the establishment of young doctors in under-
areas, and those with deprived urban regions which may served areas, to improve their working conditions (notably
also be underserved. through the creation of new multi-disciplinary medical
The overall number of doctors per capita varies across OECD homes allowing physicians and other health professionals
countries from lows of about two per 1 000 population in to work in the same location), to promote tele-medicine,
Chile, Turkey and Korea, to highs of five and more in Greece and to accelerate the transfer of competences from doctors
and Austria (see the indicator on Doctors in Chapter 5). to other health care providers (Ministry of Health, 2015). The
Beyond these cross-country differences, the number of first results from this programme are promising, although it is
doctors per capita also often varies widely across regions still too early to reach any definitive conclusions on the cost-
within the same country (Figure 7.9). A common feature in effectiveness of various measures.
many countries is that there tends to be a concentration of In Germany, the number of practice permits for new ambu-
physicians in capital cities. For example, Austria, Belgium, latory care physicians in each region is regulated, based on
the Czech Republic, Greece, Mexico, Portugal, the Slovak a national service delivery quota.
Republic and the United States have a much higher density
The effectiveness and cost of different policies to promote
of doctors in their national capital region.
a better distribution of doctors can vary significantly, with
The density of physicians is consistently greater in urban the impact likely to depend on the characteristics of each
regions, reflecting the concentration of specialised services health system, the geography of the country, physician
such as surgery and physicians preferences to practice in behaviours, and the specific policy and programme design.
urban settings. There are large differences in the density of Policies should be designed with a clear understanding of the
doctors between predominantly urban and rural regions in interests of the target group in order to have any significant
France, Australia and Canada, although the definition of and lasting impact (Ono et al., 2014).
urban and rural regions varies across countries. The distri-
bution of physicians between urban and rural regions is
more equal in Japan and Korea, but there are generally
fewer doctors in these two countries (Figure 7.10).
Doctors may be reluctant to practice in rural regions due to
concerns about their professional life (including their Regions are classified in two territorial levels. The
income, working hours, opportunities for career develop- higher level (Territorial Level 2) consists of large
ment, isolation from peers) and social amenities (such as regions corresponding generally to national adminis-
educational opportunities for their children and profes- trative regions. These broad regions may contain a
sional opportunities for their spouse). mix of urban, intermediate and rural areas. The lower
A range of policy levers may influence the choice of prac- level is composed of smaller regions classified as pre-
tice location of physicians, including: 1) the provision of dominantly urban, intermediate or rural regions,
financial incentives for doctors to work in underserved although there are variations across countries in the
areas; 2) increasing enrolments in medical education classification of these regions.
programmes of students coming from specific social or
geographic background, or decentralising the location of
medical schools; 3) regulating the choice of practice location
of doctors (for new medical graduates or foreign-trained References
doctors); and 4) re-organising health service delivery to Ministry of Health (2015), Le Pacte territoire sant [Health Terri-
improve the working conditions of doctors in underserved tory Pact], available at: www.sante.gouv.fr/le-pacte-territoire-
areas and find innovative ways to improve access to care sante-pour-lutter-contre-les-deserts-medicaux,12793.html.
for the population.
Ono, T., M. Schoenstein and J. Buchan (2014), Geographic
Many OECD countries provide different types of financial Imbalances in Doctor Supply and Policy Responses,
incentives to attract and retain doctors in underserved OECD Health Working Papers, No. 69, OECD Publishing,
areas, including one-time subsidies to help them set up Paris, http://dx.doi.org/10.1787/5jz5sq5ls1wl-en.

7. ACCESS TO CARE
7.9. Physician density, by Territorial Level 2 regions, 2013 (or nearest year)
Australia
Austria Vienna
Belgium Brussels
Canada
Chile
Czech Rep. Prague
Denmark Copenhagen Region
Estonia
Finland Helsinki
France
Germany
Greece Athens Region
Hungary
Israel
Italy
Japan
Korea
Luxembourg
Mexico Mexico city
Netherlands
New Zealand
Norway
Poland
Portugal Lisbon
Slovak Rep. Bratislava
Slovenia
Spain
Sweden
Switzerland
Turkey
United Kingdom Washington, DC.
United States
0 1 2 3 4 5 6 7 8 9 10
Density per 1 000 population
Source: OECD Regions at a Glance 2015.

1 2 http://dx.doi.org/10.1787/888933281083
7.10. Physicians density in predominantly urban and rural regions, selected countries, 2013 (or nearest year)
Urban areas Rural areas

Density per 1 000 population
5
4.5 4.6
4.4
4.1
4
3.6
3.3
3
2.5 2.5
2.2 2.2
2.1
2
1.7
1.4
1.0
1
0
ce
li a
a
da
en
n
re
an
pa
an
ra
ed
na
Ko
nl
Ja
st
Fr
Sw
Ca
Fi
Au
Note: The classification of urban and rural regions varies across countries.
Source: Australia: AIHW National Health Workforce Data Set (NHWDS) 2013; Canada: Scotts Medical Database, 2013, Canadian Institute for Health
Information; France: RPPS mdecins au 1er janvier 2015; Other: OECD Regions at a Glance 2015.
1 2 http://dx.doi.org/10.1787/888933281083

7. ACCESS TO CARE
Long waiting times for health services is an important policy Waiting times for knee replacement has come down in recent
issue in many OECD countries (Siciliani et al., 2013). Long years in the Netherlands, Denmark, Finland and Estonia,
waiting times for elective (non-emergency) surgery, such as although it remains very long in Estonia (Figure 7.13).
cataract surgery, hip and knee replacement, generates dissat- Over the past decade, waiting time guarantees have
isfaction for patients because the expected benefits of become the most common policy tool to tackle long waiting
treatments are postponed, and the pain and disability times in several countries. This has been the case in Finland
remains. While long waiting times is considered an important where a National Health Care Guarantee was introduced in
policy issue in many countries, this is not the case in others 2005 and led to a reduction in waiting times for elective
(e.g., Belgium, France, Germany, Japan, Korea, Luxembourg, surgery (Jonsson et al., 2013). In England, since April 2010,
Switzerland, United States). the NHS Constitution has set out a right to access certain
Waiting times is the result of a complex interaction services within maximum waiting times or for the NHS to
between the demand and supply of health services, where take all reasonable steps to offer a range of alternative
doctors play a critical role on both sides. The demand for providers if this is not possible (Smith and Sutton, 2013).
health services and elective surgery is determined by the These guarantees are only effective if they are enforced.
health status of the population, progress in medical tech- There are two main approaches to enforcement: setting
nologies (including the increase ease of many procedures waiting time targets and holding providers accountable for
like cataract which can now be performed as day surgery), achieving these targets; or allowing patients to choose
patient preferences (including their weighting of the alternative health providers (including the private sector) if
expected benefits and risks), and the extent of cost-sharing they have to wait beyond a maximum amount of time
for patients. However, doctors play a crucial role in convert- (Siciliani et al., 2013).
ing the demand for better health from patients in a demand
for medical care. On the supply side, the availability of
different categories of surgeons, anaesthesists and other
staff involved in surgical procedures, as well as the supply
of the required medical and hospital equipment influence There are at least two ways of measuring waiting
surgical activity rates. times for elective procedures: 1) measuring the waiting
The measure used here focuses on waiting times from the times for patients treated in a given period; or
time that a medical specialist adds a patient to the waiting 2) measuring waiting times for patients still on the list
list to the time that the patient receives the treatment. Both at a point in time. The data reported here relate to the
the average waiting time and the median are presented. first measure (data on the second measure are available
Because some patients wait for very long times, the average in the OECD health database). The data come from
is usually greater than the median. administrative databases (not surveys). Waiting times
In 2013/14, the average waiting times for cataract surgery are reported both in terms of the average and the
was just over 30 days in the Netherlands, but much longer median. The median is the value which separates a
in Chile, Estonia and Poland (Figure 7.11). In the United distribution in two equal parts (meaning that half the
Kingdom, the average waiting times for cataract surgery patients have longer waiting times and the other half
was 72 days in 2013, slightly up from 66 days in 2007. In lower waiting times). Compared with the average, the
Portugal and Spain, waiting times fell between 2007 and median minimises the influence of outliers (patients
2010, but has increased since then. In Finland and Estonia, with very long waiting times).
waiting times for cataract surgery has fallen steadily,
although the average waiting times remains high in
Estonia.
References
In 2013/14, the average waiting times for hip replacement
Jonsson, P.M. et al. (2013), Finland, Part II, Chapter 7 in
was just over 40 days in the Netherlands, but around
Waiting Time Policies in the Health Sector: What Works?,
250 days in Estonia and over 300 days in Chile and Poland
OECD Publishing, Paris,
(Figure 7.12). The median waiting times was around 40 days
http://dx.doi.org/10.1787/9789264179080-en.
in Denmark, 60 days in Israel, and between 75 and 90 days
in Hungary, the United Kingdom, Portugal, Canada and Siciliani, L., M. Borowitz and V. Moran (2013), Waiting Time
New Zealand. It reached between 120 and 150 days in Policies in the Health Sector: What Works?, OECD Publishing,
Spain, Norway and Estonia, and over 200 days in Poland Paris, http://dx.doi.org/10.1787/9789264179080-en.
and Chile. As is the case for cataract surgery, waiting times Smith, P. and M. Sutton (2013), United Kingdom, Part II,
for hip replacement fell in Portugal and Spain between 2007 Chapter 16 in Waiting Time Policies in the Health Sector:
and 2010, but has gone up since then. What Works?, OECD Publishing, Paris,
http://dx.doi.org/10.1787/9789264179080-en.

7. ACCESS TO CARE
7.11. Cataract surgery, waiting times from specialist assessment to treatment, 2007 to 2014 (or 2013)
Days 2007 2010 2014 (or 2013)

450
400
350
300
250
200
150
100
50
0
Portugal
Portugal
Finland
Netherlands
Israel
Spain
Chile
Finland
Denmark
New Zealand
Hungary
Norway
Estonia
Poland
Hungary
Israel
Denmark
New Zealand
Estonia
Spain
Chile
Australia
Norway
Poland
Canada
United Kingdom
United Kingdom
Average Median
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en. 1 2 http://dx.doi.org/10.1787/888933281097
7.12. Hip replacement, waiting times from specialist assessment to treatment, 2007 to 2014 (or 2013)
Days 2007 2010 2014 (or 2013)

500
450
400
350
300
250
200
150
100
50
0
Portugal
Portugal
Finland
Finland
Netherlands
Spain
Estonia
Chile
Chile
Norway
Hungary
Poland
Denmark
Israel
Hungary
New Zealand
Spain
Australia
Norway
Estonia
Poland
Denmark
New Zealand
Israel
United Kingdom
Canada
United Kingdom
Average Median
7.13. Knee replacement, waiting times from specialist assessment to treatment, 2007 to 2014 (or 2013)
Days 2007 2010 2014 (or 2013)

600
500
400
300
200
100
0
New Zealand
Israel
Hungary
Israel
Canada
New Zealand
Hungary
Chile
Spain
Denmark
Norway
Portugal
Spain
Chile
Denmark
Norway
Portugal
Australia
Netherlands
United Kingdom
Finland
Estonia
Poland
United Kingdom
Finland
Estonia
Poland
Average Median

8. QUALITY OF CARE
Avoidable hospital admissions
Diabetes care
Prescribing in primary care
Mortality following acute myocardial infarction (AMI)
Mortality following stroke
Waiting times for hip fracture surgery
Surgical complications
Obstetric trauma
Care for people with mental health disorders
Screening, survival and mortality for cervical cancer
Screening, survival and mortality for breast cancer
Survival and mortality for colorectal cancer
Childhood vaccination programme
Influenza vaccination for older people
Patient experience with ambulatory care

8. QUALITY OF CARE
Most health systems have developed a primary level of improvement in t he quality of p rimary care. The
care whose functions include health promotion and disease approaches countries are taking to improve the quality of
prevention, managing new health complaints, as well as primary care have been described in a series of country
long-term conditions and referring patients to hospital-based reviews undertaken by OECD. Israels Quality Indicators for
services when appropriate. A key aim is to keep people well, Community Health Care program provides an example of how
by providing a consistent point of care over the longer-term, publicly reported information on care is used to incentivise
tailoring and co-ordinating care for those with multiple providers to develop better services (OECD, 2012).
health care needs and supporting the patient in self-
education and self-management.
Asthma, chronic obstructive pulmonary disease (COPD) Definition and comparability
and congestive heart failure (CHF) are three widely prevalent
long-term conditions. Both asthma and COPD limit the The indicators are defined as the number of hospital
ability to breathe: asthma symptoms are usually intermittent admissions with a primary diagnosis of asthma, COPD
and reversible with treatment, whilst COPD is a progressive and CHF among people aged 15 years and over per
disease that almost exclusively affects current or prior 100 000 population. Rates were age-sex standardised
smokers. Asthma affects an estimated 235 million people to the 2010 OECD population aged 15 and over.
worldwide (WHO, 2013). More than 3 million people died of Disease prevalence may explain some, not all, varia-
COPD in 2012, which is equal to 6% of all deaths globally tions in cross-country rates. Differences in coding
that year (WHO, 2015). CHF is a serious medical condition practices among countries and the definition of an
in which the heart is unable to pump enough blood to meet admission may also affect the comparability of data.
the bodys needs. CHF is often caused by hypertension, For example, while the transfer of patients from one
diabetes or coronary heart disease. Heart failure is estimated hospital to another is required to be excluded from
to affect over 26 million people worldwide resulting in more the calculations to avoid double counting, this cannot
than 1 million hospitalisations annually in both the United be fully complied with by some countries. There is
States and Europe. also a risk that countries that do not have the capacity
Common to all three conditions is the fact that the evidence to track patients through the system do not identify
base for effective treatment is well established and much of all relevant admissions due to changes in diagnosis
it can be delivered at a primary care level. A high-performing coding on transfer between hospitals. The impact of
primary care system can reduce acute deterioration in people excluding admissions where death occurred has been
living with asthma, COPD or CHF and prevent their admis- investigated, given these admissions are less likely to
sion to hospital. be avoidable. The results reveal that while the impact
Figure 8.1 shows hospital admission rates for asthma and on the indicator rate varies across conditions (e.g. on
COPD together, given the physiological relationship average, reduced asthma rates by less than 1%
between the two conditions. Admission rates for asthma whereas for CHF it was nearly 9%), the changes in the
vary 11-fold across countries with Italy, Switzerland and variation of rates across countries for each condition
Mexico reporting the lowest rates and Korea, United States was minimal.
and the Slovak Republic reporting rates over twice the
OECD average. International variation in admissions for
COPD is 17-fold across OECD countries, with Japan and Italy
References
reporting the lowest rates and Hungary and Ireland the
highest rates. Combined, there is a lower 8-fold variation OECD (2012), OECD Reviews of Health Care Quality: Israel 2012:
across countries for the two respiratory conditions. Hospi- Raising Standards, OECD Publishing, Paris,
tal admission rates for CHF vary 7-fold, as shown in http://dx.doi.org/10.1787/9789264029941-en.
Figure 8.2. Mexico, United Kingdom and Korea have the WHO (2015), Chronic Obstructive Pulmonary Disease
lowest rates, while the Slovak Republic, Hungary and (COPD), Fact Sheet No. 315,
Poland report rates at least 1.8 times the OECD average. www.who.int/mediacentre/factsheets/fs315/en/.
The majority of countries report a reduction in admission WHO (2013), Asthma, Fact Sheet No. 307,
rates for CHF over recent years. This may represent an www.who.int/mediacentre/factsheets/fs307/en/.

8. QUALITY OF CARE
8.1. Asthma and COPD hospital admission in adults, 2013 (or nearest year)
COPD Asthma
Age-sex standardised rates per 100 000 population
600
500
400
300
200
100
0
i t z gal
M d
No m
De tes
nd
n
Po l y
o
il e
ec ds
d
O E in
Be 32
ay
Ca ia
Po l
Sl c e
m a
Fi g
th nd
S w p.
Ic n
da
Un Ge and
K i any
ak m
Un K .
a
Es rk
Au ia
A ria
Ze a
Hu n d
Ir e y
e
w a li
xe ni
i t e or e
ar
n
an
e
pa
ic
ur
ra
n
It a
tv
a
Re
Re
iu
Sl gdo
a
rw
an
Ch
C z lan
ed
na
la
la
N e nl a
a
CD
st
Sw r tu
to
Lu ove
Sp
a
ex
nm
ng
bo
Ne us tr
i te rm
el
l
Ja
al
La
Is
lg
St
er
Fr
h
er
d
ov
d
Note: Three-year average for Iceland and Luxembourg.
1 2 http://dx.doi.org/10.1787/888933281105
8.2. Congestive heart failure hospital admission in adults, 2008 and 2013 (or nearest years)
2008 2013
600
500
400
300
200
100
0
nd
Ic l
o
a
il e
m
De n
it z rk
No d
Be a
Po m
Ca y
th nd
s
ain
el
S en
Ge tes
d
Au c e
O E li a
30
Fi y
Au d
Sw ia
y
a
ec ny
o v e p.
H u p.
nd
ga
a
nd
re
i te eni
l
ar
n
ic
pa
an
an
ra
It a
Re
iu
do
Sw ma
rw
Ch
an
ra
na
ed
a
la
la
Ne ela
la
CD
st
r tu
N e Sp
R
ex
Ko
ng
rm
la
Ja
nl
al
Is
lg
Un lov
St
Ir e
Po
er
st
ng
Fr
n
h
ak
M
er
Ze
d
Ki
Cz
d
Sl
i te
Un
Note: Three-year average for Iceland.

1 2 http://dx.doi.org/10.1787/888933281105

8. QUALITY OF CARE
Diabetes care
Diabetes is a chronic disease that occurs when the bodys reporting rates lower than 3 per 100 000 general population
ability to regulate excessive glucose levels in the blood is and Israel, Slovenia and Portugal reporting rates above 10.
lost. Across the OECD countries, diabetes is a leading cause Rates based on the estimated diabetic population are pre-
of cardiovascular disease, blindness, kidney failure, and sented in the right panel. The rates based on the diabetic
lower limb amputation. Globally it is estimated that over population are on average 9-fold higher than for the gen-
380 million people had diabetes in 2014 and by 2035 it is eral population and display differences in the ranking of
projected that close to 600 million people will have the con- countries, providing an indication that differences in dis-
dition. Diabetes caused close to 5 million deaths in 2014 ease prevalence across countries may explain some, but
(IDF, 2014). Many countries have established comprehen- not all, cross-country variation.
sive approaches to diabetes care, but there are indications
that more can be done to prevent the disease (OECD, 2014).
Cholesterol-lowering drugs and medications to reduce
blood pressure are recommended in most national guide- Definition and comparability
lines for the care of diabetes patients (see indicator Pre-
scribing in primary care in Chapter 8) The indicator for diabetes hospital admission is
Poor control of the level of glucose in the blood over the defined as the number of hospital admissions with a
short term can lead to vomiting, dehydration and even primary diagnosis of diabetes among people aged
cause coma, whereas sustained high levels of blood glucose 15 years and over per 100 000 population. The indica-
over a number of years can result in serious diseases with tor for major lower extremity amputation in adults
ongoing consequences for a persons health and wellbeing. with diabetes is defined as the number of discharges
For example, diabetes can cause nerve damage and poor of people aged 15 years and over per 100 000 popula-
blood circulation over time. These problems make the feet tion, for the general population and the estimated dia-
vulnerable to skin ulcers that can deteriorate quickly and betic population. Rates for both indicators were age-
be difficult to treat. An ulcer that does not heal can cause sex standardised to the 2010 OECD population aged 15
severe damage to tissues and bone over time and can even- and over.
tually require amputation of a toe, foot or part of a leg. Differences in data definition and coding practices
Proper diabetes management and careful foot care can pre- between countries may affect the comparability of
vent foot ulcers. Ongoing management of diabetes usually data. For example, coding of diabetes as a principal
involves a considerable amount of self-care, and therefore, diagnosis versus a secondary diagnosis varies across
advice and education are central to the primary care of peo- countries. This is more pronounced for diabetes than
ple with diabetes. Effective control of blood glucose levels other conditions, given that in many cases admission
through routine monitoring, dietary modification and regu- is for the secondary complications of diabetes rather
lar exercise can reduce the onset of serious complications than diabetes itself. Diabetes population estimates
and the need for hospitalisation. used to calculate amputation indicator rates were
Figure 8.3 shows the avoidable hospital admissions for dia- self-reported by countries. Subject to further data
betes. The international variation in the rates is nearly 8- development, the use of diabetes population estimates
fold, with Italy, Switzerland and Spain reporting the lowest to standardise the indicator rates will be considered in
rates and Austria, Korea and Mexico reporting rates at least the future.
two times that of the OECD average. Prevalence of diabetes
may explain some of the variation in diabetes admission
rates. A positive relationship can be demonstrated between
hospital admissions for the general population and diabe-
References
tes-related hospital admissions, providing some indication
that overall access to hospital care can also play a role in International Diabetes Federation (2014), IDF Diabetes Atlas
explaining the level of hospital care among the diabetic Sixth Edition Update 2014, https://www.idf.org/sites/default/
population (OECD, 2015). files/EN_6E_Atlas_Full_0.pdf.
Hospital admissions for major lower extremity amputation OECD (2015), Cardiovascular Disease and Diabetes: Policies for
(i.e. surgical removal of lower limb, including leg or foot) Better Health and Quality of Care, OECD Health Policy Stud-
reflect the long-term quality of diabetes care. Figure 8.4 ies, OECD Publishing, Paris,
shows the rates of major lower extremity amputation in http://dx.doi.org/10.1787/9789264233010-en.
adults with diabetes. In the left panel the rates based on OECD (2014), OECD Reviews of Health Care Quality: Czech
the general population are presented. The international Republic 2014: Raising Standards, OECD Publishing, Paris,
variation in rates is over 14-fold, with Korea and Italy http://dx.doi.org/10.1787/9789264208605-en.

8. QUALITY OF CARE
Diabetes care
8.3. Diabetes hospital admission in adults, 2008 and 2013 (or nearest years)
2008 2013
450
400
350
300
250
200
150
100
50
0
l
nd
Ir e a
Au nd
ng d
No s
Ca l
O E li a
31
il e
M a
m n
o
ly
a in
Po ay
Hu d a
Sw r y
Sl en
De ni a
Fi k
d
Be g
Ge tes
m
Ze e
St .
ov any
p.
Au d
ria
er m
ga
i t e Rep
nd
re
ar
w nc
an
an
C z alan
n
a
ic
ur
ra
tv
It a
Re
iu
do
a
rw
Ch
ra
ed
na
p
la
la
la
CD
st
r tu
e
Sp
ex
Ko
nm
ng
bo
rm
Ne Fr a
la
el
nl
Lu Ja
La
Is
lg
ov
Po
er
st
Un h
ak
Ic
it z
ec
d
Ki
th
xe
Sw
Ne
d
Sl
i te
Un

1 2 http://dx.doi.org/10.1787/888933281111
8.4. Major lower extremity amputation in adults with diabetes, 2013 (or nearest year)
2.4 Korea 12.7
2.7 Italy 29.7
3.1 Switzerland
3.1 United Kingdom
3.2 Ireland
3.5 Luxembourg 26.7
3.5 Iceland
4.1 Sweden
4.5 Australia
4.7 Netherlands 42.9
4.8 Belgium
5.7 Norway
5.9 New Zealand 55.8
6.4 OECD 21/11 57.4
6.7 Spain 54.0
7.4 Canada 49.6
7.5 France
8.5 Denmark
9.2 Germany 57.0
11.9 Portugal 92.8
15.3 Slovenia 122.7
15.9 Israel 87.4
20 15 10 5 0 0 50 100 150
Age-sex standardised rates per 100 000 population Age-sex standardised rates per 100 000 people with diabetes

1 2 http://dx.doi.org/10.1787/888933281111

8. QUALITY OF CARE
Beyond consumption and expenditure information (see of developing such adverse effects compared with placebo
Chapter 10), prescribing can be used as an indicator of (Sithamparanathan et al., 2012). Figures 8.8 and 8.9 indicate
health care quality. Antibiotics, for example, should be pre- th at, ac ro s s the O EC D, o n ave rag e aroun d 2 9 p e r
scribed only where there is an evidence-based need, to 1 000 elderly patients receive long-term prescriptions for
reduce the risk of resistant strains. Likewise, quinolones benzodiazepines and related drugs ( 365 defined daily
and cephalosporins are considered second-line antibiotics doses in one year), and 62 per 1 000 have received at least
in most prescribing guidelines. Their use should be one prescription for a long-acting benzodiazepine or
restricted to ensure availability of effective second-line related drugs within the year.
therapy should first-line antibiotics fail. Total volume of To reduce the potentially harmful overuse and misuse of
antibiotics prescribed, and second-line as a proportion of medicines, diagnostic tests and procedures, the Choosing
total volume, have been validated as markers of quality in Wisely campaign was launched in 2012. Increasingly
the primary care setting. In May 2015, the World Health international, the campaign comprises evidence-based
Assembly endorsed a global action plan to tackle antimicro- information for clinicians and patients on when medica-
bial resistance (http://who.int/drugresistance/global_action_plan), tions and procedures may be inappropriate. Appropriate
which is also reflected in several national strategies. use of antibiotics and benzodiazepines is addressed
Figure 8.5 shows volume of all antibiotics prescribed in pri- (www.choosingwisely.org).
mary care, with volumes of second- line antibiotics embed-
ded within the total amount. Total volumes vary more than
four-fold across countries, with Chile, the Netherlands and
Estonia reporting the lowest volumes, and Turkey and Definition and comparability
Greece reporting volumes much higher than the OECD aver-
Defined daily dose (DDD) is the assumed average
age. Volumes of second-line antibiotics vary almost 16-fold
maintenance dose per day for a drug used for its main
across countries. The Nordic countries, the Netherlands and
indication in adults. DDDs are assigned to each active
the United Kingdom report the lowest volumes of these
ingredient in a given therapeutic class by interna-
antibiotics, and Korea, the Slovak Republic and Greece the
tional expert consensus. For instance, the DDD for
highest. Variation is likely to be explained, on the supply side,
oral aspirin equals 3 grams, which is the assumed
by differences in the regulation, guidelines and incentives
maintenance daily dose to treat pain in adults. DDDs
that govern primary care prescribers and, on the demand
do not necessarily reflect the average daily dose actually
side, by cultural differences in attitudes and expectations
used in a given country. DDDs can be aggregated
regarding the natural history and optimal treatment of
within and across therapeutic classes of the Anatomic
infective illness.
Therapeutic Classification (ATC). For more detail, see
In conjunction with additional indicators of the quality of www.whocc.no/atcddd.
primary care for diabetes (see Diabetes care), Health at a
In Figure 8.5, data for Chile include over the counter
Glance is for the first time reporting two indicators related
(OTC) drugs. Data for Canada, Israel and Luxembourg
to the quality of prescribing in primary care for diabetic
exclude drugs prescribed in hospitals, non-reimbursed
patients. In diabetic individuals with hypertension, angio-
drugs and OTC drugs. Data for Iceland refer to all
tensin-converting enzyme inhibitors (ACE-I) or angiotensin
sectors, not just primary care. Data for Portugal include
receptor blockers (ARB) are recommended in most national
OTC and non-reimbursed drugs. Data for Australia
guidelines as first-line medications to reduce blood pres-
include non-reimbursed drugs. Data for Turkey refer
sure, since they are most effective at reducing the risk of
to outpatient health care.
cardiovascular disease and renal disease. Figures 8.6
and 8.7 suggest there is wide variability across countries in Denominators comprise the population held in the
prescribing practices for diabetes patients, with 27% of national prescribing database, rather than the general
diabetic patients in the Slovak Republic given prescriptions population (with the exception of Belgian data on
for cholesterol-lowering medication, compared with 81% in benzodiazepines, which comes from a national
New Zealand. There is greater consistency in the proportion health survey).
of diabetic patients on antihypertensive agents with at
least one prescription for ACE-I or ARB, with the exception
of the Slovak Republic.
Benzodiazepines are often prescribed for elderly patients References
for anxiety and sleep disorders, despite the risk of adverse Sithamparanathan, K., A. Sadera and L. Leung (2012),
side effects such as fatigue, dizziness and confusion. A Adverse Effects of Benzodiazepine Use in Elderly People: A
meta-analysis suggests that the use of benzodiazepines in Meta-analysis, Asian Journal of Gerontology & Geriatrics,
elderly people is associated with more than double the risk Vol. 7, No. 2, pp. 107-111.

8. QUALITY OF CARE
8.5. Overall volume of antibiotics prescribed, 2013 (or nearest year)
All 2nd line (where reported)

DDS per 1 000 population, per day
45
40
35
30
25
20
15
10
5
0
Au gal
nd
ey
P o d
il e
rm l
Au a
Po a
ov nd
Ir e .
xe a in
Be y
ce
g
Gr e
a
ia
y
en
ia
No y
ay
De r ia
Ca a
Fi k
Li nd
U n O p.
Ki 29
Ic m
e
p
nd
an
ni
C z nad
li
c
ni
re
l
ar
ar
ur
ra
en
It a
tv
Re
Re
iu
do
an
rk
rw
an
ee
Ch
ra
ed
la
la
a
D
st
r tu
Lu Sp
to
ua
Ko
nm
ng
bo
la
nl
La
Is
lg
ov
i te EC
Tu
st
el
ng
Fr
Sw
Es
ak
h
th
er
Hu
m
Ge
ec
Sl
th
Ne
Sl
1. Data refer to all sectors (not only primary care).
Source: European Centre for Disease Prevention and OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en.
1 2 http://dx.doi.org/10.1787/888933281125
8.6. People with diabetes with a prescription of 8.7. People with diabetes with a prescription
cholesterol lowering medication in the past year, 2013 of recommended antihypertensive medication in the past
(or nearest year) year, 2013 (or nearest year)
% of diabetic patients % of diabetic patients
100 100
80 80
60 60
40 40
20 20
0 0
l
d
nd
a
k
en
ay
12
ia
p.
11
m
d
nd
en
p.
ia
ay
ga
ga
nd
re
ar
an
an
nd
re
ar
an
en
Re
iu
en
Re
iu
rw
ed
CD
rw
la
ed
CD
la
r tu
Ko
nm
r tu
Ko
nm
la
al
nl
lg
la
al
lg
ov
Ir e
ov
Ir e
Sw
No
Sw
ak
No
ak
Ze
er
OE
Fi
Be
Ze
er
OE
Be
Po
Po
De
Sl
De
Sl
ov
th
ov
th
w
w
Ne
Ne
Sl
Sl
Ne
Ne
8.8. Elderly people prescribed long-term benzodiazepines 8.9. Elderly people prescribed long-acting
or related drugs, 2013 (or nearest year) benzodiazepines or related drugs, 2013 (or nearest year)
Per 1 000 persons aged 65 years and over Per 1 000 persons aged 65 years and over
70 250
60
200
50
40 150
30 100
20
50
10
0 0
No al
a
l
Ca 4
ov eni a
Ir e .
Po d
De da
nd
O E ay
Be rk
m
S el
w den
th nd
Fi s
el
en
s
ay
13
da
d
ia
p.
d
ga
p
nd
re
nd
re
ar
1
n
an
an
an
ra
ra
en
Re
Re
iu
a
rw
rw
na
ed
na
CD
CD
la
la
Ne eala
r tu
r tu
Ko
nm
Ko
nm
Ne we
la
al
la
nl
nl
Is
Is
lg
ov
ov
Ir e
Sw
Ca
No
ak
ak
er
Ze
er
OE
Fi
Po
De
Sl
Sl
Z
ov
th
w
Ne
Sl
Sl
Ne

8. QUALITY OF CARE
Mortality due to coronary heart disease has declined sub- ventions and high-quality specialised health facilities such
stantially since the 1970s (see indicator Mortality from as percutaneous catheter intervention-capable centres
cardiovascular diseases in Chapter 3). Advances in the have helped to reduce 30-day case-fatality rates (OECD,
prevention such as smoking (see indicator Tobacco con- 2015a). For example, Korea had higher case-fatality rates
sumption among adults in Chapter 4) and treatment of for AMI but in 2006 it has implemented a Comprehensive
cardiovascular diseases outpaced those of many other dis- Plan for CVD, encompassing prevention, primary care and
eases (OECD, 2015a). acute CVD care (OECD, 2012). Under the Plan, specialised
A good indicator of acute care quality is the 30-day AMI services were enhanced through a creation of regional
case-fatality rate. This measure reflects the processes of cardio and cerebrovascular centres throughout the country,
care, such as timely transport of patients and effective and average waiting time from emergency room arrival to
medical interventions. The indicator is influenced by not initiation of catheterisation fell from 72.3 in 2010 to
only the quality of care provided in hospitals but also dif- 65.8 minutes in 2011, leading to a reduction in case-fatality
ferences in hospital transfers, average length of stay and (OECD, 2015a).
AMI severity.
Figure 8.10 shows the case-fatality rates within 30 days of
admission for AMI when the death occurs in the same hos- Definition and comparability
pital as the initial AMI admission. The lowest rate is found
in Australia at 4.1% and the highest rate is in Mexico at 28.2%, The case-fatality rate measures the percentage of
suggesting AMI patients do not always receive recommended people aged 45 and over who die within 30 days
care. In Mexico, the quality of pre-hospital emergency following admission to hospital for a specific acute
medical services is reportedly poor (Peralta, 2006), and the condition. Rates based on admission data refer to the
high rates of uncontrolled diabetes may also be a contribut- death occurred in the same hospital as the initial
ing factor in explaining the high AMI case-fatality rates (see admission. Admissions resulting in a transfer were
indicator Diabetes care in Chapter 8) as patients with excluded for all countries except Australia, Belgium,
diabetes have worse outcomes after AMI compared to Denmark, Hungary, Ireland, Israel, Japan, Luxembourg,
those without diabetes, particularly if the diabetes is poorly Mexico, Netherlands, Slovak Republic and Sweden.
controlled. In Japan, people are less likely to die of heart This exclusion generally increases the rate compared
disease overall, but are more likely to die once admitted with those countries which do not exclude these
into hospital for AMI compared to other OECD countries. transfers. Rates based on patient data refer to the
One possible explanation is that the severity of patients death occurred in the same hospital, a different hos-
admitted to hospital with AMI may be more advanced pital, or out of hospital.
among a smaller group of people across the population, but Rates are age-sex standardised to the 2010 OECD
could also reflect underlying differences in emergency care, population aged 45+ admitted to hospital for a specific
diagnosis and treatment patterns (OECD, 2015b). acute condition such as AMI and ischemic stroke.
Figure 8.11 shows 30-day case fatality rates where fatalities
are recorded regardless of where they occur. This is a more
robust indicator because it records deaths more widely
than the same-hospital indicator, but it requires a unique References
patient identifier and linked data which is not available in
OECD (2015a), Cardiovascular Disease and Diabetes: Policies for
all countries. The AMI case-fatality rate ranges from 7.1% in
Better Health and Quality of Care, OECD Health Policy Stud-
Canada to 18.8% in Hungary and 19.1% in Latvia.
ies, OECD Publishing, Paris,
Case-fatality rates for AMI have decreased substantially http://dx.doi.org/10.1787/9789264233010-en.
between 2003 and 2013 (Figures 8.10 and 8.11). Across the
OECD (2015b), OECD Reviews of Health Care Quality: Japan
OECD, case fatalities fell from 11.2% to 8.0% when consider-
2015: Raising Standards, OECD Publishing, Paris,
ing same hospital deaths and from 14.3% to 9.5% when
http://dx.doi.org/10.1787/9789264225817-en.
considering deaths occurred in and out of hospital. The
rate of decline was particularly striking in the Slovak OECD (2012), OECD Reviews of Health Care Quality: Korea 2012:
Republic, the Netherlands and Australia for the first indicator Raising Standards, OECD Publishing, Paris,
and in Finland and Poland for the second indicator, with http://dx.doi.org/10.1787/9789264173446-en.
more than 6% annual average reduction per year compared Peralta, L.M.P. (2006), The Prehospital Emergency Care Sys-
to an OECD average of respectively 3 and 4 %. Better access tem in Mexico City: A Systems Performance Evaluation,
to high-quality acute care for heart attack, including timely P re h o s p i t a l a n d D i s a s t e r M e d i c i n e , Vo l . 2 1 , N o. 2 ,
transportation of patients, evidence-based medical inter- pp. 104-111.

8. QUALITY OF CARE
8.10. Thirty-day mortality after admission to hospital for AMI based on admission data, 2003 to 2013 (or nearest years)
2003 2008 2013
Age-sex standardised rate per 100 admissions of adults aged 45 years and over
30
25
20
15
10
Au al
De a tes
e l
bo d
nd
O E in
Ge e a
o
Sw lia
Sl nd
ly

Ir e k
Ze d
Ca d
a
Is .
ay
32
Po ny
Es ia
Ja a
Hu an
y
il e
M ia
g
ak e
Be p.
Po
ia
i te er la
er
p
K i nds
en
Un t h um
m
nd
C z nad
ni
ov n c
xe l a n
w an
an
g
en
tv
i te t a
a
Re
ar
ar
ic
ur
rw
Ch
a
ra
Re
la
la
S w gdo
CD
st
r tu
to
ra
Sp
ed
p
Ko
la
rm
Sl Fr a
I
N e F inl
al
La
ex
N e l gi
ng
nm
ov
St
L u Ic e
No
st
it z
ec
n
m
Au
d
Un
Note: 95% confidence intervals represented by H. Three-year average for Iceland and Luxembourg.
1. Admissions resulting in a transfer are included.
1 2 http://dx.doi.org/10.1787/888933281135
8.11. Thirty-day mortality after admission to hospital for AMI based on patient data, 2003 to 2013 (or nearest years)
2003 2008 2013
25
20
15
10
0
nd
l
a
da
en
el
ly
ay
nd
a in
ia
20
p.
ia
ga
nd
re
ni
ar
ar
an
an
ur
ra
en
It a
tv
Re
do
rw
ed
na
la
la
CD
r tu
to
Sp
Ko
nm
ng
bo
la
nl
al
La
Is
ov
Po
er
ng
Sw
Ca
No
Es
h
er
Ze
Fi
OE
Hu
m
Po
it z
De
ec
Sl
Ki
th
xe
w
Cz
Sw
Ne
d
Ne
Lu
i te
Un
Note: 95% confidence intervals represented by H. Three-year average for Luxembourg.

1 2 http://dx.doi.org/10.1787/888933281135

8. QUALITY OF CARE
Stroke and other cerebrovascular diseases accounted for Between 2003 and 2013, case-fatality rates for ischemic
around 7% of all deaths in OECD countries in 2013. Isch- stroke have decreased substantially (Figures 8.12 and 8.13).
emic stroke represented around 85% of all cerebrovascular Across the OECD, case fatalities fell from 10.2% to 8.4%
disease cases. It occurs when the blood supply to a part of when considering same hospital rates and from 12.7% to
the brain is interrupted, leading to a necrosis (i.e. the cells 10.1% when considering in- and out-of-hospital rates. The
that die) of the affected part. Treatment for ischemic stroke United Kingdom and the Netherlands for the first indicator
has advanced dramatically over the last decade. Clinical and the United Kingdom, Estonia and Finland for the sec-
trials have demonstrated clear benefits of thrombolytic ond indicator were able to reduce their rates by an average
treatment for ischemic stroke as well as receiving care in annual reduction of more than 6% compared to an OECD
dedicated stroke units to facilitate timely and aggressive average of respectively 2 and 2.5%. Better access to high-
diagnosis and therapy for stroke victims (Hacke et al., 1995; quality stroke care, including timely transportation of
Seenan et al., 2007). patients, evidence-based medical interventions and high-
Figure 8.12 shows the case-fatality rates within 30 days of quality specialised facilities such as stroke units have
admission for ischemic stroke when the death occurred in helped to reduce 30-day case-fatality rates (OECD, 2015).
the same hospital as the initial stroke admission. Despite the progress seen so far, there is still room to
Figure 8.13 shows the case-fatality rate where deaths are improve implementation of best practice acute care for car-
recorded regardless of where they occurred. This indicator diovascular diseases including stroke across countries. To
is more robust because it captures fatalities more compre- shorten acute care treatment time, targeted strategies can
hensively. Although more countries can report the more be highly effective. But to encourage the use of evidence-
partial same-hospital measure, an increasing number of based advanced technologies in acute care, wider
countries are investing in their data infrastructure and are approaches are needed. Adequate funding and trained pro-
able to provide more comprehensive measures. fessionals should be made available, and health care deliv-
Across OECD countries 8.4% of patients in 2013 died within ery systems should be adjusted to enable easy access
30 days in the same hospital in which the initial admission (OECD, 2015).
for ischemic stroke occurred (Figure 8.12). The case-fatality
rates were highest in Mexico (19.5%) and Latvia (18.4%).
Rates were less than 5% in Japan, Korea and the United Definition and comparability
States. With the exception of Japan and Korea, countries
that achieve better results for ischemic stroke also tend to Case-fatality rates are defined in indicator Mortality
report good case-fatality rates for acute myocardial following acute myocardial infarction in Chapter 8.
infarction (AMI). This suggests that certain aspects of acute
care may be influencing outcomes for both stroke and AMI
patients. By contrast, Japan reports the lowest rates for
ischemic stroke but high case-fatality rates for AMI. This References
somewhat paradoxical result requires further investigation Hacke, W. et al. (1995), Intravenous Thrombolysis with
but may be associated with the severity of disease in the Recombinant Tissue Plasminogen Activator for Acute
country that is not captured in the data (see indicator Mor- Hemispheric Stroke. The European Co-operative Acute
tality following acute myocardial infarction in Chapter 8 Stroke Study (ECASS), Journal of the American Medical
for more details). Association, Vol. 274, No. 13, pp. 1017-1025.
Across the 19 countries that reported in- and out-of-hospi- OECD (2015), Cardiovascular Disease and Diabetes: Policies for
tal case-fatality rates, 10.1% of patients died within 30-days Better Health and Quality of Care, OECD Publishing, Paris,
of being admitted to hospital for stroke (Figure 8.13). This http://dx.doi.org/10.1787/9789264233010-en.
figure is higher than the same-hospital based indicator Seenan, P., M. Long and P. Langhorne (2007), Stroke Units
because it captures deaths that occur not just in the same in Their Natural Habitat: Systematic Review of Observa-
hospital but also in other hospitals and out-of-hospital. tional Studies, Stroke, Vol. 38, pp. 1886-1892.

8. QUALITY OF CARE
8.12. Thirty-day mortality after admission to hospital for ischemic stroke based on admission data, 2003 to 2013
(or nearest years)
2003 2008 2013
25
20
15
10
l
er nd
M ia
Fr
31
De il e
Ire y
o
n
a
es
U n xe m a r k
ng
Po da
p.
Sl ni a
ia
Au om
Be lia
H u e p.

a in
No d
ay
e l
Au y
Ge tr ia
S w ny
n
ce
w and
OE nd
ov g a
s
C z ium
K i ur g
nd
re
l
an
en
tv
It a
nd
ar
S w ede
pa
ic
at
rw
an
Ch
na
a
Re
ra
a
CD
ra
Sl r tu
to
Sp
R
Ko
la
s
rm
N e z er l
N e Ic el
nl
al
La
ex
d
ng
L u nm
i te bo
ov
St
la
Ja
lg
Is
Ca
Es
st
Ze
Fi
ak
ec
d
it
th
i te
d
Un
Note: 95% confidence intervals represented by H. Three-year average for Iceland and Luxembourg.
1. Admissions resulting in a transfer are included.
1 2 http://dx.doi.org/10.1787/888933281146
8.13. Thirty-day mortality after admission to hospital for ischemic stroke based on patient data, 2003 to 2013
(or nearest years)
2003 2008 2013
30
25
20
15
10
0
l
d
ia
a
da
19
l
nd
y
ay
ly
ds
en
ain
ia
d
om
ga
ae
ep
ni
re
ar
ar
lan
lan
ur
en
tv
It a
rw
ed
na
lan
CD
ala
Isr
r tu
to
Sp
hR
Ko
nm
gd
ng
bo
La
ov
F in
er
Sw
Ca
No
Es
er
Ze
OE
in
Hu
m
Po
it z
De
ec
Sl
th
dK
xe
w
Cz
Sw
Ne
Ne
Lu
i te
Un
Note: 95% confidence intervals represented by H. Three-year average for Luxembourg.

1 2 http://dx.doi.org/10.1787/888933281146

8. QUALITY OF CARE
The main risk factors for hip fracture are associated with country. In Canada, the percentage of patients operated on
ageing an increased risk of falling and loss of skeletal within the two day benchmark increased from 87% in 2008
strength from osteoporosis. With increasing life expec- to 92% in 2013, but there is considerable variation in this
tancy across most OECD countries, it is anticipated that hip indicator between provinces and hospitals (CIHI, 2015).
fracture will become a more significant public health issue Portugal saw a decline of hip fracture repair within two
in coming years. days of admission from 57% in 2008 to 45% in 2013.
In most instances following hip fracture, surgical interven- Time to surgery for hip fracture patients is influenced by
tion is required to repair or replace the hip joint. There is many factors, including hospitals surgical theatre capacity,
general consensus that early surgical intervention maxi- flow and access. Improvement in timely surgery for patients
mises patient outcomes and minimises the risk of compli- with a particular diagnosis or injury (e.g. hip fracture) may be
cations. General agreement is that surgery should occur achieved at the expense of timeliness in others (e.g. hip or
within two days (48 hours) of hospitalisation. Guidelines knee replacements).
in some countries call for even earlier intervention. For
example, the National Institute for Health and Care
Excellence (NICE) clinical guidelines recommend hip fracture
surgery to be performed on the day of hospital admission Definition and comparability
or the next day (National Institute for Health and Care
Excellence, 2014). This indicator is defined as the proportion of patients
This is the first time Health at a Glance is reporting on the aged 65 years and over admitted to hospital in a
time taken to initiate hip fracture surgery after hospital specified year with a diagnosis of upper femur
admission. Timely surgery can be considered an indicator fracture, who had surgery initiated within two
of the quality of acute care received by patients with hip calendar days of their admission to hospital. Data are
fracture. also provided for the proportion of those patients who
had surgery within one day of their admission to
In 2013, on average across the OECD over 80% of patients
hospital, and for patients who had surgery on the
admitted for hip fracture underwent surgery within two
same day as their hospital admission. While the
days (Figure 8.14). In Denmark, Iceland and the Netherlands,
capacity to capture time of admission and surgery in
the proportion was greater than 95%. Countries with the
hospital administrative data varies across countries,
lowest proportion of patients operated on within two days of
most countries are able to distinguish between
admission were Spain (43%), Italy (45%) and Portugal (45%).
patients who stay overnight and have surgery within
Many patients were treated sooner than two days following
24 hours from patients who have surgery on the day of
admission. In the Netherlands and the Czech Republic, for
admission. Some countries supplied results for surgery
example, over 40% of patients admitted for hip fracture
within two calendar days only.
underwent surgery on the day of admission.
Figure 8.15 shows the proportion of hip-fracture repairs
occurring within two days of admission in OECD countries
between 2003 and 2013. The OECD average increased from
76% to 81% over that time. The greatest improvement was
References
observed in Italy, where the proportion increased from 28% CIHI Canadian Institute for Health Information (2015),
in 2008 to 45% in 2013, and in Israel, where it increased Wait Times for Priority Procedures in Canada, Ottawa.
from 70% in 2003 to 85% in 2013. A policy of comparative National Institute for Health and Care Excellence (2014),
public reporting of hospital indicators, including time to sur- Hip Fracture: The Management of Hip Fracture in
gery following hip fracture, implemented by Italian authori- Adults, NICE Clinical Guideline No. 124, issued June 2011,
ties may partly explain the improvement observed in that last modified March 2014.

8. QUALITY OF CARE
8.14. Hip fracture surgery initiation after admission to hospital, 2013 (or nearest year)
2 days Next day Same day
% of patients aged 65 years and over
100
90
80
70
60
50
40
30
20
10
0
d
nd
l
y
el
nd
k
en
da
ay
p.
ria
ia
ly
22
ia
a in
ga
nd
an
ni
ar
ar
an
an
an
ra
en
tv
It a
Re
iu
do
rw
ed
na
la
la
st
CD
to
r tu
Sp
nm
ng
rm
la
el
nl
al
La
Is
lg
ov
Ir e
er
ng
Au
Sw
Ca
No
Es
h
Ic
er
Ze
Fi
OE
Be
Hu
Po
it z
De
Ge
ec
Sl
Ki
th
w
Cz
Sw
Ne
Ne
i te
Un

1 2 http://dx.doi.org/10.1787/888933281152
8.15. Hip fracture surgery initiation after admission to hospital, 2003 to 2013 (or nearest years)
2003 2008 2013
% of patients aged 65 years and over being operated within two days
100
90
80
70
60
50
40
30
20
10
0
l
d
nd
nd
el
k
en
da
ay
ria
d
m
p.
22
ia
ly
a in
ia
ga
nd
an
ni
ar
ar
an
an
an
ra
en
tv
It a
Re
iu
do
rw
ed
na
la
la
CD
st
to
r tu
Sp
nm
ng
rm
la
el
nl
al
La
Is
lg
ov
Ir e
er
ng
Au
Sw
Ca
No
Es
h
Ic
er
Ze
Fi
OE
Be
Hu
Po
it z
De
Ge
ec
Sl
Ki
th
w
Cz
Sw
Ne
Ne
i te
Un

1 2 http://dx.doi.org/10.1787/888933281152

8. QUALITY OF CARE
Patient safety remains one of the most prominent issues in

health policy and public debate. High rates of error during Definition and comparability
the delivery of medical care have been demonstrated
repeatedly, including the landmark report by the Institute Surgical complications are defined as the number of
of Medicine which estimated that more people die from discharges with ICD codes for complication in any
medical errors than from traffic injuries or breast cancer secondary diagnosis field for the surgical admission
(Kohn et al., 2000). Robust comparison of performance with and any diagnosis field for any subsequent related re-
peers is fundamental to securing improvement. Two types admission within 30 days, divided by the total number
of patient safety event can be distinguished for this purpose: of discharges for patients aged 15 and older. Contrary
never events, those events that should never occur, such as to the data presented in Health at a Glance 2013, the
failure to remove surgical foreign bodies at the end of a indicator rates have not been adjusted by the average
procedure; and adverse events, such as post-operative sepsis, number of secondary diagnoses, given a strong posi-
which can not be avoided in all cases given the high-risk tive correlation between the number of secondary
nature of some procedures, although increased incidence at diagnoses and indicator rates reported by countries
an aggregate level may indicate a systemic problem. was not evident in the most recent data.
Figure 8.16 shows rates for two related adverse events, A fundamental challenge in international comparison
pulmonary embolism (PE) or deep vein thrombosis (DVT) of patient safety indicators centres on the quality of
after hip or knee replacement surgery. These are high risk the underlying data. Variations in how countries
procedures most commonly associated with postoperative record diagnoses and procedures and define hospital
DVT and PE complications. PE and DVT cause unnecessary admissions can affect calculation of rates. For exam-
pain and in some cases death, but can be prevented by anti- ple differences in the use of the present on admission
coagulants and other measures before, during and after flag for diagnosis and disease (e.g. ICD-9-CM and ICD-
surgery. Figure 8.17 shows rates for another adverse event, 10-AM) and procedure classification systems are
sepsis after abdominal surgery. Abdominal surgery is also a known to affect data comparability. In some cases,
high risk procedure. Likewise, sepsis after surgery, which higher adverse event rates may signal more developed
may lead to organ failure and death, can in many cases be patient safety monitoring systems and a stronger
prevented by prophylactic antibiotics, sterile surgical tech- patient safety culture rather than worse care. Recent
niques and good postoperative care. Figure 8.18 illustrates a analysis of dispersion of postoperative PE or DVT rates
never event (events that should never occur), rates of foreign across hospitals within OECD countries revealed
body left in during procedure. The most common risk factors extremely large variations in reported rates, including
for this never event are emergencies, unplanned changes in implausibly high and low rates for hospitals in the
procedure, patient obesity and changes in the surgical same country even after risk adjustment. Hence, dif-
team; preventive measures include counting instruments, ferences in the national rates presented here are likely
methodical wound exploration and effective communica- to reflect differences in coding and recording practices
tion among the surgical team. both between and within countries and mask true
differences in care quality. There is a need for greater
The left panel of Figures 8.16, 8.17 and 8.18. shows the rate
consistency in reporting of patient safety events
of the three respective postoperative complications based
across countries and significant scope exists for
on the surgical admission, the hospital admission when
improved data quality within national patient safety
the surgery took place. The right panel of these figures
programs. Wider analysis of coding comparability will
shows rates based on the surgical admission and all subse-
inform future strategies for improvement.
quent re-admissions to hospital within 30 days, whether at
the same hospital or in another hospital. The use of a
unique patient identifier is required to calculate the indica-
tor rates in the right panel, which is currently not available
in some countries.
Caution is needed in interpreting the extent to which these
References
indicators accurately reflect international differences in
patient safety rather than differences in the way that coun- Kohn, L.T., J.M. Corrigan and M.S. Donaldson (Editors) (2000),
tries report, code and calculate rates of adverse events (see To Err Is Human: Building a Safer Health System, Institute of
Definition and comparability box). Medicine, National Academy Press, Washington, DC.

8. QUALITY OF CARE
8.16. Postoperative pulmonary embolism (PE) or deep vein thrombosis (DVT) in hip and knee surgeries, 2013
(or nearest year)
DVT PE
Surgical admission based All admission based
Poland
Finland
Spain
Italy
Norway
Portugal
Belgium
Sweden
United Kingdom
Switzerland
Ireland
United States
Israel
OECD9/9
Slovenia
Canada
Australia
New Zealand
France
2 500 2 000 1 500 1 000 500 0 0 500 1 000 1 500 2 000 2 500
Per 100 000 hospital discharges Per 100 000 hospital discharges
Note: Rates have not been adjusted by the average number of secondary diagnoses.
1. The average number of secondary diagnoses is < 1.5.
1 2 http://dx.doi.org/10.1787/888933281167
8.17. Postoperative sepsis in abdominal surgeries 2013 (or nearest year)

Poland
Finland
Korea
Canada
New Zealand
Italy
Sweden
Switzerland
Norway
Israel
Portugal
United Kingdom
OECD8/10
United States
Slovenia
Belgium
Spain
Australia
Ireland
3 500 3 000 2 500 2 000 1 500 1 000 500 0 0 500 1 000 1 500 2 000 2 500 3 000 3 500
1 2 http://dx.doi.org/10.1787/888933281167
8.18. Foreign body left in during procedure, 2013 (or nearest year)
Poland
Slovenia
Italy
Norway
Finland
Belgium
Portugal
New Zealand
Israel
Spain
OECD8/9
Ireland
Sweden
United Kingdom
United States
Canada
Australia
Switzerland
25 20 15 10 5 0 0 5 10 15 20 25
1 2 http://dx.doi.org/10.1787/888933281167

8. QUALITY OF CARE
Obstetric trauma
Patient safety during childbirth can be assessed by looking and Israel to more than 10% in the United States, Sweden,
at potentially avoidable tearing of the perineum during Denmark and Canada.
vaginal delivery. Tears that extend to the perineal muscles Rates of obstetric trauma after vaginal delivery without
and bowel wall require surgery. They are more likely to instrument (Figure 8.20) display equally large variation
occur in the case of first vaginal delivery, high baby birth across countries, ranging from 0.3% or less in Poland and
weight, labour induction, occiput posterior baby position, Slovenia to 2.8% or above in the United Kingdom, Sweden
prolonged second stage of labour and instrumental deliv- and Canada. There is a strong relationship between the two
ery. Possible complications include continued perineal pain indicators, with Poland and Slovenia reporting the lowest
and incontinence. rates and Sweden and Canada reporting amongst the high-
These types of tears are not possible to prevent in all cases, est rates for both indicators.
but can be reduced by employing appropriate labour man-
agement and high quality obstetric care. Hence, the propor-
tion of deliveries involving higher degree lacerations is a
useful indicator of the quality of obstetric care. Obstetric Definition and comparability
trauma indicators have been used by the US Joint Commis-
The two obstetric trauma indicators are defined as the
sion as well as by different international quality initiatives
proportion of instrument assisted/non-assisted vagi-
seeking to assess and improve obstetric care (AHRQ, 2006).
nal deliveries with third- and fourth-degree obstetric
Episiotomy is a surgical incision of the perineum per- trauma codes in any diagnosis and procedure field.
formed to widen the vaginal opening for the delivery of an Therefore, any differences in the definition of princi-
infant. Wide variation in the use of episiotomy during vag- pal and secondary diagnoses have no influence on the
inal deliveries currently exists across Europe, ranging from calculated rates. Several differences in data reporting
around 70% of births in Portugal and Poland in 2010 to less across countries may influence the calculated rates of
than 10% in Sweden, Denmark and Iceland (Euro-Peristat, obstetric patient safety indicators. These relate pri-
2013). The selective use of episiotomy to decrease severe marily to differences in coding practice and data
perineal lacerations during delivery is controversial, with sources. Some countries report the obstetric trauma
claims that there are currently inadequate data to properly rates based on administrative hospital data and oth-
evaluate safety and effectiveness considerations (Lappen ers based on obstetric register data. There is some evi-
and Gossett, 2010). dence that registries produce higher quality data and
Obstetric trauma indicators are considered relatively reli- report a greater number of obstetric trauma events
able and comparable across countries, particularly given compared to administrative datasets (Baghestan et al.,
they are less sensitive to variations in secondary diagnosis 2007).
coding practices across countries. Nevertheless, differ-
ences in the consistency with which obstetric units report
these complications may complicate international compar-
ison. Fear of litigation, for example, may cause under- References
reporting; conversely systems that rely on specially trained
AHRQ Agency for Health Research and Quality (2006),
administrative staff to identify and code adverse events
Patient Safety Indicators Overview: AHRQ Quality Indicators
from patients clinical records may produce more reliable
February 2006, AHRQ, Rockville, United States.
data.
Baghestan, E. et al. (2007), A Validation of the Diagnosis of
Obstetric trauma with instrument refers to deliveries using
Obstetric Sphincter Tears in Two Norwegian Databases,
forceps or vacuum extraction. As the risk of a perineal lac-
the Medical Birth Registry and the Patient Administra-
eration is significantly increased when instruments are
tion System, Acta Obstetricia et Gynecologica, Vol. 86,
used to assist the delivery, rates for this patient population
pp. 205-209.
are reported separately. The average rate of obstetric
trauma with instrument (6.0 per 100 instrument-assisted Euro-Peristat (2013), European Perinatal Health Report:
vaginal delivery) across 21 OECD countries in 2013 was Health and Care of Pregnant Women and Babies in
nearly 4 fold the rate without instrument (1.6 per Europe in 2010, INSERM, Paris.
100 vaginal delivery without instrument assistance). The Lappen, J.R. and D.R. Gossett (2010), Changes in Episiotomy
rate of obstetric trauma after vaginal delivery with instru- Practice: Evidence-based Medicine in Action, Expert
ment (Figure 8.19) shows high variation across countries. Review of Obstetrics and Gynecology, Vol. 5, No. 3,
Reported rates vary from below 2% in Poland, Slovenia, Italy pp. 301-309.

8. QUALITY OF CARE
Obstetric trauma
8.19. Obstetric trauma, vaginal delivery with instrument, 2013 (or nearest year)
Crude rates per 100 instrument-assisted vaginal deliveries
18
17.1
16
14 13.2 13.4
12
10.3
10
8.4
8.1
8 7.3
7.2 7.2
6.0
6
4.8 4.8
4.2
4 3.7
3.2
2.9
2.6
2.3
1.9
2 1.4
0.8 0.8
0
s1
d1
ay 1
en 1
k1
l
nd
nd
es
nd
el
ia
ly
ce
a in
m
21
li a
d
m
da
ga
an
an
ra
en
It a
nd
ar
iu
do
an
at
an
ra
na
la
la
la
CD
r tu
Sp
rw
ed
rm
al
Is
lg
nm
ov
St
Po
Ir e
er
st
la
nl
ng
Fr
Ca
Ze
Sw
OE
Be
No
Po
er
Au
Fi
it z
Ge
Sl
d
Ki
De
th
i te
w
Sw
Ne
Ne
Un
i te
Un
1. Based on registry data.
1 2 http://dx.doi.org/10.1787/888933281174
8.20. Obstetric trauma, vaginal delivery without instrument, 2013 (or nearest year)
Crude rates per 100 vaginal deliveries without instrument assistance

3.5
3.1
3.0 2.8 2.8

2.6 2.6 2.6
2.5
2.4
2.5
2.1
1.9
1.5
1.6
1.5
1.4
1.0
0.9
0.8
0.6 0.6
0.5 0.5 0.5 0.5
0.3
0.1
0
d1
ay 1
s1
k1
en 1
l
es
nd
nd
nd
el
li a
d
ia
ly
ce
a in
21
da
ga
an
an
ra
en
It a
nd
ar
iu
do
an
at
an
ra
na
la
la
la
CD
r tu
Sp
rw
ed
rm
al
Is
lg
nm
ov
St
Po
Ir e
er
st
la
nl
ng
Fr
Ca
Ze
OE
Sw
Be
No
Po
er
Au
Fi
it z
Ge
Sl
Ki
De
th
i te
w
Sw
d
Ne
Ne
Un
i te
Un
1. Based on registry data.

1 2 http://dx.doi.org/10.1787/888933281174

8. QUALITY OF CARE
The burden of mental illness is substantial, affecting an secondary prevention was sufficient (OECD, 2014a; OECD,
estimated one in four of the OECD population at any time, 2014b).
and one in two across the life course (OECD, 2014a). High
quality, timely care has the potential to improve outcomes
and may help reduce suicide and excess mortality for indi-
viduals with psychiatric disorders.
High quality care for mental disorders in inpatient settings The inpatient suicide indicator is composed of a
is vital. Figure 8.21 shows rates of inpatient suicide denominator of patients discharged with a principal
amongst all psychiatric hospital admissions. Inpatient sui- diagnosis or first two listed secondary diagnosis code
cide is a never event, which should be closely monitored of mental health and behavioural disorders (ICD-10
as an indication of how well inpatient settings are able to codes F10-F69 and F90-99) and a numerator of the
keep patients safe from harm. Most countries report rates number of patients who committed suicide (ICD-10
below 0.1 per 100 patients; Denmark and Estonia are excep- codes: X60-X84). There are often fewer than ten inpa-
tions with rates of 0.1 and 0.3 respectively. Steps to prevent tient suicides in a given year, meaning that reported
inpatient suicide include identification and removal of rates can vary. Where possible a 3-year average has
likely opportunities for self-harm, risk assessment of been calculated to give more stability to the indicator.
patients, monitoring and appropriate treatment plans. This was not possible for the Czech Republic, Portugal,
Suicide rate after discharge can be an indicator of the qual- and Switzerland. The data should be interpreted with
ity of care in the community, and co-ordination between caution due to a very small number of cases.
inpatient and community settings. The risk of suicide in Suicide within 30 days and within one year of dis-
the first year after discharge from psychiatric inpatient charge is established by linking discharge following
care is much greater than for the general population. Sui- hospitalisation with a principal diagnosis or first two
cide rate amongst patients who had been hospitalised in listed secondary diagnosis code of mental health and
the previous year was 0.43 per 100 patients, compared to a behavioural disorders (ICD-10 codes F10-F69 and
suicide rate of 0.01 per 100 for the general population in F90-99), with suicides recorded in death registries
2012 across OECD countries for which these data are avail- (ICD-10 codes: X60-X84). In cases with several admis-
able. Patients with a psychiatric illness are particularly at sions during the reference year, the follow-up period
risk immediately following discharge from hospital; in all starts from the last discharge.
countries suicide within 30 days of discharge amounted to For the excess mortality indicators the numerator is
at least one quarter of all suicides within the first year fol- the overall mortality rate for persons aged between 15
lowing discharge (Figure 8.22). Good discharge planning and 74 years old ever diagnosed with schizophrenia or
and follow-up, and enhanced levels of care immediately bipolar disorder. The denominator is the overall mor-
following discharge can help reduce suicide in the high-risk tality rate for the general population aged between 15
days immediately following discharge (OECD, 2014a). and 74 years old. The relatively small number of people
Individuals with a psychiatric illness have a higher mortal- with bipolar disorder dying in any given year can
ity rate than the general population. An excess mortality cause substantial variations from year to year in some
value that is greater than one implies that people with countries. The available data in most countries did
mental disorders face a higher risk of death than the rest of not allow the calculation of 2-year averages.
the population. Figures 8.23 and 8.24 show the excess mor- The data have been age-sex standardised to the 2010
tality for schizophrenia and bipolar disorder, which is OECD population structure, to remove the effect of
above two in all countries. A higher rate of physical illness different population structures across countries.
and chronic disease related to risk factors such as smoking,
drug and alcohol abuse, side effects of psychotropic treat-
ment and poor physical health care and increased risk of
suicide contribute to excess mortality. A multifaceted References
disease-related approach is needed to reduce this excess OECD (2014a), Making Mental Health Count. The Social and Eco-
mortality, including primary care prevention of physical ill nomic Costs of neglecting mental health care, OECD Publish-
health among people with mental disorders, better integra- ing, Paris, http://dx.doi.org/10.1787/9789264208445-en.
tion of physical and mental health care, behavioural inter-
OECD (2014b), OECD Reviews of Health Care Quality: Norway:
ventions, and changing professional attitudes. For
Raising Standards, OECD Publishing, Paris,
example, Sweden monitors the use of inpatient physical
http://dx.doi.org/10.1787/9789264208469-en.
care for patients with a mental disorder diagnosis that
could have been avoided if primary care and/or primary or

8. QUALITY OF CARE
8.21. Inpatient suicide amongst patients with 8.22. Suicide following hospitalisation for a psychiatric
a psychiatric disorder, 2013 (or latest year) disorder, within 30 days and one year of discharge, 2012
Within 30 days of discharge Within one year of discharge
Age-sex standardised rates per 100 patients Age-sex standardised rates per 100 patients
0.30 1.2
0.25
0.25 1.0
0.20 0.8
0.15 0.6
0.12
0.09
0.10
0.08
0.4
0.08
0.07
0.06
0.05
0.04
0.05
0.03
0.2
0.02
0.02
0.02
0.00
0 0
l
ak d
m
Po nd
De el
a
p.
i t z in
No .
w way
Be d
Es k
m
ga
p
ni
ar
Sl r l an
an
il e
m
p.
ia
el
d
en
ia
ra
S w Sp a
Re
Re
iu
do
a
a
ar
an
an
r tu
to
nm
ra
en
tv
Re
do
i te an
al
nl
Is
lg
Ch
ed
r
ng
nm
h
nl
al
La
Is
Ze
Fi
ov
C
ng
Sw
ec
Ze
Ki
Fi
ov
De
ec
Sl
Ki
Cz
w
Ne
Cz
d
Ne
i te
Un
Un
Note: Three-year average for most countries.

1 2 http://dx.doi.org/10.1787/888933281184
1 2 http://dx.doi.org/10.1787/888933281184
8.23. Excess mortality from schizophrenia, 2013 8.24. Excess mortality from bipolar disorder, 2013
(or latest year) (or latest year)
Men Women Men Women
Ratio Ratio
7 7
6.6
6.1
6 6
5.6
5.3
5.2
5.2
4.9
5 5
4.5
4.5
4.4
4.2
3.9
3.9
4 4
3.8
3.7
3.6
3.5
3.3
3.3
3.3
3.2
3.0
2.8
3 3
2.7
2.6
2.6
2.6
2.5
2.4
2 2
1.7
1 1
0 0
ia
el
el
a
k
en
ay
en
ay
d
re
re
ar
ar
an
an
an
an
ra
ra
tv
rw
rw
ed
ed
Ko
Ko
nm
nm
al
nl
al
nl
La
Is
Is
Sw
Sw
No
No
Ze
Ze
Fi
Fi
De
De
w
w
Ne
Ne
Note: Excess mortality is compared to the mortality rate for the general Note: Excess mortality is compared to the mortality rate for the general
population. population.

8. QUALITY OF CARE
Cervical cancer is highly preventable if precancerous

changes are detected and treated before progression Definition and comparability
occurs. The main cause of cervical cancer, which accounts
for approximately 95% of all cases, is exposure to the Screening rates are based on surveys or encounter
human papilloma virus (HPV) through sexual activity data, which may influence the results. Survey-based
(IARC, 2005). results may be affected by recall bias. Programme data
are often calculated for monitoring national screening
Countries follow different policies with regards to the pre-
programmes and differences in target population and
vention and early diagnosis of cervical cancer. About half of
screening frequency may also lead to variations in
OECD countries have cervical cancer screening organised
screening coverage across countries.
through population-based programmes but their periodicity
and target age groups vary (OECD, 2013). Some countries Relative survival is the ratio of the observed survival
with low cervical cancer incidence such as Israel and experienced by cancer patients over a specified period
Switzerland do not have an organised screening pro- of time after diagnosis to the expected survival in a
gramme but women in the eligible age group can have a comparable group from the general population in
Pap smear test performed every three years for free. WHO terms of age, sex and time period. Survival data for
recommends HPV vaccinations as part of national immun- Chile, Germany and Italy are based on a sample of
isation programmes, primarily for girls aged 9-13 years, in patients. The number of countries which monitor and
countries where the prevention of cervical cancer is a pub- report cancer survival has been increasing in recent
lic health priority, the introduction is feasible and finan- years and an international study (Allemani et al.,
cially sustainable, and cost-effectiveness has been 2015) also shows that a wide range of countries have
evaluated (WHO, 2014). Nowadays, most OECD countries cancer registries which enable international compari-
have HPV vaccination programmes. sons of cancer survival.
Screening rates for cervical cancer ranged from 20.7% in Countries use either period analysis or cohort analysis
Mexico to 84.5% in the United States in 2013 and have to calculate cancer survival. Period analysis gives an
increased from 57.0% to 61.6% on average across OECD up-to-date estimate of cancer patient survival using
countries over the past decade (Figure 8.25). The coverage more recent incidence and follow-up periods than
increase was particularly large in Korea where the screen- cohort analysis which uses survival information of a
ing programme was rolled out nationwide in the mid- complete five-year follow-up period. The reference
2000s. In about half of OECD countries, however, screening periods for diagnosis and follow-up years vary across
coverage declined, which may be related to the introduc- countries.
tion of HPV vaccinations, starting from the late 2000s Cancer survival presented have been age-standardised
(OECD, 2013). using the International Cancer Survival Standard
Cancer survival is one of the key measures of the effective- (ICSS) population.
ness of cancer care systems, taking into account both early See indicator Mortality from cancer in Chapter 3 for
detection of the disease and the effectiveness of treatment. definition, source and methodology underlying cancer
Five-year relative cervical cancer survival ranges widely mortality rates.
from 45.3% in Chile to 81.2% in Norway in recent years
(Figure 8.26). Some countries with relatively high screening
coverage such as the United States, Austria, the United
Kingdom, New Zealand and Ireland have lower survival,
References
but four of the five countries have low mortality. During the Allemani, C. et al. (2015), Global Surveillance of Cancer
past decade, five-year relative survival for cervical cancer Survival 1995-2009: Analysis of Individual Data for
improved in many countries. 25 676 887 Patients from 279 Population-based Registries
Mortality rates reflect the effect of cancer care over the past in 67 Countries (CONCORD-2), The Lancet, Vol. 385,
years and the impact of screening, as well as changes in pp. 977-1010.
incidence. The mortality rates for cervical cancer declined IARC International Agency for Research on Cancer (2005),
i n m o s t O E C D c o u n t r i e s b e t we e n 2 0 0 3 a n d 2 0 1 3 Cervix Cancer Screening, IARC Handbooks of Cancer Pre-
(Figure 8.27). In Greece, however, the mortality rate from vention, Vol. 10, International Agency for Research on
cervical cancer increased substantially by 47% during the Cancer, Lyon.
same period, although it is still below the OECD average. OECD (2013), Cancer Care: Assuring Quality to Improve Sur-
The incidence is low and decreasing over time and it is vival, OECD Publishing, Paris, http://dx.doi.org/10.1787/
likely that Greece can control the increasing burden of cer- 9789264181052-en.
vical cancer by providing more effective cervical cancer
WHO (2014), Human Papillomavirus Vaccines: WHO Posi-
treatment.
tion Paper, October 2014, Weekly Epidemiological Record,
No. 43, 89, 465492, Geneva.

8. QUALITY OF CARE
8.25. Cervical cancer screening in women aged 20-69, 8.26. Cervical cancer five-year relative survival, 1998-2003
2003 to 2013 (or nearest years) and 2008-2013 (or nearest periods)
2013 2003 2008-2013 1998-2003
United States Norway1

Austria 2 Korea1
Sweden1
United Kingdom1 Italy1
New Zealand1 Japan 2
Ireland1 Denmark 3
Switzerland2
Norway1 Finland1
France 2 Iceland*2
Canada 2 Estonia 3
Slovenia1
Sweden 2
Greece 2
Poland 2 Israel 2
Spain 2 Australia1
Finland1 Canada 2
Turkey1
Netherlands1 France 2
Denmark1 OECD22
Iceland*1 Czech Rep. 2
OECD24
New Zealand1
Chile1
Belgium1 Germany1
Australia1 Netherlands2
Luxembourg1 Belgium1
Portugal2
Germany2 Portugal2
Czech Rep.1 Austria 2
Korea1 Ireland1
Estonia1
United States 2
Slovak Rep.1
Lithuania1 United Kingdom 2
Japan 2 Latvia 2
Italy1 Slovenia1
Hungary1
Latvia1 Poland1
Mexico 1 Chile 2
0 25 50 75 100 0 25 50 75 100
% of women screened Age-standardised survival (%)
1. Programme. 2. Survey. * Three-year average. 1. Period analysis. 2. Cohort analysis. 3 Different analysis methods used
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en. for different years. * Three-period average. 95% confidence intervals
1 2 http://dx.doi.org/10.1787/888933281196 represented by H.
1 2 http://dx.doi.org/10.1787/888933281196
8.27. Cervical cancer mortality, 2003 to 2013 (or nearest years)
2003 2013
Age-standardised rates per 100 000 women
18
16
14
12
10
8
6
4
2
0
st d
Ic I t a l y
S w F in *
i t z land
Tu lia
Ca ey
Fr da
er in
Ne I ds
Lu Ze el
Be tes
S w gal
nm n
Gr c e
Ne S ce
i te b nd
Un Kin rg*
d m
N o um
Au ay
rm a
r tu y
Ja k
Sl p a n
C z Ir e l a a
ec nd
E s il e
CD a
Ko 4
ak r y
Br .
Ru P o i l
ia nd
s t ni a
La a
th ia
M ni a
lo o
a
S l u n g p.
Ch .
p
d
Po an
Ge s tr i
bi
O E eni
re
ar
ic
az
d
Au l an
De ede
C o ex i c
3
w sr a
Li t v
th pa
Re
H e
i t e gdo
Fe
ov a
rk
rw
an
ee
ra
na
n
an
ss la
Un xem a l a
aR
m
C o to
a
ua
R
i
d ou
la
lg
ov
St
er
n
el
* Three-year average.
1 2 http://dx.doi.org/10.1787/888933281196

8. QUALITY OF CARE
Breast cancer is the most prevalent form of cancer in tries such as Estonia, the Czech Republic and Latvia,
women across OECD countries. One in nine women will although survival after breast cancer diagnosis is still
have breast cancer at some point in their life and one in below the OECD average. The improvement may be related to
thirty will die from the disease. Risk factors that increase a strengthening of cancer care governance in these countries.
person's chance of getting this disease include age, family For instance, the Czech Republic intensified its effort to detect
history of breast cancer, genetic predisposition, reproduc- breast cancer patients early through the introduction of a
tive factors, oestrogen replacement therapy, and lifestyles screening programme in 2002 and implemented a National
including obesity, physical inactivity, diet and alcohol con- Cancer Control Programme in 2005 to improve the quality
sumption. of cancer care and cancer survival. Cancer care delivery
Most OECD countries have adopted breast cancer screening was reorganised by reducing the number of comprehensive
programmes as an effective way for detecting the disease cancer centres while aiming to optimise the population
early, though the periodicity and population target groups coverage of each centre, and skilled professionals and neces-
vary across countries (OECD, 2013). Due to recent progress sary investment were allocated at each centre. The current
in treatment outcomes and concerns about false-positive cancer care delivery model is considered to be well organised
results, over-diagnosis and overtreatment, breast cancer and distributed adequately around the country, and, partly
screening recommendations have been re-evaluated in due to the more equal access, variations in cancer survival
recent years. Taking account of recent research findings, across regions have been reduced (OECD, 2013; OECD, 2014).
WHO recommends organised population-based mammog- Mortality rates have declined in most OECD countries over
raphy screening if women are able to make an informed the past decade (Figure 8.30). The reduction is a reflection
decision based on the benefits and risks of mammography of improvements in early detection and treatment of breast
screening (WHO, 2014). cancer. Improvements were substantial in the Czech
Screening rates ranged from less than 20% in Mexico to Republic, Norway and the Netherlands with a decline of
over 80% in Finland, Slovenia, Denmark and the United over 20% in a decade. Denmark also reported a consider-
States in 2013 (Figure 8.28). Screening coverage increased able decline, but its mortality rate was still the highest in
substantially among countries with low rates a decade ago. 2013. On the other hand, in Korea, Turkey and Japan, the
Mexico and Chile had an increase of more than ten-fold, mortality rate from breast cancer increased over the past
Korea an over four-fold increase, and the Slovak Republic decade, although it remains the lowest among OECD coun-
and Lithuania a three-fold rise. On the other hand, countries, and the incidence of breast cancer has doubled or
tries that had the highest screening rates in the early 2000s more in the past decade.
experienced some reductions, including Finland, the
United States, the Netherlands, Ireland and Norway. In
Ireland, the screening programme, which was commenced
on a phased basis in 2000, completed its nationwide roll- Definition and comparability
out in 2009, but it is still at a stage too early to evaluate the
coverage trend over time. Screening rates and survival are defined in indicator
Breast cancer survival reflects early diagnosis as well as
in Chapter 8. See indicator Mortality from cancer in
improved treatments. All OECD countries have attained
Chapter 3 for definition, source and methodology
five-year relative breast cancer survival of 80% except Esto-
underlying cancer mortality rates.
nia, Poland and Chile (Figure 8.29). Relative survival of peo-
ple with cervical and colorectal cancers is also the lowest
for Poland and Chile (see indicators Screening, survival
and mortality for cervical cancer and Survival and mor- References
tality for colorectal cancer). In both countries, access to
care is limited due to fewer numbers of cancer care centres OECD (2014), OECD Reviews of Health Care Quality: Czech
and radiotherapy facilities. In Chile, some cancer drugs and Republic 2014: Raising Standards, OECD Publishing, Paris,
other medical technologies are not widely available, and http://dx.doi.org/10.1787/9789264208605-en.
there are not enough specialised professionals, resulting in OECD (2013), Cancer Care: Assuring Quality to Improve
a long waiting time for cancer treatment (OECD, 2013). Survival, OECD Publishing, Paris,
Over the last decade, the five-year relative breast cancer http://dx.doi.org/10.1787/9789264181052-en.
survival has improved in all OECD countries. Relative survival WHO (2014), WHO Position Paper on Mammography
has increased considerably in some Eastern European coun- Screening, Geneva.

8. QUALITY OF CARE
8.28. Mammography screening in women aged 50-69, 8.29. Breast cancer five-year relative survival, 1998-2003
2003 to 2013 (or nearest years) and 2008-2013 (or nearest periods)
2013 2003 2008-2013 1998-2003
Finland1 Sweden 2
Slovenia1
United States 2
Denmark1
United States 2 Norway1
Austria 2 Finland1
Netherlands1 Australia1
Spain 2
United Kingdom1 Portugal2
Norway1 Israel 2
Portugal2 Canada 2
New Zealand
Japan 2
Canada 2
Germany2 Iceland*2
Israel1 Denmark 3
Ireland1 New Zealand
Korea1
Italy1 Belgium1
Luxembourg*1 Korea1
OECD27 Italy1
Iceland*1
Germany1
Czech Rep.1
Poland 2 Netherlands 2
Belgium1 OECD22
Australia1 Latvia 2
Estonia1
France 1 France 2
Greece 2 Slovenia1
Switzerland2 Austria 2
Hungary1
Ireland1
Japan 2
Slovak Rep.1 United Kingdom 2
Lithuania1 Czech Rep. 2
Turkey1 Chile 2
Latvia1
Chile1 Poland1
Mexico 1 Estonia 3
0 25 50 75 100 0 20 40 60 80 100
% of women screened Age-standardised survival (%)
1. Programme. 2. Survey. * Three-year average. 1. Period analysis. 2. Cohort analysis. 3 Different analysis methods used
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en. for different years. * Three-period average. 95% confidence intervals
1 2 http://dx.doi.org/10.1787/888933281202 represented by H.
1 2 http://dx.doi.org/10.1787/888933281202
8.30. Breast cancer mortality in women, 2003 to 2013 (or nearest years)
2003 2013
Age-standardised rates per 100 000 women
45
40
35
30
25
20
15
10
5
0
Tu ea
Ja y
h al
M an
lo o
a
il e
aR l
Sp a
No a in
So or ay
Fi ic a
Un Aus en
Au nd
S w and
d li a
OE land
C z an 4
h a
S w Gr e i a
t h p.
er e
ria
Ru L a l y
Sl i an i a
F r p.
Ne Es ce
L u Z e ni a
ng *
er el
Ic n d s
Sl nd*
Hu ium
a k d.
i te b nd
Ge dom
N e Is n y
B e nd
Ir e n i a
nm y
k
Po tes
st a zi
e
bi
ec ad
ic
it z ec
Ki rg
ar
De ar
C o ex i c
C 3
ut ug
th r a
ss at v
Re
L i Re
ov F e
rk
P w
r
Ch
an
i te tr a
ed
p
a
la
la
Un xem a l a
st
CD
r
It
m
w to
ua
e
Ko
ng
d ou
Co r
rm
nl
la
a
lg
Af
r
ov
B
St
el
1 2 http://dx.doi.org/10.1787/888933281202

8. QUALITY OF CARE
Colorectal cancer is the third most commonly diagnosed 55.9% for females. Slovenia, Latvia and Sweden also have a
form of cancer after prostate and lung cancers for men, and comparatively large difference.
the second most common cancer after breast cancer for Most countries experienced a decline in mortality of
women, across OECD countries. Colorectal cancer incidence colorectal cancer in recent years, with the average rate
is high in Korea, the Slovak Republic, Hungary, Denmark and across OECD countries falling from 27.4 to 24.2 deaths per
the Netherlands at 40 or more cases per 100 000 population 100 000 population between 2003 and 2013 (Figure 8.33).
while it is low in Mexico, Greece, Chile and Turkey at less The decline was particularly large in the Czech Republic,
than half this rate. Incidence is significantly higher for men Austria and Australia with a reduction of over 25%. The
than women across countries. There are several factors main exceptions to this general trend were Turkey, Brazil,
that place certain individuals at increased risk for the dis- Chile and Mexico where the mortality rate from colorectal
ease, including age, ulcerative colitis, a personal or family cancer increased by more than 10% over the last decade,
history of colorectal cancer or polyps, and lifestyle factors although the rate remains much lower than the OECD aver-
such as a diet high in fat and low in fibre, lack of physical age. Despite some progress, Central and Eastern European
activity, obesity, and tobacco and alcohol consumption. countries, particularly Hungary, the Slovak Republic, Slovenia
Following screening for breast and cervical cancers, and the Czech Republic, continue to have higher mortality
colorectal cancer screening has become available, and an rates than other OECD countries.
increasing number of countries have introduced free pop- Across countries, colorectal cancer continues to be an
ulation-based screening, targeting people in their 50s important cause of cancer death for both men and women
and 60s (OECD, 2013). Partly because of uncertainties about (see indicator Mortality from cancer in Chapter 8) and
the cost-effectiveness of screening (Lansdorp-Vogelaar et countries will need to make further effort to promote not
al., 2010), countries are using different methods (i.e. faecal only early diagnosis and effective treatment but also
occult blood test, colonoscopy and flexible sigmoidoscopy). healthy lifestyles to reduce its risk factors (see Chapter 8
Multiple methods are also available within the screening Non-medical determinants).
programme in some countries. In most countries that
provide faecal occult blood test, screening is available every
two years. The screening periodicity schedule is less frequent
with colonoscopy and flexible sigmoidoscopy, generally
every ten years, making it difficult to compare screening
coverage across countries.
Survival and mortality rates are defined in indicator
Advances in diagnosis and treatment of colorectal cancer Screening, survival and mortality for cervical cancer
including improved surgical techniques, radiation therapy in Chapter 8. See indicator Mortality from cancer in
and combined chemotherapy and their wider and timelier Chapter 3 for definition, source and methodology
access have contributed to increased survival over the last underlying cancer mortality rates. Survival and mor-
decade. All OECD countries showed improvement in five- tality rates of colorectal cancer are based on ICD-10
year relative survival for colorectal cancer. On average, five- codes C18-C21 (colon, rectosigmoid junction, rectum,
year colorectal cancer survival improved from 55.8% to and anus).
62.2% for people with colorectal cancer during 1998-2003 to
2008-2013 respectively (Figure 8.31). Poland, Estonia and
the Czech Republic also had a considerable improvement,
but cancer survival in these countries is still the lowest References
among OECD countries at less than 55%. Korea and Israel Lansdorp-Vogelaar, I., A.B. Knudsen and H. Brenner (2010),
had the highest survival at over 70%. Cost-effectiveness of Colorectal Cancer Screening An
In most OECD countries, colorectal cancer survival is higher Overview, Best Practice & Research Clinical Gastroenterol-
for women but in Chile, Korea, Israel, Japan, Portugal, Austria ogy, Vol. 24, pp. 439- 449.
and the Netherlands, men have a slightly higher survival OECD (2013), Cancer Care: Assuring Quality to Improve Sur-
(Figure 8.32). The gender difference is the largest in Estonia vival, OECD Publishing, Paris,
with the five-year relative survival of 48.4% for males and http://dx.doi.org/10.1787/9789264181052-en.

8. QUALITY OF CARE
8.31. Colorectal cancer, five-year relative survival, 8.32. Colorectal cancer, five-year relative survival
1998-2003 and 2008-13 (or nearest periods) by gender, 2008-13 (or nearest periods)
2008-2013 1998-2003 Men Women
Korea1 Korea1
Israel 2 Israel 2
Iceland*2
Australia1
Australia1
Japan 2
Japan 2
Belgium1
Belgium1
Sweden 2
Sweden 2
Finland1
Finland1
Austria 2
Austria 2
United States 2
United States 2
Germany1
Germany1
Netherlands 2 Netherlands 2
Italy1 Italy1
Norway1 Norway1
Canada 2 Canada 2
New Zealand1 New Zealand1
OECD21 OECD25
Portugal2 Portugal2
Slovenia1 Slovenia1
Denmark 3 Denmark 3
Ireland1 Ireland1
Latvia 2 Latvia 2
France 2 United Kingdom 2
United Kingdom 2 Czech Rep. 2
Czech Rep. 2 Estonia 3
Estonia 3 Poland1
Poland1 Chile 2
0 20 40 60 80 100 0 20 40 60 80 100
Age-standardised survival (%) Age-standardised survival (%)
1. Period analysis, 2. Cohort analysis. 95% confidence intervals 1. Period analysis. 2. Cohort analysis. 3 Different analysis methods used
represented by H. for different years. * Three-period average. 95% confidence intervals
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en. represented by H.
1 2 http://dx.doi.org/10.1787/888933281219 Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en.
1 2 http://dx.doi.org/10.1787/888933281219
8.33. Colorectal cancer mortality, 2003 to 2013 (or nearest years)
2003 2013
50
45
40
35
30
25
20
15
10
5
0
h co
B a
l o il
Co Turk a
st ey
Gr i c a
Sw C s
Un F i e c e
er e
st d
d nd
Ic r a li a
Au d*
m
Ko i a
i te F r e a
ng e
B e om
Sw ly
rm n
Ja a
C a ny
xe Is n
bo l
Ne Po gal
OE rg*
th 4
N e Ir e i a
er d
E s ds
Ru S n i a
a in in
r tu y
N o e d.
Ze nd
C z nm d
h k
Sl lov i a
ak ia
ng .
y
L a p.
m r ae
Hu Rep
bi
e
i t z hil
d
Ki c
Po a
ric
Co r a z
ec ar
ar
Au l an
Ge ede
pa
th lan
De lan
Li D 3
n
ov en
r
It a
S tv
ss pa
iu
Re
at
d r an
rw
ut xi
a
na
n
an
i t e nl a
w la
st
aR
m
ua
to
F
d
u
e
la
lg
Af
So Me
St
a
el
Lu
Un
1 2 http://dx.doi.org/10.1787/888933281219

8. QUALITY OF CARE
All OECD countries have established vaccination pro- national infant immunisation programme. Across the
grammes based on their interpretation of the risks and OECD, the average immunisation coverage for hepatitis B
benefits of each vaccine. Figures 8.34 and 8.35 show that for children aged one year old is 92%. In countries such as
the overall vaccination of children against measles and China, the Czech Republic and Korea, it reaches 99%. How-
diphtheria, tetanus and pertussis (DTP) is high in OECD ever, a number of countries do not require children to be
countries. On average, 95% of children receive the recom- vaccinated, and consequently the rates for these countries
mended DTP vaccination and 94% receive measles vaccina- are significantly lower than other countries. For example,
tions in accordance with national immunisation schedules. in Denmark, Sweden and the United Kingdom, vaccination
Rates for DTP vaccinations are below 90% only in Indonesia, against hepatitis B is not part of the general infant vaccina-
Austria, Mexico, India and South Africa. Rates for measles tion programme, but is provided to high-risk groups such as
vaccinations are below 90% in Denmark, France, Mexico, children with mothers who are infected by the hepatitis B
Indonesia, Austria, India and South Africa. virus. Other OECD countries that do not include vaccination
While national coverage rates are high in many countries, against hepatitis B in their infant programmes are Iceland,
some parts of the population remain exposed to certain Finland, Hungary, Japan, Slovenia and Switzerland. In
diseases. For example, the United States reported 189 measles Canada, not all jurisdictions immunise infants against
cases between 1 January and 18 Septembre 2015. Most of hepatitis B, with some doing this at school age.
these cases were linked to an amusement park in
California. The Centers for Disease Control and Prevention
reported that most of the measles cases in 2015 were in
unvaccinated people. In the previous year, over 650 cases of Definition and comparability
measles were reported in the United States, the highest
number of cases since measles elimination was docu- Vaccination rates reflect the percentage of children
mented in 2000. Many of the cases were associated with a that receives the respective vaccination in the recom-
large outbreak that originated in the Philippines (Centers mended timeframe. The age of complete immunisation
for Disease Control and Prevention, 2015). In July 2015, the differs across countries due to different immunisation
first death related to measles since 2003 was reported in schedules. For those countries recommending the
the United States (Washington State Department of Health, first dose of a vaccine after age one, the indicator is
2015). calculated as the proportion of children less than two
years of age who have received that vaccine. Thus,
Parts of Europe also reported large number of measles
these indicators are based on the actual policy in a
cases in 2015. During the 12 months to June 2015, more
given country.
than 4 000 cases were reported across 30 countries. More
than half the cases were in Germany, with over 400 cases Some countries administer combination vaccines (e.g.
reported in Italy. The measles-related death of an DTP for diphtheria, tetanus and pertussis) while oth-
18-month toddler in Germany was reported in February ers administer the vaccinations separately. Some
2015. Most of the cases across Europe were among unvacci- countries ascertain vaccinations based on surveys
nated people (European Centre for Disease Prevention and and others based on encounter data, which may influ-
Control, 2015). Catch-up programmes in older children may ence the results.
be needed to avoid the risk of, or respond to, measles out-
breaks. Such a campaign was conducted in the United
Kingdom in 2013.
Figure 8.36 shows the percentage of children aged one year References
vaccinated for hepatitis B. The hepatitis B virus is transmit-
Centers for Disease Control and Prevention (2015), Measles
ted by contact with blood or body fluids of an infected per-
Cases and Outbreaks, available at:
son. A small proportion of infections become chronic, and
www.cdc.gov/measles/cases-outbreaks.html (accessed 12/
these people are at high risk of death from cancer or cirrhosis
10/2015).
of the liver. A vaccination has been available since 1982 and
is considered to be 95% effective in preventing infection European Centre for Disease Prevention and Control (2015),
and its chronic consequences. Since a high proportion of Surveillance Report: Measles and Rubella Monitoring,
chronic infections are acquired during early childhood, the July 2015.
WHO recommends that all infants should receive their first Washington State Department of Health (2015), Measles
dose of hepatitis B vaccine as soon as possible after birth, Led to Death of Clallam Co. Woman; First in US in a
preferably within 24 hours (WHO, 2015). Dozen Years, available at www.doh.wa.gov/Newsroom/
Most countries have followed the WHO recommendation to 2015NewsReleases/15119WAMeaslesRelatedDeath.
incorporate hepatitis B vaccine as an integral part of their WHO (2015), Hepatitis B, Fact Sheet No. 204, Geneva.

8. QUALITY OF CARE
8.34. Vaccination against diphteria, tetanus and 8.35. Vaccination against measles, children aged 1, 2013
pertussis, children aged 1, 2013
Belgium Brazil
China China
Czech Rep. Czech Rep.
France Greece
Greece Hungary
Hungary Korea
Korea Poland
Luxembourg Portugal
Poland Russian Fed.
Finland Slovak Rep.
Japan Turkey
Portugal Finland
Slovak Rep. Germany
Sweden Israel
Turkey Sweden
Italy Latvia
Netherlands Netherlands
Russian Fed. Canada
Canada Japan
Germany Luxembourg
Ireland Spain
Spain United Kingdom
Switzerland OECD34
United Kingdom Australia
OECD34 Estonia
Brazil Slovenia
Costa Rica Ireland
Latvia Lithuania
Slovenia Norway
Denmark Switzerland
Estonia Belgium
Israel Colombia
Norway New Zealand
United States Iceland
Lithuania Costa Rica
New Zealand United States
Australia Chile
Chile Italy
Iceland Denmark
Colombia France
Indonesia Mexico
Austria Indonesia
Mexico Austria
India India
South Africa South Africa
0 25 50 75 100 0 25 50 75 100
% of children vaccinated % of children vaccinated
Source: WHO/UNICEF. Source: WHO/UNICEF.

8.36. Vaccination against hepatitis B, children aged 1, 2013

% of children vaccinated
100
80
60
40
20
0
Au 24
le
Ge tes
ce
a
a
ak l
nd
Ze a
O E nd
C o r a li a
do y
M ia
Ca o
Fr a
Es rg
L i ni a
Au ia
ec na
Be a
p.
ia y
ey
Ir e i l
nd
th v ia
s t in
m a
Ko .
m
ce
Po ael
Tu .
ov g a
p
nd
bi
In a n
u t di
w a ni
ric
re
ss It al
xe i c
az
ic
i t e hi
r
Co a
Re
Re
iu
Fe
u
rk
an
ee
na
C z Chi
a
la
la
CD
st
ne
Lu a R
r
Ne L at
S o In
to
Sl tu
ex
bo
Br
rm
C
la
al
Is
lg
Af
N e t hu
St
S
Po
st
Gr
lo
r
h
er
h
d
Ru
Un
Source: WHO/UNICEF.
1 2 http://dx.doi.org/10.1787/888933281226

8. QUALITY OF CARE
Influenza is a common infectious disease affecting 5%-10% included in most 2009-10 vaccination programmes. In part,
of adults and 20%-30% of children. There are an estimated this may be due to the easing of concerns about the threat
3 to 5 million cases of severe influenza-related illness of H1N1 amongst the general population by the time the
worldwide each year, and 250 000 to 500 000 deaths (WHO, vaccine became available. Studies have shown that the
2014). Influenza can also have a major impact on health most important determinant for individuals to take-up
care systems. In the United States, it is estimated that each H1N1 vaccine was previous exposure to seasonal flu vaccine,
year, more than 200 000 people are hospitalised for respira- leading some researchers to argue that higher vaccination
tory and heart condition illnesses associated with seasonal rates for seasonal flu may help take-up during potential
influenza virus infections (Thompson et al., 2004). At certain future pandemics (Nguyen et al., 2011).
times of the year, influenza can place health systems under
significant stress. For example, in Ontario, Canada, the
average annual rate of emergency department visits
attributable to seasonal influenza is 500 per 100 000 popu- Definition and comparability
lation. This rate increased to an estimated 1 000 per
100 000 population during the H1N1 pandemic in 2009 Influenza vaccination rate refers to the number of
(Schanzer et al., 2013). people aged 65 and older who have received an annual
In 2003, countries participating in the World Health Assem- influenza vaccination, divided by the total number of
bly committed to the goal of attaining vaccination coverage people over 65 years of age. In some countries, the
against influenza of at least 50% of the elderly population data are for people over 60 years of age. The main lim-
by 2006 and 75% by 2010. Figure 8.37 shows that in 2013, the itation in terms of data comparability arises from the
OECD average influenza vaccination rate for people aged 65 use of different data sources, whether survey or pro-
and over was 48%. Vaccination rates are as low as 1.1% in gramme, which are susceptible to different types of
Estonia, where influenza vaccination is recommended but errors and biases. For example, data from population
not free. Only four countries have attained the 75% target: surveys may reflect some variation due to recall errors
Mexico, Korea, Chile and the United Kingdom. Australia and irregularity of administration.
came close to meeting the target.
Figure 8.38 indicates that between 2003 and 2013, the vacci-
nation rate against influenza among the elderly population
has remained stable on average among the group of OECD References
countries that have trend data over this period, but with no Kelly, H. et al. (2011), The Age-specific Cumulative Inci-
uniform trend across countries. In some countries, such as dence of Infection with Pandemic Influenza H1N1 2009
New Zealand, Israel, Germany, Denmark, the Czech Republic Was Similar in Various Countries Prior to Vaccination,
and the United Kingdom, the percentage of the population PLoS One, Vol. 6, No. 8:e21828.
aged 65 and over vaccinated against influenza has
Nguyen, T. et al. (2011), Acceptance of A Pandemic Influ-
increased, while it has come down in other countries such
enza Vaccine: A Systematic Review of Surveys of the
as the Netherlands, Spain, France, the Slovak Republic and
General Public, Infection and Drug Resistance, Vol. 4,
Slovenia.
pp. 197-207.
In June 2009, the WHO declared an influenza pandemic.
Schanzer, D.L., B. Schwartz and M.J. Mello (2013), Impact of
The H1N1 influenza virus (also referred to as swine flu)
Seasonal and Pandemic Influenza on Emergency Depart-
infected an estimated 11% to 18% of the global population
ment Visits, 2003-2010, Ontario, Canada, Academic Emer-
(Kelly et al., 2011). Mexico was at the centre of the pan-
gency Medicine, Vol. 20, No. 4, pp. 388-397.
demic, being among the first countries where swine flu was
detected and also where mortality rates were reportedly Thompson, W.W. et al. (2004), Influenza-Associated Hospi-
higher than those in many other countries. The high rate of talizations in the United States, Journal of American
seasonal vaccinations that are still being observed in Mexico Medical Association, Vol. 292, No. 11, pp. 1333-1340.
may come as a result of the H1N1 experiences in that coun- WHO (2014), Influenza (Seasonal), Fact Sheet No. 211, avail-
try. In other countries, however, the take-up rate of H1N1 able at: www.who.int/mediacentre/factsheets/fs211/en/
vaccine was lower than expected, despite the vaccine being (accessed 23/06/2015).

8. QUALITY OF CARE
8.37. Influenza vaccination coverage, population aged 65 and over, 2013 (or nearest year)
%
100
90
80 78.5 77.4
76.5 75.5
74.6
70 69.0 68.8
66.5
64.1
61.2
60 59.2 58.6 58.0
56.4
54.2
51.9
50.0 49.9
50 48.0
46.0 45.8 45.6
43.3
41.4 41.0
40 36.7 36.1
30
22.1 20.8
20
15.6
13.2 13.0 12.1
10
1.1
0
OE gal
a
s
es
da
o
Ir e e l
Ge and
S w nd
De en
Au om
a
th nd
B e ny
a in
Fr y
ce
Po n
it z 3 3
m k
Gr g
Fi e
Au y
Hu d
ec ia
ak y
N o p.
Tu .
Sl e y
Po a
nd
a
p
l
i te nd
a
re
w a li
ni
c
i
l
xe a r
ar
ic
pa
an
ur
d Chi
ra
en
It a
C z s tr
Re
Re
iu
rk
at
Sl r w
an
ee
ed
na
la
la
Ne eala
Sw CD
r tu
Sp
to
ex
Ko
L u nm
d
ng
bo
Ne s tr
rm
Un er l a
Ja
nl
Is
lg
ov
St
er
ng
Ca
Es
h
M
d
Ki
ov
i te
Un

1 2 http://dx.doi.org/10.1787/888933281235
8.38. Influenza vaccination coverage, population aged 65 and over, 2003 and 2013 (or nearest years)
% 2003 2013
90
80
70
60
50
40
30
20
10
0
l
a
il e
es
da
el
nd
m
li a
ce
en
ria
p.
d
a in
ly
26
p.
ia
ga
nd
an
re
ar
ar
pa
an
an
ur
ra
en
It a
Re
Re
iu
do
at
an
Ch
ra
ed
na
la
CD
st
r tu
Sp
Ko
nm
ng
bo
rm
la
Ja
nl
al
Is
lg
ov
St
Ir e
st
ng
Au
Fr
Sw
Ca
ak
er
Ze
Fi
OE
Be
Hu
m
Po
Au
Ge
De
ec
Sl
d
Ki
ov
th
xe
i te
w
Cz
Ne
d
Sl
Ne
Lu
Un
i te
Un

1 2 http://dx.doi.org/10.1787/888933281235

8. QUALITY OF CARE
Delivering health care that is responsive and patient-centered patients report positive experiences with regards to time
is playing a greater role in health care policy across OECD spent with the doctor (Figure 8.39), easy-to-understand
countries. Measuring and monitoring patient experience explanations (Figure 8.40), opportunities to ask questions
empowers patients and the public, involves them in deci- or raise concerns (Figure 8.41), as well as involvement in
sions on health care delivery and governance, and provides care and treatment decisions (Figure 8.42). For all four
insight into the extent to which they are health-literate and aspects of patient experience, Belgium and Luxembourg
have control over the treatment they receive. Across coun- score high at above 95% of patients reporting positive expe-
tries, using the health care user as a direct source of infor- riences. Poland has lower rates with fewer than one in two
mation is becoming more prevalent for health system patients reporting having been given the opportunity to ask
monitoring, planning and decision making, and efforts to questions or been involved in their care and treatment dur-
measure and monitor patient experiences have actually led ing consultation. The proportion of patients with positive
to health care quality improvements (Fujisawa and Klazinga, experience has decreased since 2010 in Australia, France,
forthcoming). the Netherlands and Switzerland, but countries with
Since the mid-1990s, there have been efforts to institution- lower rates such as Sweden and Poland have improved
alise measurement and monitoring of patient experiences. some aspects of patient experiences in recent years (Com-
In many countries, responsible organisations have been monwealth Fund, 2010).
established or existing institutions have been taking charge
of measuring and reporting patient experiences. They
developed survey instruments for regular collection of
patient experience data and standardised procedures for Definition and comparability
analysis and reporting. An increasing number of countries
collect not only Patient-Reported Experience Measures In order to measure and monitor general patient
(PREMs) but also Patient-Reported Outcome Measures experience in the health system, the OECD recom-
(PROMs) which collect patients perception on their specific mends collecting data on patient experience with any
medical conditions and general health, including mobility, doctor in ambulatory settings. An increasing number of
pain/discomfort and anxiety/depression, before and after a countries have been collecting patient experience data
specific medical intervention such as hip and knee replace- based on this recommendation through nationally rep-
ment. resentative population surveys while Japan and Portugal
collect them through nationally-representative service
A growing number of countries are using patient-reported
user surveys. Some countries, however, collect data
data to drive quality improvements in health systems.
on patient experience with a regular doctor. For about
Patient experience data are reported in periodic national
half the countries presented, the Commonwealth
health system reports or on public websites, showing dif-
Fund's International Health Policy Surveys 2010 and 2013
ferences across providers, regions and over time. Korea,
were used, even though there are critiques relating to
Norway, Sweden and the United Kingdom use patient expe-
the sample size and response rates. Data from this
rience measures in payment mechanisms or for fund allo-
survey refer to patient experience with a regular doc-
cations to promote quality improvement and patient-
tor rather than any doctor.
centred care, and Australia, Canada, the Czech Republic,
Denmark and France use them to inform health care regu- Rates are age-sex standardised to the 2010 OECD pop-
lators for inspection, regulation and/or accreditation. ulation, to remove the effect of different population
Patient-reported measures are also used in some Canadian structures across countries.
jurisdictions, Denmark, France and the Netherlands to pro-
vide specific feedback for providers quality improvement.
In England, PROMs and patients feedback about their
experience are used to inform patient choice and to incen-
References
tivise service improvement. For example, PROMs data for CIHI Canadian Institute for Health Information (2015),
patients undergoing some procedures such as hip and knee CIHI Proms Forum, Background Document,
replacement are used for benchmarking hospitals. The use www.cihi.ca/proms.
of PROMs can also enable the potential shift from a vol- Commonwealth Fund (2010), 2010 International Health
ume-based to a value-based model of health system Policy Survey in Eleven Countries, Chartpack, Common-
resource management (Canadian Institute for Health Infor- wealth Fund, New York.
mation, 2015).
Fujisawa, R. and N. Klazinga (forthcoming), Measuring
Patients generally report positive experiences when it Patient Experiences (PREMs): Progress Made by the OECD
comes to communication and autonomy in the ambulatory and its Member Countries 2006-2015, OECD Health Work-
health care system. Across countries, the majority of ing Papers, Paris.

8. QUALITY OF CARE
8.39. Doctor spending enough time with patient 8.40. Doctor providing easy-to-understand explanations,
in consultation, 2013 (or nearest year) 2013 (or nearest year)
Belgium1 97.5 Belgium1 97.8
Czech Rep.1 97.2 Luxembourg1 97.5
Luxembourg1 95.6 Portugal1, 2 96.3
New Zealand1 92.6 Japan1 96.3
Portugal1, 2 89.6 Czech Rep.1 96.2
Germany 2
88.2 New Zealand1 90.9
Estonia1 86.9 Germany2 90.7
Australia1 86.5 United Kingdom 2 89.5
United Kingdom 2 86.3 OECD19 87.9
Netherlands 2 85.1
Estonia1, 2 87.4
OECD18 84.9
Netherlands2 86.8
United States 2 86.3

Switzerland2 83.6
Australia 2 85.9
Israel1 81.8
2
Canada 2 85.4
United States 80.9
Norway2 84.1
France 2 80.0
France 2 83.7
Norway2 79.6
Israel1 83.2
Canada 2 79.3
Switzerland2 81.9
Sweden 2 78.3
Sweden 2 81.8
Poland1, 2 59.6
Poland1, 2 69.5
0 20 40 60 80 100 0 20 40 60 80 100
Age-standardised rates per 100 patients Age-standardised rates per 100 patients
Note: 95% confidence intervals represented by H. Note: 95% confidence intervals represented by H.
1. National sources. 2. Data refer to patient experiences with regular doctor. 1. National sources. 2. Data refer to patient experiences with regular doctor.
Source: Commonwealth Fund International Health Policy Survey 2013 and Source: Commonwealth Fund International Health Policy Survey 2013
other national sources. and other national sources.
8.41. Doctor giving opportunity to ask questions or raise 8.42. Doctor involving patient in decisions about care and
concerns, 2013 (or nearest year) treatment, 2013 (or nearest year)
Belgium1 97.7 Luxembourg1 95.5
Luxembourg 1
95.3 Belgium1 95.1
Switzerland2 94.4 Portugal1 90.9
Czech Rep. 2 94.0 New Zealand1 88.2
Germany2 94.0 United Kingdom 2 88.0
Netherlands 2
92.5 Germany2 87.7
United Kingdom 2 92.2 Australia 2 86.0
New Zealand 2 92.0 United States 2 83.9
Portugal1 91.8 Netherlands2 83.9
Canada 2
88.3 Norway2 83.3
Australia 2 88.3 Canada 2 83.0
United States2 86.7 Czech Rep.1 81.8
OECD19 85.0 Switzerland2 81.4
Norway 2
83.4 OECD19 81.3
Estonia1, 2 83.2 Sweden 2 80.5
France 2 82.8 Israel1 79.7
Israel1 78.4 France 2 78.8
Sweden 2 75.8 Estonia1, 2 67.4
Japan1 69.8 Spain1, 2 62.1
Poland 1, 2 33.6 Poland1, 2 47.9
0 20 40 60 80 100 0 20 40 60 80 100
Age-standardised rates per 100 patients Age-standardised rates per 100 patients
Note: 95% confidence intervals represented by H. Note: 95% confidence intervals represented by H.
1. National sources. 2. Data refer to patient experiences with regular doctor. 1. National sources. 2. Data refer to patient experiences with regular doctor.
Source: Commonwealth Fund International Health Policy Survey 2010 Source: Commonwealth Fund International Health Policy Survey 2013
and other national sources. and other national sources.

9. HEALTH EXPENDITURE AND FINANCING
Health expenditure per capita
Health expenditure in relation to GDP
Health expenditure by function
Financing of health care
Expenditure by disease and age
Capital expenditure in the health sector
international law.

The amount that each country spends on health, for both this level of spending growth has been constant since 2009.
individual and collective services, and how this changes Preliminary estimates for 2014 point towards a slight slow-
over time can be the result of a wide array of social and eco- down in health spending in Japan, after recent strong
nomic factors, as well as the financing and organisational growth.
structures of a country's health system. In the United States, health spending grew by 1.5% in 2013,
In 2013, the United States continued to outspend all other less than half the average annual growth rate prior to 2009.
OECD countries by a wide margin, with the equivalent of The latest forecasts from the Centers for Medicare and
USD 8 713 for each US resident (Figure 9.1). This level of Medicaid Services point to faster growth in 2014 as more
health spending is two-and-a-half times the average of all Americans gain health insurance coverage (Keehan et al.,
OECD countries (USD 3 453) and nearly 40% higher than the 2015).
next biggest spender, Switzerland (adjusted for the differ- Canada has seen a sustained period of low growth since
ent purchasing powers see Definition and comparability 2010. This is in contrast to the average 3.5% growth per year
box). Compared with some other G7 countries, the United between 2005 and 2009. With health spending growth esti-
States spends around twice as much on health care per mated to have continued below economic growth, health
person as Germany, Canada and France. Countries spend- spending as a share of GDP has also declined from a high of
ing less than half the OECD average include many of the 10.6% in 2009 to 10.2% in 2013.
central European members of the OECD, such as Hungary
and Poland, together with Chile. The lowest per capita
spenders on health in the OECD were Mexico and Turkey
with levels of less than a third of the OECD average. Outside
of the OECD, among the key partner countries, China and
India spent 13% and 4% of the OECD average on health in
Expenditure on health measures the final consump-
per capita terms in 2013.
tion of health goods and services (i.e. current health
Figure 9.1 also shows the breakdown of per capita spending expenditure). This includes spending by both public
on health into public and private sources (see the indicator and private sources on medical services and goods,
on Financing of health care). In general, the ranking public health and prevention programmes and
according to per capita public expenditure remains compa- administration, but excludes spending on capital for-
rable to that of total spending. Even if the private sector in mation (investments).
the United States continues to play the dominant role in
To compare spending levels between countries, per
financing, public spending on health per capita is still
capita health expenditures are converted to a com-
greater than that in all other OECD countries, with the
mon currency (US dollar) and adjusted to take account
exception of Norway and the Netherlands.
of the different purchasing power of the national cur-
Per capita spending on health across the OECD edged up rencies. Economy-wide (GDP) PPPs are used as the
slightly in 2013 continuing a trend of recent years. This slow most available and reliable conversion rates.
rise comes after health spending growth ground to a halt in
For the calculation of growth rates in real terms, econ-
the wake of the global financial and economic crisis. Between
omy-wide GDP deflators are used. In some countries
2009 and 2013, average annual health spending growth across
(e.g. France and Norway) health-specific deflators
the OECD was 0.6%, in contrast to the 3.4% in the period
exist, based on national methodologies, but these are
between 2005 and 2009 (Figure 9.2). There has been a differ-
not used due to limited comparability.
ence of health spending growth between Europe and the rest
Note: Ireland is currently implementing a project to report
of the OECD with some European countries facing dramatic
increased detail on health expenditure and financing data in
reductions in health spending from 2010 onwards. accordance with international guidelines. Data for 2013 is
There have been some significant changes in the annual therefore not available and revisions to this and the following
indicators will be made available on completion of the project.
growth rates in health spending in the years before and
during the financial crisis in a number of countries. Annual
increases have been reversed in Greece (5.4% vs. -7.2%) and
Ireland (5.3% vs. -4.0%) and have slowed down in the vast
majority of OECD countries. Only six countries Hungary,
Mexico, Switzerland, Israel, Japan and Chile recorded References
higher average growth following the crisis than pre-2009. Keehan, S.P. et al. (2015), National Health Expenditure Pro-
Chile, Korea and Turkey saw health spending increase by jections, 2014-24: Spending Growth Faster Than Recent
more than 5% in real terms in 2013. For Chile and Korea, Trends, Health Affairs, Vol. 34, No. 8, pp. 1407-1417.

Un
i te
-10
-5
0
5
10
15
0
1 000
2 000
3 000
4 000
5 000
6 000
7 000
8 000
9 000
L u Gr e Sw d St
USD PPP
xe e c it z ate
m e 5.4 er s 8 713
bo -7.2 la
ur
6 325
Ne No nd
-0.4 th r w
-4.3
Ir e g
la er ay
la 5 862
Po nd 5.3 Sw nds
r tu -4.0 5 131
ga 1.3
Ge ede
2. Data refers to 2012.

Sp l -3.0
rm n 4 904
De an
1. Includes investments.
3.5
ain nm y 4 819
-1.7
Annual average growth rate (%)

It a Lu A ar
De l y 0.5 xe u s k 4 553
-1.6
nm m tr i
a
bo a
u 4 553
Ic r k 3.4 Ca rg 2
-0.8 4 371

el
an n
0.4
Sl B e ad
o d lg a 4 351
C v -0.4 iu
Un z e eni a 3.2
F m 4 256
c Au r an
i te h -0.3 st ce
d Re 4 124
Public
K i p. 5.4 ra
li
-0.2
ng
do Ja a 2 3 866
3.6 Ic p a n
Ca m
-0.1 el 3 713
na
d 3.5
Ir e a n d
3 677
Es a
0.3
la
to n
n 3 663

6.7
2005-09
Be ia OE d 2
lg 0.3 Ne Fi CD
iu Un w nl 3 453
m 3.2 i te Ze and
Sl O E 0.5 d al 3 442
K i an
3.4
Note: Expenditure excludes investments, unless otherwise stated.

ov C D
Ne ak R 0.6
ng d
do 3 328
w e 11.3
m 3 235
Z e p. It a
al 0.6

an S ly 3 077
Hu d 4.1 Po p ai
ng 0.6 r tu n 2 898
ar -2.3
1. Mainland Norway GDP price index used as deflator. 2. CPI used as deflator.
Sl g a
Fr y 0.8 ov l 2 514
an en
c 1.5 Is i a 2 511
Au e 0.9 r
st
r 2.2
Gr a e l
ee 2 428
Sw ia
1.0 c
ed C z Ko e 2 366
e 1.7 e r
Tu n 1.0 Sl ch e a 2 275
rk ov Re
5.0
2 040
No ey a k p.
rw 1.2
Ru H u R e p
ay 1 1.9 ss ng . 2 010
Un F in 1.2 i a ar
i te an l n y 1 719
d 1.7 Fe
Private
d
Ne S t a 1.3 d. 1
1 653
th tes 2.3 L i Chi
th le
er
la 1.5 ua 1 606
3.3
nd
9.1. Health expenditure per capita, 2013 (or nearest year)
E s ni a 1 573
2009-13
M s 1.7 to
ex
Au ico 1.7
Po ni a 1 542
1.7
la
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en; WHO Global Health Expenditure Database.
st
r C o Br nd 1 530
Ge a li a 2.9 st a zi
rm 2.0 aR l 1 1 471
an 2.8 ic
P y
So a 1 380
2.0 ut L at 1
S w ola h vi
8.4 Af a 1 216
i t z nd
er 2.3 r
la M ic a 1 1 121
nd 1.3 ex
2.5 ic
Is 1 048
ra 1.9
Co Turk o
e lo ey
Ja l 3.6 m 941
bi
pa
n 3.2 a 864
Ko 3.9 In C h i 1
do n a 1
re 9.0 649
9.2. Annual average growth rate in per capita health expenditure, real terms, 2005 to 2013 (or nearest years)
a ne
5.4 si
Ch
In a 1 293
il e 2 5.9 di
6.4 a1 215
165
1 2 http://dx.doi.org/10.1787/888933281252
1 2 http://dx.doi.org/10.1787/888933281252
The change in how much a country spends on healthcare The United States has seen its health spending to GDP ratio
in relation to spending on all the other goods and services remain consistent at 16.4% since 2009, in contrast to the
in the economy can depend on both fluctuations in the rate earlier steep rise whereby the share increased almost two
of health spending itself as well as growth in the economy percentage points between 2005 and 2009. Canada also
as a whole. The 2000s were characterised by a period of experienced a steady rise through the second half on the
health spending growth above that of the overall economy 2000s to reach a peak in 2009. Since then, with health
so that health expenditure as a share of GDP rose sharply in spending growth lower than economic growth, the share of
many OECD countries. However, the economic crisis that GDP has gradually decreased. Japan, on the other hand, has
took hold in 2008 resulted in an initial rise followed by a seen its health spending share of GDP rise steadily from the
reduction in the health spending to GDP ratio across many OECD average in 2005 to continue increasing to more than
OECD countries. 10% of GDP by 2013 as a result of a deliberate policy to
Health spending accounted for 8.9% of GDP (excluding increase public spending on health.
investment) on average across OECD countries in 2013, In Europe, France and Germany also have seen their health
unchanged from 2012 and up marginally from 8.8% in 2011 spending to GDP ratio stabilise since 2009 as health spend-
(Figure 9.3). Including capital spending (see the indicator ing growth has aligned with economic growth. Other Euro-
on Capital expenditure in the health care sector), expen- pean countries, such as Portugal and Ireland saw health
diture on health as a share of GDP is estimated to have been spending growth decline much more than GDP, resulting in
9.3% on average in 2013. a rapidly decreasing health spending to GDP ratio, after sig-
In 2013, the United States spent 16.4% of GDP on health, nificant increase prior to 2009, as health spending signifi-
remaining well above the OECD average and more than five cantly outpaced economic growth. Greece, where there
percentage points above a group of high-income countries all have been significant cuts in health spending, has seen the
at around 11%, which include the Netherlands, Switzerland, health spending to GDP ratio fluctuate but overall remain at
Sweden, Germany and France. Almost half of OECD coun- a similar level to the mid-2000s as the overall economy has
tries spend in a band between 8 and 10% of GDP on health suffered to the same extent.
services. Among OECD countries, Mexico and Estonia
devoted around 6% of GDP to health around two-thirds of
the OECD average, while Turkey reported the lowest share Definition and comparability
at just over 5% of GDP. Among the key partner countries,
China and India spent 5.6% and 4.0% of GDP respectively See indicator Health expenditure per capita for a
in 2013, while Brazil (9.1%) and South Africa (8.9%) spent definition of expenditure on health.
close to the OECD average (all including investment). Gross domestic product (GDP) = final consumption +
The health spending to GDP ratio jumped sharply in 2009 to gross capital formation + net exports. Final consump-
reach 9.0% on average up from 8.3% in 2008 as overall eco- tion of households includes goods and services used
nomic conditions rapidly deteriorated but health spending by households or the community to satisfy their indi-
continued to grow or was maintained in many countries. vidual needs. It includes final consumption expendi-
In the subsequent context of reducing public deficits, the ture of households, general government and non-
subsequent reductions in (public) spending on health profit institutions serving households.
have resulted in the share of GDP first falling and since In countries, such as Ireland and Luxembourg, where
stabilising as health expenditure growth has become a significant proportion of GDP refers to profits
aligned to economic growth in many OECD countries exported and not available for national consumption,
(Figures 9.4, and 9.5). GNI may be a more meaningful measure than GDP.

9.3. Health expenditure as a share of GDP, 2013 (or nearest year)

Public Private
% GDP
18
16.4
16
14
11.1
10.9
12
11.0
11.0
11.1
10.4
10.2
10.2
10.2
10.1
9.9
9.5
10
9.2
9.1
9.1
8.9
8.9
8.9
8.8
8.8
8.8
8.7
8.7
8.6
8.5
8.1
7.6
7.5
7.4
8
7.3
7.1
6.9
6.8
6.6
6.5
6.4
6.2
6.1
6.0
5.6
5.3
6
5.1
4.0
4
2.9
2
0
Ze a 2
Af l 2
No ic a 2
li a 1
ak d 1
Ru e m b b i a 2
1
P o d. 2
La a 2
ne 2
a2
ut r a l
S w l and
C a ium
S w er l es
er s
h e
rm n
Fr ny
nm e
Ja r k
Be pan
Co Au da
Ne t a tr ia
Gr n d
rt e
ay
st D
Sp l y
Ic ain
ov d
i te F eni a
ng d
I s p.
ng l
Cz C r y
C o Ko .
x m a
M and
th o
E s ni a
Ch i a
Tu via
In In d y
S l Ir e l m
So B uga
Hu r ae
ia rg
do i a
i t z and
e c hil
e
De anc
Po eec
Lu lo re
Ge ede
Sl l an
K i an
L i ex i c
h zi
Au OEC
n
It a
Re
Re
do
in
si
w Ric
a
ov an
rk
th t at
rw
a
na
Fe
s s ou
t
to
ua
ra
s
d inl
l
al
r
lg
e
Ne d S
n
s
i te
Un
Un
Note: Excluding investments unless otherwise stated.
1. Data refers to 2012.
2. Including investments.
1 2 http://dx.doi.org/10.1787/888933281263
9.4. Health expenditure as a share of GDP, selected G7 9.5. Health expenditure as a share of GDP, selected
countries, 2005-13 European countries, 2005-13
Canada France Estonia Greece
Germany Japan Ireland Portugal
United States OECD34 Spain OECD34
% GDP % GDP
18 12
16
10
14
12 8
10
6
8
6 4
2005 2007 2009 2011 2013 2005 2007 2009 2011 2013

Spending on inpatient care and outpatient care combined the H1N1 influenza epidemic, which led to significant one-
covers the major part of health expenditure across OECD off expenditure for vaccination in many countries around
countries almost two-thirds of current health expendi- 2009.
ture on average in 2013 (Figure 9.6). A further 20% of health While spending on long-term, outpatient and inpatient
spending was allocated to medical goods (mainly pharma- care have continued to grow, the rates have also signifi-
ceuticals), while 12% went towards long-term care and the cantly reduced since 2009. Expenditure growth for outpa-
remaining 6% on collective services, such as public health tient care was reduced by more than half overall (1.7% vs.
and prevention services as well as administration. 3.9%), but has still remained positive in three quarters of
Greece has the highest share of spending on inpatient care OECD countries. Some governments decided to protect
(including day care in hospitals) among OECD countries: it expenditure for primary care and front-line services whilst
accounted for 42% of total health spending in 2013, up from looking for cuts elsewhere in the health system. The
36% in 2009, as a consequence of larger decreases in spend- annual average growth rate for hospital care dropped to a
ing for outpatient care and pharmaceuticals. In Poland, quarter of its previous growth rate, down from 2.4%, and
France and Austria, the hospital sector also plays an impor- was negative between 2009 and 2013 in a dozen OECD
tant role, with inpatient spending comprising more than a countries. Reducing wages in public hospitals, postponing
third of total costs. While the United States consistently staff replacement and delaying investment in hospital
reports the highest share of outpatient care (and by conse- infrastructure were among the most frequent measures
quence the lowest inpatient share), it should be noted that taken in OECD countries to balance health budgets.
this figure includes remunerations of physicians who inde-
pendently bill patients for hospital care. Other countries
with a high share of outpatient spending include Portugal
and Israel (48% and 46%).
The other major category of health spending is medical The System of Health Accounts (OECD, 2000; OECD,
goods. In the Slovak Republic and Hungary, medical goods Eurostat and WHO, 2011) defines the boundaries of
represent the largest spending category at 36% and 33% of the health care system. Current health expenditure
all health expenditure, respectively. With around 30%, the comprises personal health care (curative care, rehabil-
share is also high in Greece and Mexico. In Denmark and itative care, long-term care, ancillary services and
Norway, on the other hand, spending on medical goods medical goods) and collective services (prevention
represents only 10-11% of total health spending. and public health services as well as health adminis-
There are also differences between countries in their tration). Curative, rehabilitative and long-term care
expenditure on long-term care (see the indicator on Long- can also be classified by mode of production (inpa-
term care expenditure in Chapter 11). Countries such as tient, day care, outpatient and home care). Concern-
Norway, the Netherlands, Sweden and Denmark which ing long-term care, only the health aspect is normally
have established formal arrangements for the elderly and reported as health expenditure, although it is difficult
the dependent population, allocate around a quarter or in certain countries to separate out clearly the health
more of total health spending to long-term care. In many and social aspects of long-term care. Some countries
southern or central European countries with a more infor- with comprehensive long-term care packages focus-
mal long-term care sector, the expenditure on formal long- ing on social care might be ranked surprisingly low
term care services accounts for a much smaller share of based on SHA data because of the exclusion of their
total spending. social care. For example, an ongoing review of Japanese
long-term care boundaries concerning SHA will likely
The slowdown in health spending experienced in many
lead to a significant increase in health spending based
OECD countries in recent years has affected all spending
on SHA2011 to be released in 2016. Thus, estimations
categories, but to varying degrees (Figure 9.7). Expenditure
of long-term care expenditure are one of the main fac-
for pharmaceuticals has been cut annually by nearly 2%
tors limiting comparability across countries.
after recording positive annual increases of 2% in the pre-
crisis years still down on previously strong growth in phar-
maceutical spending in the 1990s and early 2000s (see the
indicator on Pharmaceutical expenditure in Chapter 10). References
Despite initially ring-fencing and protecting public health
OECD (2000), A System of Health Accounts, OECD Publishing,
budgets, prevention spending turned negative in around
Paris, http://dx.doi.org/10.1787/9789264181809-en.
half of OECD since 2009. Overall, spending on preventive
care contracted by -0.3% on an annual basis, after recording OECD, Eurostat and WHO (2011), A System of Health Accounts,
very high growth rates during the period 2005-09 (5.6%). 2011 Edition, OECD Publishing, Paris,
Part of the reversal in spending growth can be explained by http://dx.doi.org/10.1787/9789264116016-en.

9.6. Current health expenditure by function of health care, 2013 (or nearest year)
% Inpatient care* Outpatient care** Long-term care Medical goods Collective services
100 3
4 5 5 6 5 4 6 4 5 4 5 6 6 4 5 5 5 7
11 8 8 8 8 10 8 8
14
90 12 10 11
15 16 13
20 16 13 12
22 16 22 20 16
13 23 24 23 20 20
80 32 24 20 22
33
2 8 36
6 23 24
70 5 9 15 19 29 30
4 9 20 25
6 9 12 26
12 14 22
1 10 14 14
60 4
21
50 48 46 40 37 36 30
32 29 25
52 37 35 34 33 33 30 28 30 30 28 25 22
40 35 34 34 34 35
30
20 42
36 35 35
30 31 32 30 32
26 26 28 28 28 28 28 30 29 28 29 29
25 24 23 22 22
10 18 21
0
l
nd
el
en
da
s
p.
nd
ria
ce
a in
27
ce
ia
p.
ay
m
ga
es
nd
an
ni
re
ar
ar
an
an
pa
ur
ic
ra
en
Re
Re
iu
rw
ee
an
ed
na
la
la
CD
st
at
r tu
to
Sp
ex
Ko
nm
ng
bo
rm
nl
Ja
el
la
Is
lg
ov
Po
er
Au
Gr
Fr
Sw
Ca
St
No
Es
ak
M
Ic
er
Fi
OE
Be
Hu
m
Po
it z
De
Ge
ec
Sl
ov
th
d
xe
Cz
Sw
i te
Ne
Sl
Lu
Un
Note: Countries are ranked by curative-rehabilitative care as a share of current expenditure on health. * Refers to curative-rehabilitative care in
inpatient and day care settings. ** Includes home-care and ancillary services.
1. Inpatient services provided by independent billing physicians are included in outpatient care for the United States.
1 2 http://dx.doi.org/10.1787/888933281277
9.7. Growth rates of health spending for selected functions per capita, OECD average, 2005-13
2005-09 2009-13
Average annual growth rates in real terms, %
7
6 5.9
5.6
4 3.9 3.8
3 2.7
2.4
1.9
2 1.7
1 0.7 0.8
0
-0.3
-1
-2 -1.8
-3
Inpatient care Outpatient care Long-term care Pharmaceuticals Prevention Administration

1 2 http://dx.doi.org/10.1787/888933281277

Across all OECD countries, health care is financed by a mix Private health insurance (PHI) can play different roles in
of public and private spending. In some countries, public health systems. Whereas PHI provides primary health care
health spending is mostly confined to spending by the gov- coverage for large population groups in the United States
ernment using general revenues. In others, social insur- and Chile, it complements or supplements public coverage
ance funds finance the bulk of health expenditure. Private for the vast majority of the population in countries such as
financing of health care consists mainly of payments by France, Belgium and Slovenia. In other countries, such as
households (either as standalone payments or as part of Australia and Ireland, it serves as duplicate insurance pro-
co-payment arrangements) as well as various forms of pri- viding access to a larger group of providers. Spending for
vate health insurance. PHI accounts for only 6% of overall health spending in the
In nearly all OECD countries, the public sector is the main OECD, but it represents a sizeable share in a number of
source of health care financing. Around three-quarters of countries, particularly in the United States (35%) and Chile
health care spending was publicly financed in 2013 (20%). While health spending growth through private
(Figure 9.8). In Denmark, Sweden and the United Kingdom, health insurance slowed down significantly in the period
central, regional or local governments financed more than 2009-11, spending grew by 2.9% between 2011 and 2013
80% of all health spending. In the Czech Republic, France, also as a response to some cost-shifting and loss of cover-
Luxembourg, Japan and Germany, social health insurance age in some countries.
financed 70% or more of all health expenditure. Only in
Chile and the United States was the share of public spend-
ing on health below 50%. In these countries, a great propor-
tion of health spending is financed either directly by Definition and comparability
households (Chile) or by private insurance (United States).
The financing of health care can be analysed from the
Health care is competing for public resources with different
point of view of the sources of funding (households,
sectors such as education, defence and housing. The size of
employers and the state), financing schemes (e.g.
the public budget allocated to health is determined by a
compulsory or voluntary insurance) and financing
number of factors including the type of health and long-
agents (organisations managing the financing
term care system, the demographic composition of the
schemes). Here financing is used in the sense of
population and the relative budget priorities. On average,
financing schemes as defined in the System of Health
15% of total government expenditure was dedicated to
Accounts (OECD, 2000; OECD, Eurostat and WHO, 2011).
health care in 2013 (Figure 9.9). There are, however, impor-
Public financing includes expenditure by the general
tant variations across OECD member states. Whereas a
government and social security funds. Private financ-
fifth of government spending is allocated to health care in
ing covers households out-of-pocket payments, pri-
countries such as New Zealand and Switzerland, this falls
vate health insurance and other private funds (NGOs
to around 10% in Hungary and Greece.
and private corporations). Out-of-pocket payments
Developments in overall health spending are largely driven are expenditures borne directly by patients. They
by the trends in public spending. Strong pre-crisis growth include cost-sharing and, in certain countries, estima-
resulted in average public expenditure on health increasing tions of informal payments to health care providers.
at an annual rate of almost 4% (Figure 9.10). In 2010, growth
Total government expenditure is used as defined in
in public health spending came to a halt with reductions in
the System of National Accounts and includes as
many countries. Since then spending growth has been very
major components intermediate consumption, com-
slow, often in line with overall economic growth.
pensation of employees, interest, social benefits,
After public financing, the main source of funding tends to social transfers in kind, subsidies, other current
be out-of-pocket payments. On average private households expenditure and capital expenditure payable by cen-
directly financed 19% of health spending in 2013. The share tral, regional and local governments as well as social
of out-of-pocket payments was above 30% in Mexico, Korea, security funds.
Chile and Greece and 10% or lower in France and the United
Kingdom. Out-of-pocket spending has continued to grow
since 2009, albeit at a slower rate, partly as a result of cost-
sharing measures introduced in a number of countries.
References
Measures taken include increasing co-payments and rais-
ing reimbursement thresholds for pharmaceuticals, reduc- OECD (2000), A System of Health Accounts, OECD Publishing,
ing benefits for dental treatment, increasing user charges Paris, http://dx.doi.org/10.1787/9789264181809-en.
for hospital care, introducing cost-sharing for certain activ- OECD, Eurostat and WHO (2011), A System of Health Accounts,
ities in primary care and removing entitlements for public 2011 Edition, OECD Publishing, Paris, http://dx.doi.org/
coverage for particular groups of the population. 10.1787/9789264116016-en.

9.8. Expenditure on health by type of financing, 2013 (or nearest year)

General government Social security Private out-of-pocket Private insurance Other
% of current health expenditure
100 2 1 2 5 4 1
4 5 3 4
6 3 5 9 5 2 6 4 7 6
15 14 3
5 14 15 15 15 13 9 13 11
90 10 14 12 18 22 22 20
13 17 18 17 23
13 19 24
7 19 24 31
80 11 27 26 28
13 14 20 17 35
8 23 37 45
70 5 1
28 14 1
33
60
45 36 12
50 80
58
78 73 74 66 67 40 4
84 84 83 75 70 47
40 77 68 61 47
74 72 68 56 29
67 69 68 67 65 45
30 61
53
48
20 42
37
31
25 22
10 20 19 16
10 9 11 11 10 9 11
7 6 7 7
4 3
0
i t z gal
No s
ng n
Es m
Po a
Po nd
Gr d
Hu c e
m n
Ze
Be ey
a
Ge t al y
y
el
Un M a

Au y
F ia
ak d
O E p.
34
Sl a in
Ca ia
Au and
Ir e l i a
St o
il e
De ay
ec rk
i t e S w p.
Lu Ja
Ic r g
Fr d
Tu e
w nd
es
m
an
d
ni
re
c
ar
n
K i de
ov n
xe p a
an
c
ra
en
r
nd
Re
Re
iu
C z ma
rk
rw
an
ee
Ch
ra
na
i t e ex i
la
Sl inl a
st
CD
do
to
Sw r tu
Sp
at
Ko
ng
la
Ne ela
bo
rm
I
e
l
al
Is
lg
ov
er
la
st
n
h
er
d
th
Ne
d
Un
1. The Netherlands report compulsory cost-sharing in health care insurance and in Exceptional Medical Expenses Act under social security rather
than under private out-of-pocket, resulting in an underestimation of the out-of-pocket share.
2. Data refer to total health expenditure (= current health expenditure plus capital formation).
3. Social security reported together with general government.
9.9. Health expenditure as share of total government 9.10. Growth of health spending by financing, OECD
expenditure, 2013 (or nearest year) average, 2005-2013
New Zealand 22
2005-07 2007-09 2009-11 2011-13
Switzerland 22
Netherlands 21
United States 20
Japan 20
19 3.7
Germany
Canada 18
Sweden 17 General government/ 4.3
Norway 17
Social security
Australia 16 0.2
United Kingdom 16
Iceland 16 0.5
Denmark 15
Austria 15
France 15
OECD34 15
Belgium 15
2.8
Chile 14
Spain 14
0.7
Czech Rep. 14
Private out-of-pocket
Slovak Rep. 14
0.7
Italy 13
Ireland 13
Luxembourg 12 1.2
Korea 12
Portugal 12
Estonia 12
Mexico 11
Finland 11 4.9
Israel 11
Poland 11 4.0
Turkey 10 Private insurance
Slovenia 10 1.8
Greece 10
Hungary 10
2.9
0 5 10
15 20 25
% total government expenditure
1. Data refer to total health expenditure (= current health expenditure 0 2 4 6
plus capital formation). Annual growth rates per capita in real terms (%)
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en;
OECD National Accounts; Eurostat Statistics Database; IMF World
1 2 http://dx.doi.org/10.1787/888933281280
Economic Outlook Database.
1 2 http://dx.doi.org/10.1787/888933281280

Attributing health care expenditure by disease and age is Figure 9.13 shows that the share of spending increases with
important for health policy makers in order to analyse age after an initial peak of spending linked to birth and
resource allocations in the health care system. This infor- early childhood illnesses. The share of current health
mation can also play a role in assessing the impact of pop- spending remains relatively constant until around the 50 to
ulation ageing and changing disease patterns on spending. 54 age group before increasing sharply as people grow
Furthermore, the linking of health expenditures by disease older. As a result, a significant share of current health
to appropriate measures of outputs (e.g. hospital dis- spending is consumed by elderly population. Those aged 65
charges by disease) and outcomes (e.g. survival rates after and above consume around 60% of the current health
heart attack or cancer) helps in monitoring the perfor- spending on average in all three countries. In addition, in
mance of health care systems at a disease-based level Korea and the Netherlands more than 20% of current
(Heijink et al., 2006). health spending is accounted for by those aged 85 years
Figure 9.11 shows the distribution of hospital inpatient and above, while in the Czech Republic the share is much
expenditure according to seven main diagnostic categories. lower. This may be explained by a lower level of long-term
These categories account for between 60% and 80% of all care spending in Czech Republic.
inpatient acute care expenditure across the group of coun-
tries. Circulatory diseases account for the highest share of
inpatient spending in each of the countries except for
Korea and the Netherlands, where spending on cancer and Definition and deviations
mental and behavioural disorders is the largest category,
respectively. The differences between countries can be Expenditure by disease and age allocates current
influenced by a number of factors, including demographic health expenditure by patient characteristics. Guide-
structure and disease patterns, as well as institutional lines developed propose disease categories according
arrangements and clinical guidelines for treating different to ICD-10. To ensure comparability between countries,
diseases. For example, in the Netherlands, mental and expenditures are also linked to the System of Health
behavioral disorders account for around 23% of all inpa- Accounts (SHA) framework and a common methodol-
tient spending around twice the level as that of Germany, ogy is proposed advocating primarily a top-down allo-
Finland and Japan. This may be partly explained by the cation of expenditures based on principal diagnosis.
large number of acute mental health hospitals with very The main comparability issues relates to the treat-
long average lengths of stay (OECD, 2015). Similarly, longer ment of non-allocated and non-disease-specific
than average lengths of stay in Japan for some of the spe- expenditures. In the former case this is due to data
cific circulatory diseases such as cerebrovascular disease limitations (often in outpatient and pharmaceutical
(stroke) might explain why more than 22% of hospital inpa- expenditure) and in the latter case mainly prevention
tient expenditures are allocated to the treatment of circula- and administration expenditure.
tory diseases. Discharges related to circulatory diseases Note that the charts cover allocated spending only
only account for 12% of all discharges in Japan a propor- and the following country limitations apply. Canada
tion similar to other countries. excludes Quebec and mental health hospitals; the
Figure 9.12 compares expenditure by hospital discharge for Czech Republic refers to expenditure by the Health
circulatory diseases and cancers. Generally, the cost per Insurance Fund only; Germany refers to total hospital
discharge between these two main disease categories is expenditure; and the Netherlands refers to curative
similar in all countries, apart from Japan where spending care in general and specialty hospitals.
per discharge for circulatory diseases is more than twice
that of cancer. Japan has the highest expenditure per dis-
charge compared to the other countries for circulatory dis-
ease, again due to the much longer lengths of stay, while
References
the Netherlands has the highest expenditure per discharge Heijink, R., M.A. Koopmanschap and J.J. Polder (2006), Inter-
for cancer treatment. national Comparison of Cost of Illness, RIVM, Bilthoven.
Different cost patterns can also be due partly to demographic OECD (2015), Addressing Dementia: The OECD Response, OECD
factors. The allocation of current health spending by age Health Policy Studies, OECD Publishing, Paris,
group in the Czech Republic, Korea and the Netherlands in http://dx.doi.org/10.1787/9789264231726-en.

9.11. Share of hospital inpatient expenditures by main diagnostic category, 2011 (or nearest year)
Circulatory diseases Cancer Injury, poisoning and other consequences of external causes
% Mental and behavioural disorders Respiratory system Digestive diseases Diseases of musculoskeletal system and connective tissue
25
20
15
10
0
Canada Czech Rep. Finland Germany Hungary Japan Korea Netherlands Slovenia Sweden Switzerland
(2008) (2011) (2010) (2008) (2006) (2010) (2009) (2011) (2011) (2011) (2011)
Source: OECD Expenditure by Disease, Age and Gender Database.

1 2 http://dx.doi.org/10.1787/888933281298
9.12. Expenditure per hospital discharge for two 9.13. Share of current health spending by age group, 2011
diagnostic categories, 2011 (or nearest year) (or nearest year)
Circulatory disease Cancer Czech Rep. Korea Netherlands
%
30
8 558
Switzerland 10 925
5 753
Sweden 6 444
4 773
Slovenia 5 162
11 962
20
Netherlands 14 809
4 742
Korea (2009) 5 081
Japan (2010) 17 366

8 001
1 817 10
Hungary (2006) 1 586
5 616
Germany (2008) 5 819
4 881
Finland (2010) 4 847
Czech Rep. 2 909 0

2 406
25 4
55 4
75 4
80 9
+
4
10 9
15 4
20 9
30 9
35 4
40 9
45 4
50 9
60 9
65 4
70 9
4
-7
-1
-1
-2
-7
-3
-4
-5
-6
5-
-2
-3
-4
-5
0-
-6
-8
85
5 800
Canada (2008) 5 776
0 5 000 10 000 15 000
1 2 http://dx.doi.org/10.1787/888933281298
USD PPP
1 2 http://dx.doi.org/10.1787/888933281298

Knowing how much a health system is investing in hospi- In parallel with current health spending, capital spending
tals, medical technology and other equipment is very rele- has been affected by the global economic crisis with out-
vant for policy making and analysis. Although health lays on health system infrastructure and equipment
systems remain a highly labour-intensive sector, capital often being a prime target for reduction or postpone-
has been increasingly important as a factor of production ment. Overall, capital spending grew strongly in the
of health services over recent decades. This is illustrated, period up to 2008 annual capital expenditure was 22%
for example, by the growing importance of diagnostic and higher than in 2005 in real terms on average. During the
therapeutic equipment or the expansion of information, next three years, the annual outlay fell back by almost
computer and telecommunications technology in health 15%. Since 2011, there has been a return to growth in capi-
care over the last few years. The availability of statistics on tal spending (Figures 9.15 and 9.16).
capital expenditure is essential to the analysis of the health The country differences also mirror the trends in current
systems production capacity (that is, whether capacity is spending. Outside of Europe, investment in the health sec-
appropriate, deficient or excessive), which is needed in tor has been generally less affected by the economic down-
turn to inform policy implementation (for example, if turn. Australia and Korea, for example, report capital
excess capacity exists, the marginal cost of expanding cov- spending more than 40% higher in 2013 compared with
erage will be lower than if the health care system is already 2005.
straining to fill current demand).
A number of European countries have seen severe reduc-
On average, OECD countries invested around 0.45% of their tions in capital spending. Figures for Greece show that the
GDP in 2013 in terms of capital spending in the health sec- outlay was less than 40% of the 2005 level in 2013, with an
tor. This compares with 8.9% of GDP on average across the acceleration of the fall in 2010. Similarly, Spain experienced
OECD for current spending on health care services and med- a sharp reversal after 2008, with capital spending in 2012 at
ical goods (see the indicator on Health expenditure in rela- half the level of 2005.
tion to GDP). As with current spending, there are both
differences in the current levels of investment expenditure
between countries and in the recent trends observed.
At the higher end of the scale, Belgium spent more than
0.8% of GDP on capital investment in 2013, followed by a Gross fixed capital formation in the health care sys-
group of countries, including France, Germany and the tem is measured by the total value of the fixed assets
United States, all spending more than 0.6% of GDP. Around that health providers have acquired during the
half the OECD countries are in a relatively narrow band of accounting period (less the value of the disposals of
plus or minus 25% of the average ranging from the United assets) and that are used repeatedly or continuously
Kingdom to Australia. At the lower end, Turkey, Chile and for more than one year in the production of health
Hungary spent around half the OECD average, while services. The breakdown by assets includes infra-
Greece, Iceland and Mexico spent around 0.1% of GDP on structure (e.g. hospitals, clinics, etc.), machinery and
capital infrastructure and equipment in the health care equipment (including diagnostic and surgical machin-
sector. ery, ambulances, and ICT equipment), as well as soft-
Data from National Accounts provides an idea of the type ware and databases.
of assets and capital spending. While capital spending can Gross fixed capital formation is reported by many
fluctuate from year to year, overall in the health sector countries under the System of Health Accounts. It is
there is an even split between spending on construction also reported under the National Accounts broken
(i.e. building of hospitals and other health care facilities) down by industrial sector according to the Interna-
and spending on equipment (medical machinery, ambu- tional Standard Industrial Classification (ISIC) Rev. 4
lances, as well as ICT equipment). Together they account using Section Q: Human health and social work activ-
for 85% of capital expenditure. The remaining 15% is ities or Division 86: Human health activities. The for-
accounted for by intellectual property products the result mer is normally broader than the SHA boundary while
of research, development or innovation. This can vary sig- the latter is narrower.
nificantly between countries.

9.14. Gross fixed capital formation in the healthcare sector as a share of GDP, 2013 (or nearest year)
Public Private Total (no breakdown)
% GDP
1.0
0.8
0.6
0.4
0.2
0
m1
d2
Au ny 1
s1
ly 1
S l p. 1
Ic e 1
d1
No l
es
el
nd
ria
ak a
Ze rk
Ge c e
S li a
er n
Tu .
ey
Hu il e
Gr r y
o
Ca d
da
Es y
a
CD
ec nd
m a
g
m
ga
p
a in
a
ni
xe ni
e
th de
an
ic
ur
d sr a
nd
Re
c
do
a
an
an
a
rk
at
rw
an
Sl Kor
Ch
ra
na
It a
Cz la
la
Re
st
iu
OE
r tu
to
Lu ve
ex
ee
N e enm
ng
bo
Ne we
nl
Sp
St
Po
Ir e
rm
st
la
el
ng
al
Au
lg
Fr
I
o
M
Fi
h
Po
Be
Ki
ov
D
i te
w
Un
i te
Un
1. Refers to gross fixed capital formation in ISIC 86: Human health activities (ISIC Rev. 4).
2. Refers to gross fixed capital formation in ISIC Q: Human health and social work activities (ISIC Rev. 4).
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en; OECD National Accounts Database.
1 2 http://dx.doi.org/10.1787/888933281305
9.15. Gross fixed capital formation, selected 9.16. Gross fixed capital formation, selected European
non-European countries, 2005-13 countries, 2005-13
Australia Canada Korea Ireland Greece Portugal
United States OECD34 Spain OECD34
2005 = 100 2005 = 100
160 160
140 140
120 120
100 100
80 80
60 60
40 40
20 20
2005 2007 2009 2011 2013 2005 2007 2009 2011 2013

10. PHARMACEUTICAL SECTOR
Pharmaceutical expenditure
Financing of pharmaceutical expenditure
Pharmacists and pharmacies
Pharmaceutical consumption
Share of generic market
Research and development in the pharmaceutical sector
international law.

Pharmaceuticals play a vital role in the health system and ing up overall health expenditure (Figure 10.2). Average real
policy makers must balance the access of patients to new annual growth in pharmaceutical spending outpaced over-
effective medicines with limited health care budgets, while all health spending growth more than 5% on average each
providing the right incentives to manufacturers to develop year between 1990 and 2004, compared with average health
new generations of drugs. After inpatient and outpatient spending growth of less than 4% per year. However, in the
care, pharmaceuticals represent the third largest expendi- second half of the 2000s there was a significant drop in
ture item of health care spending and accounted for more average pharmaceutical spending growth which then
than a sixth (17%) of health expenditure on average across intensified following the global economic crisis. In this
OECD countries in 2013, not taking into account spending period, policy makers in many OECD countries were con-
on pharmaceuticals in hospitals. cerned about reining in public pharmaceutical spending in
The total retail pharmaceutical bill across OECD countries an effort to limit total public spending (see Indicator
was around USD 800 billion in 2013. However, there are Financing of pharmaceutical expenditure). Thus, a num-
wide variations in pharmaceutical spending per capita ber of countries introduced a series of measures: price cuts
across countries, reflecting differences in volume, patterns (achieved through negotiations with the pharmaceutical
of consumption and pharmaceutical prices (Figure 10.1). manufacturers, introduction of reference pricing, applica-
With more than USD 1 000 in 2013, the United States spent tion of compulsory rebates, decrease of pharmacy margins,
far more on pharmaceuticals than any other OECD country reductions of the value added tax applicable for pharma-
on a per capita basis, and double the OECD average. Japan ceuticals), promoting the use of generics, reduction of
(USD 752), Greece (USD 721) and Canada (USD 713) also package sizes, reduction in coverage (excluding pharmaceuti-
spent significantly more on medicines than other OECD cals from reimbursement) and increases in co-payments by
countries. At the other end of the scale, Denmark (USD 240) households.
had relatively low spending levels, less than half the aver-
age across OECD countries. It is important to note that
these figures refer only to retail pharmaceuticals, that is,
pharmaceuticals dispensed directly to patients with a
medical prescription or over-the-counter purchases. Phar- Pharmaceutical expenditure covers spending on pre-
maceuticals can also be administered to patients when scription medicines and self-medication, often
they are in hospital, but these are not taken into account referred to as over-.the-counter products. In some
here. Figures available for a small number of OECD coun- countries, other medical non-durable goods are also
tries suggest that this can add another 10-20% on average included. Pharmaceuticals consumed in hospitals and
to the retail spending, but can vary according to different dis- other health care settings are excluded. Final expendi-
pensing and budgetary practices (Belloni et al., forthcoming). ture on pharmaceuticals includes wholesale and retail
Around 80% of total retail pharmaceutical spending is for margins and value-added tax. It also includes phar-
prescribed medicines; the rest being spent on over-the- macists remuneration when the latter is separate
counter (OTC) medicines. OTC medicines are pharmaceuti- from the price of medicines. Total pharmaceutical
cals that can be bought without prescription and their costs spending refers in most countries to net spending,
are generally borne by patients. In some cases, however, i.e. adjusted for possible rebates payable by manufac-
OTC drugs can also be reimbursed by public payers. turers, wholesalers or pharmacies.
Depending on country-specific legislation, OTC pharma-
ceuticals can be sold outside of pharmacies, for example, in
supermarkets, other retail stores or via the internet. In Aus-
tralia, Spain and Poland, the share of OTC medicines is rel- References
atively high in the latter case accounting for half of Belloni, A., D. Morgan and V. Paris (forthcoming), Pharma-
pharmaceutical spending. ceutical Expenditure and Policies: Past Trends and
During the 1990s and early 2000s, increasing spending on Future Challenges, OECD Working Paper, OECD Publish-
retail pharmaceuticals acted as a major contributor in driving, Paris.

10.1. Expenditure on pharmaceuticals per capita, 2013 (or nearest year)

Prescribed medicines Over-the-counter medicines Total (no breakdown)
USD PPP
1 200
1 026
1 000
800 752
721 713
678 666 652
603 596 590
600 572
536 533 526 515 503
481
459 459
436
397 396 392 381
400 367 364
326
287 273
240
200
0
ce 1
ly 1
p. 1
s1
l1
g2
el 1
es
nd
n
d
da
ce
li a
ria
a in
29
ia
d
en
p.
ay
nd
k
nd
an
re
ni
ar
ar
an
pa
an
ga
en
nd
Re
iu
ur
at
rw
an
ra
ed
na
ra
It a
la
la
Re
CD
st
to
Sp
ee
Ko
nm
ng
la
rm
el
Ja
nl
r tu
lg
ov
bo
St
Po
er
st
la
Is
Au
Fr
Sw
Ca
No
Es
Ir e
h
Gr
Ic
Fi
OE
Be
Hu
ak
er
Au
it z
Ge
De
ec
Sl
m
Po
d
ov
th
i te
xe
Cz
Sw
Ne
Sl
Un
Lu
1. Includes medical non-durables (resulting in an over-estimation of around 5-10%).
2. Excludes spending on over-the-counter medicines.
1 2 http://dx.doi.org/10.1787/888933281318
10.2. Average annual growth in pharmaceutical and total health expenditure per capita, in real terms, average across
OECD countries, 1990 to 2013
Health expenditure (including pharmaceutical spending) Pharmaceutical expenditure

%
8
-2
-4
1990 1995 2000 2005 2010 2013

1 2 http://dx.doi.org/10.1787/888933281318

In all OECD countries, pharmaceuticals are financed by a cant reversals in growth of public pharmaceutical spending
mix of public and private spending. Tax-funded schemes or following the crisis compared to the pre-crisis period. The
social health insurance cover a significant amount of pre- reduction in public spending on pharmaceuticals has not
scribed pharmaceuticals in most countries, sometimes been restricted to Europe. Public spending also came down
complemented by private health insurance. Patients typi- in Canada and Australia (both -2.1%). Japan, on the other
cally have to cover some part of the cost of prescription hand, continues to see substantial annual increases (4.9%).
drugs themselves, although exemptions often exist for vul- Reduction of public pharmaceutical spending in most
nerable segments of the population such as children, the OECD countries was achieved by a wide range of policy
elderly and patients suffering from certain chronic ill- measures (see Indicator Pharmaceutical expenditure),
nesses. Over-the-counter (OTC) pharmaceuticals are nor- including reforms that have aimed to shift some of the
mally financed entirely by private households. burden of pharmaceutical spending away from the public
Pharmaceutical spending represents around 1.4% of GDP purse to private payers. These measures included the de-
on average across OECD countries ranging from 0.5% in listing of products (i.e. excluding them from reimbursement)
Denmark to 2.8% in Greece (Figure 10.3). Public funds repre- and the introduction or increase of user charges for retail
sent slightly less than 60% on average just under 1% of prescription drugs (Belloni et al., forthcoming). In recent
GDP across OECD countries. However, this share is signifi- years, measures of this kind have been taken by around a
cantly higher in Japan (1.5%) and Greece (1.9%) and much dozen OECD countries. Ireland, for example, introduced a
lower in Denmark and Norway (both 0.3%). The proportion 50-cent prescription fee for Medical Card holders in 2010
of private expenditure in GDP is highest in Hungary and the which was subsequently increased. At the same time, the
United States (both 1.3%), and also high in Canada (1.0%). monthly drug reimbursement threshold was raised by 20%
Public protection against the costs of pharmaceuticals is to EUR 120 for non-Medical Card holders, followed by
not as developed as for other health services, such as inpa- subsequent increases. As a result of these policy measures,
tient and outpatient care (Figure 10.4). On average across the share of private financing of pharmaceuticals has
OECD countries, the public sector covered a much higher increased substantially in a number of countries. In Spain,
proportion of the costs of health services (79%) compared 39% of pharmaceutical costs were covered out-of-pocket in
with pharmaceuticals (57%) in 2013. This is true for all 2013, up from 24% in 2009. In Greece and Iceland, the pro-
countries with the exception of Greece where public cover- portion of pharmaceutical spending paid for by households
age for pharmaceuticals is higher (67% vs. 64%). Public cov- directly went up by 10 percentage points or more since
erage for pharmaceuticals is high in countries such as 2009.
France, Japan and Germany where coverage by public
financing schemes accounts for 70% or more of total costs.
Private sources have to cover more than half of the total
pharmaceutical bill in eight OECD countries, with public Definition and comparability
coverage being the lowest in Poland (32%), the United
States (34%) and Canada (36%). However, in the United See indicator on pharmaceutical expenditure for defi-
States and Canada, private health insurance plays a signif- nition of what is included and possible limitations.
icant role in covering parts of the pharmaceutical costs for See indicator on financing of health care for definition
patients. Poland reports large spending on privately of public and private spending on health.
financed OTC pharmaceuticals. Health services refer to inpatient and outpatient care
The growth in public spending on pharmaceuticals has (including day cases), long-term health care and
remained below total health spending growth over the last auxiliary services.
decade (see Indicator Pharmaceutical expenditure) with
recent growth rates in sharp decline as compared to pre-
crisis years (Figure 10.5). Between 2009 and 2013, public
expenditure on pharmaceuticals dropped by 3.2% on aver- References
age across OECD countries while it increased by 2.7% each Belloni, A., D. Morgan and V. Paris (forthcoming), Pharma-
year in the 2005-09 period. The reduction was particularly ceutical Expenditure and Policies: Past Trends and
steep in Portugal (-11.1%), Denmark (-10.4%) and Iceland Future Challenges, OECD Working Paper, OECD Publish-
(-9.9%). Greece and the Netherlands have also seen signifi- ing, Paris.

10.3. Expenditure on pharmaceuticals as a share of GDP, 10.4. Public share of expenditure on health services
2013 (or nearest year) and goods, 2013 (or nearest year)
Public Private Pharmaceuticals1 Health services
Greece 1 2.8 Luxembourg 82
Hungary 2.2 Netherlands2 80

Japan 2.1 Germany 75
Slovak Rep.1 2.0 Japan 71
United States 1.9 France 69
Slovenia 1.7
Austria 68
Canada 1.7
Slovak Rep. 67
Italy1 1.6
Greece 67
Spain 1.6
Belgium 66
France 1.6
Germany 1.5 Switzerland 65
Belgium 1.4 Czech Rep. 62
Ireland1 1.4 Spain 61

Portugal1 1.4 Norway 58
OECD29 1.4 OECD26 57
Poland 1.4 Portugal 55
Australia 1.3
Korea 55
Czech Rep. 1.3
Estonia 54
Korea 1.3
Finland 53
Austria 1.2
Switzerland 1.2 Sweden 52
Finland 1.2 Australia 49
Estonia 1.1 Slovenia 48

Sweden 1.0 Hungary 43
Israel1 1.0 Denmark 43
Iceland 0.9 Iceland 38
Netherlands1 0.9
Canada 36
Norway 0.6
United States 34
Luxembourg 2 0.6
Poland 32
Denmark 0.5
0 1 2 3 0 50 100
% GDP %
1. Includes medical non-durables. 1. Includes medical non-durables.

2. Excludes spending on over-the-counter medicines. 2. The shares for the Netherlands are overestimated as they include
Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en. compulsory co-payments by patients to health insurers.
1 2 http://dx.doi.org/10.1787/888933281325 Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en.
1 2 http://dx.doi.org/10.1787/888933281325
10.5. Average annual growth in public pharmaceutical expenditure1 per capita, in real terms, 2005-09 and 2009-13
(or nearest periods)
% 2005-09 2009-13
15
11.6
10.3
10.3
8.5
10
7.8
5.5
5.4
5.4
5.0
4.9
3.8
3.8
5
3.4
3.4
2.8
2.7
2.7
2.2
2.1
1.6
1.2
1.1
0.8
0.8
0.5
0.1
0.1
0
-0.3
-0.3
-0.5
-0.6
-0.7
-1.1
-1.1
-1.2
-1.4
-1.5
-1.9
-2.1
-2.1
-2.5
-2.9
-3.2
-3.2
-3.6
-3.9
-5
-4.2
-5.3
-5.7
-5.8
-6.4
-6.4
-6.6
-7.7
-10
-9.6
-9.9
-10.4
-11.1
-15
n
l
nd
es
d
28
nd
a
k
ce
p.
a in
nd
ly
ia
en
p.
li a
da
ay
d
ce
y
ria
ga
pa
nd
an
re
ni
ar
ar
an
an
ur
en
It a
Re
Re
iu
at
rw
ee
an
ra
ed
na
la
la
la
CD
st
r tu
to
Sp
Ko
nm
ng
Ja
bo
rm
la
el
nl
lg
ov
St
Po
Ir e
er
st
Au
Gr
Fr
Sw
Ca
No
Es
h
ak
Ic
er
Fi
OE
Be
Hu
m
Po
Au
it z
De
Ge
ec
Sl
d
ov
th
xe
i te
Cz
Sw
Ne
Sl
Lu
Un
1. Includes medical non-durables.

1 2 http://dx.doi.org/10.1787/888933281325

Pharmacists assist people in obtaining medication and including branch pharmacies and supplementary phar-
ensuring that these are used in a safe and proper fashion. macy units attached to the main pharmacy (Vogler et al.,
The role of the pharmacists has changed over the recent 2012).
years. Although their main role is still to dispense medications
The range of products and services provided by the pharma-
in community pharmacies, pharmacists are increasingly
cies varies across countries. In most European countries, for
providing direct care to patients (e.g. flu vaccinations in
example, pharmacies can also sell cosmetics, food supple-
Ireland), both in community pharmacies and as part of inte-
ments, medical devices and homeopathic products and in a
grated health care provider teams.
few countries pharmacies can also sell reading glasses and
OECD countries generally have between 50 and 130 pharma- didactic toys (Martins et al., 2015). Depending on countries
cists per 100 000 population. Japan has by far the highest legislation, pharmacies can provide services such as vacci-
density of pharmacists, at twice the OECD average, while nation, medication use review, unit dose dispensing,
the density of pharmacists is low in Turkey, Chile and the generic substitution, point of care testing, medication
Netherlands (Figure 10.6). Between 2000 and 2013, the num- administration, needle exchange programme, take back
ber of pharmacists per capita has increased in nearly all medicines (disposal of medicines), etc.
OECD countries, with the exception of Switzerland. It
increased most rapidly in Portugal, Ireland, Japan, Spain and
Hungary. Definition and comparability
In Japan, the strong increase in the number of pharmacists
can be attributed to a large extent to governments efforts to Practicing pharmacists are defined as the number of phar-
separate more clearly drug prescribing from drug dispensing. macists who are licensed to practice and provide direct
Traditionally, the vast majority of prescription drugs in services to clients/patients. They can be either salaried
Japan were dispensed directly by doctors. Over the years, or self-employed, and work in community pharmacies,
the Japanese government has taken steps to encourage the hospitals and other settings. Assistant pharmacists and
separation of drug prescribing from dispensing. The Medical the other employees of pharmacies are normally
Service Law was first amended in 1997 and then in 2006 to excluded.
recognise community pharmacies as facilities providing In Ireland, the figures include all pharmacists registered
health goods and services. Following these amendments, with the Pharmaceutical Society of Ireland, possibly
the percentage of prescriptions dispensed by pharmacists including some pharmacists who are not in activity. In
rose from less than 40% of all prescriptions in 2000 to 67% in addition they include assistant pharmacists, pharma-
2013, while the number of community pharmacies ceutical assistants, and doctors who are dispensing
increased from 48 252 to 57 071 (Japanese Pharmaceutical medication (approximately 140 in 2007), resulting in an
Association, 2015). over-estimation compared with the data provided by
Most pharmacists work in community pharmacies, but other countries. Assistant pharmacists are also
some also work in hospital, industry, research and academia included in Iceland.
(FIP, 2015). For instance, in Canada in 2012, more than Community pharmacies are premises which in accordance
three-quarters of practising pharmacists worked in a com- to the local legal provisions and definitions may operate
munity pharmacy while about 25% worked in hospitals and as a facility in the provision of pharmacy services in the
community settings. The number of community phar-
other health care facilities (CIHI, 2013). In Japan, around
macies reported are the number of premises where
55% of pharmacists worked in community pharmacies in
dispensing of medicines happened under the supervision
2012, while around 20% worked in hospitals or clinics and
of a pharmacist.
the other 25% worked in other settings (Japanese Pharma-
ceutical Association, 2015).
The number of community pharmacies varies widely References
across OECD countries (Figure 10.7). This big variation can CIHI Canadian Institute for Health Information (2013),
be explained by the more or less active planning role of Pharmacist Workforce, 2012 Provincial/Territorial High-
governments and agencies; by the remuneration model lights, Ottawa, Canada.
used in the country, as well as by different dispensing FIP International Pharmaceutical Federation (2015), Global
channels of medicines. In addition to community pharma- Trends Shaping Pharmacy Regulatory Frameworks, Distribu-
cies, medicines can be dispensed through hospital phar- tion of Medicines and Professional Services, The Hague.
macies (serving both inpatients and outpatients) or can be Japan Pharmaceutical Association (2015), Annual Report of
provided directly by doctors in a few countries. For example, JPA 2014-2015, Tokyo.
the relatively low number of community pharmacies in the Martins, S.F. et al. (2015), The Organizational Framework of
Netherlands may be explained partly by the fact that Community Pharmacies in Europe, International Journal
patients can also purchase their prescription drugs directly of Clinical Pharmacy, May 28.
from some doctors (Vogler et al., 2012). There are about RIVM National Institute for Public Health and the Environ-
400 GPs who are selling medicines in the Netherlands, pro- ment (2014), The Dutch National Atlas of Public Health,
viding access to drugs especially in rural areas where the Bilthoven.
nearest pharmacy may be quite far away (RIVM, 2014). Vogler, S. et al. (2012), Impact of Pharmacy Deregulation
Denmark has few, but large, community pharmacies and Regulation in European Countries, Vienna.

10.6. Practising pharmacists, 2000 and 2013 (or nearest year)

2000 2013
180
161
160
140
127
119
120 114 112 111
108 106
100 97
92 91
83
80 80
80 77 77 76 75 73 72 72 70 69
66 65 64 62 61
60 58
54
50
40 35 35
21
20
S w al
De and
n
a in
Gr d
Fr
ce
Un C a
Au tes
bo l
Ge r e a

th il e
s
d
m
St a
Un O a li a
Ki 33
Ze r y
Po om
N e un n
Po g
Au d
No ia
Es y
a
ov ny
e c p.
S l e p.
i t z ni a
Tu rk
xe r a e
ce
ly
ey
nd
nd
a
i te nad
ni
an
pa
an
H e
n
ur
g
C z Re
iu
a
an
w ga
rw
an
ed
Sl r ma
It a
la
st
D
rk
ee
r tu
Sp
to
Sw e
R
Ne Ch
Ko
nm
d
la
r
a
la
el
Ja
nl
l
L u Is
lg
v
i te EC
er
st
al
ng
Ir e
ak
h
o
Ic
er
Fi
Be
m
d
1. Data include not only pharmacists providing direct services to patients, but also those working in the health sector as researchers, for
pharmaceutical companies, etc.
2. Data refer to all pharmacists licensed to practice (resulting in a large over-estimation of the number of practising pharmacists).
10.7. Community pharmacies, 2015 (or nearest year)

50
47.2
45.0
45 43.9
41.8
40
37.5
35.7
35 34.0
31.5
29.9
30 28.0
26.7
25.2 25.1
25 23.6 23.1
22.1 21.3
20.2
20
15.4 15.0 14.9
15 13.3
11.7
10.1
10
6.0
5 3.9
0
l
nd
s*
*
y
li a
d*
ria
a in
nd
ay
el
m
nd
y
ce
ly
da
25
k
ga
ke
an
re
es
en
ar
ar
an
pa
ra
It a
iu
do
rw
nd
an
ra
na
an
la
la
la
CD
st
r tu
Sp
Ko
nm
ng
r
at
rm
ed
nl
Ja
Is
lg
Tu
Ir e
Po
er
st
ng
Au
Fr
Ca
la
el
No
St
Fi
OE
Be
Hu
Po
Sw
Au
it z
Ge
De
er
Ic
Ki
d
th
Sw
i te
d
Ne
i te
Un
Un
* Estimates.
Source: FIP (2015), Global Trends Shaping Pharmacy Regulatory Frameworks, Distribution of Medicines and Professional Services. 2013-1015.
1 2 http://dx.doi.org/10.1787/888933281337

In general, pharmaceutical consumption continues to Where antidepressants consumption is very low Korea,
increase, partly driven by a growing demand for drugs to Chile, Estonia there may be a case for addressing unmet
treat ageing-related and chronic diseases and by changes needs. In other countries with particularly high antidepres-
in clinical practice. This section examines consumption of sants consumption, there is a need to assess the appropri-
four categories of pharmaceuticals: antihypertensive, cho- ateness of prescribing patterns, and the availability of
lesterol-lowering, antidiabetic and antidepressant drugs. alternative treatments for depression.
Consumption is measured in defined daily doses (DDD)
(see the box on Definition and comparability).
Consumption of antihypertensives has nearly doubled in Definition and comparability
OECD countries between 2000 and 2013. It has more than
tripled in Estonia and quadrupled in Luxembourg Defined daily dose (DDD) is the assumed average
(Figure 10.8). It is highest in Germany and Hungary, almost maintenance dose per day for a drug used for its main
five-fold the level of Korea and Turkey. These variations indication in adults. DDDs are assigned to each active
reflect both differences in the prevalence of high-blood ingredient(s) in a given therapeutic class by interna-
pressure and in clinical practice. In 2008, 16% of the Korean tional expert consensus. For instance, the DDD for
population had high blood pressure, against 26% in oral aspirin equals 3 grams, which is the assumed
Germany and 37% in Hungary, while the average number of maintenance daily dose to treat pain in adults. DDDs
DDD prescribed per patient with high blood pressure was do not necessarily reflect the average daily dose actu-
lower in Korea (0.5) than in Hungary (1.1) and Germany (1.2) ally used in a given country. DDDs can be aggregated
(OECD, 2015). within and across therapeutic classes of the Anatomic-
The use of cholesterol-lowering drugs has more than tri- Therapeutic Classification (ATC). For more detail, see
pled in OECD countries between 2000 and 2013 (Figure 10.9). www.whocc.no/atcddd.
The Slovak Republic, the United Kingdom and Australia The volume of hypertension drugs consumption pre-
had the highest consumption per capita in 2013, with levels sented in Figure 10.8 refers to the sum of five ATC2
over 40% higher than the OECD average. Prescription clini- categories which can all be prescribed against hyper-
cal guidelines for anti-cholesterol treatments have been tension (antihypertensives, diuretics, beta-blocking
updated several times since the 1990s, recommending agents, calcium channel blockers and agents acting
wider screening, earlier treatments, and higher dosages. on the renin-angiotensin system).
This explains part of the high growth observed during the Data generally refer to outpatient consumption only,
period. except for the Czech Republic, Estonia, Italy and Sweden
The use of antidiabetics has almost doubled in OECD coun- where data also include hospital consumption. The
tries between 2000 and 2013 (Figure 10.10). This growth can data for Canada relate to three provinces only (British
be explained by a rising prevalence of diabetes, largely Columbia, Manitoba and Saskatchewan). The data for
linked to increases in the prevalence of obesity (see Spain refer to outpatient consumption for prescribed
indicator on overweight and obesity in Chapter 4), a major drugs covered by the National Health System (public
risk factor for the development of type-2 diabetes. In 2013, insurance). Data for Luxembourg are underestimated
the consumption of antidiabetics was highest in Finland, due to incomplete consideration of products with
Germany and the United Kingdom. multiple active ingredients.
Consumption of antidepressants has increased consider-
ably in most OECD countries since 2000 (Figure 10.11). This
might reflect some narrowing of the treatment gap for
References
depression. However, there is significant variation in con-
sumption of antidepressants between countries. Iceland Grandfils, N. and C. Sermet (2009), Evolution 1998-2002 of
reported the highest level of consumption of antidepres- the Antidepressant Consumption in France, Germany
sants in 2013, twice the OECD average, followed by Australia, and the United Kingdom, Document de travail IRDES,
Portugal and Canada. Chile, Korea and Estonia reported low No. 21, Paris.
consumption levels. Moore, M. et al. (2009), Explaining the Rise in Antidepres-
The level of antidepressants consumption depends on the sant Prescribing: A Descriptive Study Using the General
prevalence of depression in each country, and on how Practice Research Database, British Medical Journal,
depression is diagnosed and treated. This, in turn, depends Vol. 339:b3999.
on other available therapies, local guidelines, and prescrib- OECD (2015), Cardiovascular Disease and Diabetes: Policies for
ing behavior (OECD, 2014; Moore et al., 2009). These factors Better Health and Quality of Care, OECD Publishing, Paris,
vary between countries. In England and in France, the http://dx.doi.org/10.1787/9789264233010-en.
increase in antidepressants consumption has been associ- OECD (2014), Making Mental Health Count: The Social and Eco-
ated with a longer duration of drug treatment (Grandfils nomic Costs of Neglecting Mental Health Care, OECD Pub-
and Sermet, 2009; Moore et al., 2009). lishing, Paris, http://dx.doi.org/10.1787/9789264208445-en.

10.8. Antihypertensive drugs consumption, 10.9. Cholesterol-lowering drugs consumption,

2000 and 2013 (or nearest years) 2000 and 2013 (or nearest years)
2000 2013 2000 2013
Turkey 124 Chile 10

Korea 141 Turkey 26
Austria 184 Estonia 44
Greece 194 Korea 45
Israel 217 Austria 69
Luxembourg 223
Germany 73
Australia 239
Italy 83
Portugal Sweden 86
250
France Iceland 91
266
Spain France 92
269
Iceland OECD27 95
274
Spain 96
Norway 279
Canada 99
Netherlands 303 Portugal 102
Canada 311 Finland 102
OECD26 318 Greece 103
Estonia 323 Czech Rep. 103
Belgium 328 Hungary 105
Sweden 368 Israel 109
United Kingdom 380 Slovenia 110
Denmark 398 Netherlands 112
Italy 399 Luxembourg 117
Slovak Rep. 410 Norway 120
Slovenia 414 Denmark 126
Finland 427 Belgium 130
Czech Rep. 442 Australia 134
Hungary 543 United Kingdom 135
Germany 575 Slovak Rep. 153
0 100 200 300 400 500 600 0 30 60 90 120 150 180

Defined daily dose, per 1 000 people per day Defined daily dose, per 1 000 people per day
10.10. Antidiabetic drugs consumption, 2000 and 2013 10.11. Antidepressant drugs consumption, 2000 and 2013
(or nearest years) (or nearest years)
2000 2013 2000 2013
Chile 9 Chile 13
Austria 40 Korea 20
Iceland 43 Estonia 21
Norway 49 Hungary 28
Denmark 52 Turkey 35
Estonia 55 Slovak Rep. 35
Israel 55 Israel 42
Sweden 56
Italy 43
Turkey 56
Netherlands 43
Canada 58
Greece 44
OECD27 Czech Rep. 49
62
Australia France 50
62
Germany 53
Portugal 63
Slovenia 53
Luxembourg 64
Luxembourg 54
Belgium 65
Norway 56
Korea 65 OECD28 58
Slovak Rep. 66 Austria 59
France 66 Spain 65
Greece 66 Finland 69
Spain 67 Belgium 72
Italy 67 New Zealand 73
Slovenia 73 Denmark 80
Netherlands 75 United Kingdom 82
Hungary 78 Sweden 84
Czech Rep. 80 Canada 85
United Kingdom 82 Portugal 88
Germany 83 Australia 96
Finland 86 Iceland 118
0 20 40 60 80 100 0 20 40 60 80 100
Defined daily dose, per 1 000 people per day Defined daily dose, per 1 000 people per day

All OECD countries see the development of generic markets differential price between brand-name and generic drugs is
as a good opportunity to increase efficiency in pharmaceuti- much higher in the United Kingdom than in Germany.
cal spending but many do not fully exploit the potential of One way to exert pressure on generic prices is tendering,
generics (Figure 10.12). In 2013, generics accounted for more which has been used in New Zealand, the Netherlands and
than three-quarters of the volume of pharmaceuticals sold in Germany with some success. Many countries, however, prefer
the United States, the United Kingdom, Chile, Germany and regulating the price of generics at market entry by reference
New Zealand, while they represented less than one-quarter of to the price of the originator (a practice known as generic
the market in Luxembourg, Switzerland, Italy, and Greece. price linkage). Several countries have recently increased this
Some of the differences in generic uptake can be explained by gap. In Canada, several provinces have introduced or reduced
market structures, notably the number of off-patent medi- the reimbursement prices of generics included in public
cines, and by prescribing practices, but generic uptake also plans formularies since 2010. As a result, generic price caps
very much depends on policies implemented by countries are around 25% of brand name products price (PMPRB, 2015).
(EGA, 2011; Vogler, 2012). Several countries have expanded France and Greece also increased the gap between originator
their efforts to encourage generic uptake since the onset of and generic prices to 40% and 60% respectively (Belloni et al.,
the economic crisis in 2008. forthcoming).
Prescribing in International Non-proprietary Name (INN) is
permitted in two-thirds of OECD countries and is mandatory
in a few countries (e.g. Estonia since 2010, Portugal and Spain Definition and comparability
since 2011, and France since 2015). Similarly, pharmacists are
allowed to substitute brand-name drugs with generics in a A generic is defined as a pharmaceutical product
majority of OECD countries. While generic substitution is which has the same qualitative and quantitative com-
mandatory in some countries (e.g., Denmark, Finland, Spain, position in active substances and the same pharma-
Sweden, Italy), New Zealand and the United Kingdom have ceutical form as the reference product, and whose
high generic penetration without any substitution mandate. bioequivalence with the reference product has been
demonstrated. Generics can be classified in branded
Financial incentives for physicians, pharmacists and patients
generics (generics with a specific trade name) and
have been implemented to boost the development of generic
unbranded generics (which use the international non-
markets. For instance, France (in 2009 and 2012) introduced
proprietary name and the name of the company).
incentives for GPs to prescribe generics through a pay-for-per-
formance scheme while Japan (in 2012) increased the share of In many countries, the data cover all pharmaceutical
generics in total prescribing leading to a bonus. consumption. However, several countries provide data
covering only the community pharmaceutical market or
Pharmacies are often paid through mark-ups based on the
the reimbursed pharmaceutical market.
price of medicines. This disincentive to substitute a generic
for a more expensive drug has been addressed in some coun- The share of generic market expressed in value can be
tries. France guarantees pharmacists an equivalent mark-up, the turnover of pharmaceutical companies, the
while in Switzerland, pharmacists receive a fee for generic amount paid for pharmaceuticals by third-party payers,
substitution. In several countries, pharmacists have the obli- or the amount paid by all payers (third-party and
gation to inform patients about the possibility of a cheaper consumers). The share of generic market in volume
alternative. can be expressed in DDDs or as a number of packages/
boxes or standard units.
Patients have a financial interest to choose cheaper drugs
when their co-payment is lower for generic drugs than for its
equivalent. This is generally the case in all systems using
reference prices (or fixed reimbursement amount) for clusters References
of products. In Greece, patients choosing originator over
Belloni, A., D. Morgan and V. Paris (forthcoming), Pharmaceu-
generic drugs are now required to pay for the difference. In
tical Expenditure and Policies: Past Trends and Future
France, since 2010, patients refusing generic substitution have
Challenges, OECD Working Paper, OECD Publishing, Paris.
to pay in advance for their drugs and are reimbursed later.
EGA European Generic Medicines Association (2011), Mar-
These policies, associated with patent expiries of several
ket Review The European Generic Medicines Markets, EGA.
blockbusters in recent years, have contributed to the
increase in generic market share observed over the past PMPRB Patented Medicine Prices Review Board (2015),
decade (Figure 10.13). In Portugal, the generic reimbursed NPDUIS CompassRx: Annual Public Drug Plan Expenditure
market grew from virtually zero in 2000 to 39% in volume Report 2012/13, 1st Edition.
and 23% in value in 2013. In Spain, the generic reimbursed Vogler, S. (2012), The Impact of Pharmaceutical Pricing and
market share reached 47% in volume and 21% in value in Reimbursement Policies on Generic Uptake: Implemen-
2013, up from 3% in 2000. Beyond encouraging generic take- tation of Policy Options on Generics in 29 European
up, it is also important to promote the lowest possible price Countries An Overview, Generics and Biosimilars Initia-
for generics. Figure 10.12 suggests, for instance, that the tive Journal, Vol. 1, No. 2, pp. 44-51.

10.12. Share of generics in the total pharmaceutical market, 2013 (or nearest year)
% Value Volume
90
84 83
80 80
77
72
70 70
60 59
55 54
50
48 47
46 46 45
41 40 40 39
37 37
35
33 34
32 30
30 28 29 29 28
24 24
23
21
19 20 19
17 17 17
16
14 16 16 15 16 17
14
11 11
8
4
0
m1
il e 2
y1
d1
s1
k1
ia 2
l1
m1
ce 1
nd 1
ce 1
g1
nd
y
p.
da
26
ay
ly
d
p.
a
es
ria
a in
ke
ni
an
pa
ga
It a
nd
an
Re
ar
Re
an
ur
rw
na
at
la
en
CD
iu
do
to
an
ee
Ch
la
st
r
nl
Ja
r tu
Sp
nm
bo
Tu
rm
er
St
la
al
lg
Ca
No
ov
Es
Ir e
ak
h
ng
Au
Gr
Fr
Fi
OE
er
Ze
it z
ec
Be
m
Po
De
Ge
d
Sl
ov
Ki
th
i te
xe
Cz
Sw
w
Sl
Ne
d
Ne
Un
Lu
i te
Un
1. Reimbursed pharmaceutical market.

2. Community pharmacy market.
1 2 http://dx.doi.org/10.1787/888933281352
10.13. Trend in share of generics in the reimbursed pharmaceutical market, selected countries, 2000 to 2013
In value In volume
Germany Portugal Germany Portugal
% Spain United Kingdom % Spain United Kingdom
40 90
75
30
60
20 45
30
10
15
0 0
2000 2002 2004 2006 2008 2010 2012 2000 2002 2004 2006 2008 2010 2012

1 2 http://dx.doi.org/10.1787/888933281352

The pharmaceutical industry devotes significant resources trend may reverse in the coming years due to changes in
to research and development (R&D). In 2011, the industry the R&D model, or the emergence of new technology (e.g.
spent USD 92 billion on R&D (OECD, 2015). This represents precision medicine).
10-15% of industry revenues.
While pharmaceutical and biotechnology companies are
the greatest contributors to pharmaceutical R&D, pharma-
ceutical R&D financing is a complex mix of private and Definition and comparability
public funding. The industry receives R&D tax credits in
many countries, and the development of medicines draws Business enterprise expenditure on R&D (BERD) cov-
heavily on knowledge and innovation derived from other ers R&D activities carried out in the private sector by
sectors including higher education and NGOs (Kezselheim performing firms and institutes, regardless of the origin
et al., 2015). of funding. This includes all firms, organisations and
institutions whose primary activity is the production
Worldwide, most pharmaceutical R&D activity takes place
of goods and services for sale to the general public at
in OECD countries. In 2011, the pharmaceutical industry
an economically significant price, and the private and
spent close to USD 50 billion for R&D in the United States,
not-for-profit institutions serving them. BERD will
11.5 billion in Japan, 5.2 in Germany and 3.7 in France. As a
register in the country where the R&D activity took
share of GDP, pharmaceutical industry R&D spending is
place, not the country of origin of the organisations
highest in Switzerland (0.63%), Belgium (0.45%), Slovenia
funding the activity.
(0.45%) and Denmark (0.36%) (Figure 10.14). In the United
States and Japan, the percentages were 0.30 and 0.26 Data are provided by participating countries using a
respectively. survey. When assessing changes in BERD over time, it
is necessary to take account of changes in methods
In some countries, pharmaceutical R&D accounts for one-
and breaks in series, notably in terms of the extension
fourth to one-third of total private R&D expenditure,
of survey coverage, particularly in the services sector,
reflecting a high degree of specialisation. This is the case in
and the privatisation of publicly owned firms. Identi-
Belgium (31%), Switzerland (30%), the United Kingdom
fying new and occasional R&D performers is also a chal-
(28%), Hungary (26%) and Slovenia (25%). Sixteen and ten
lenge and OECD countries take different approaches in
per cent of private R&D was spent on pharmaceuticals in
their BERD surveys.
the United States and Japan respectively.
Gross domestic product (GDP) = final consumption +
Expenditure on R&D in the pharmaceutical industry in
gross capital formation + net exports. Final consump-
OECD countries doubled in real terms between 2000 and
tion of households includes goods and services used
2011 (Figure 10.15). Expenditure growth was the highest in
by households or the community to satisfy their indi-
the United States (+85%), followed by Japan (+76%) and
vidual needs. It includes final consumption expendi-
Europe (+38%). Outside the OECD, China has seen pharma-
ture of households, general government and non-
ceutical R&D spending increase by 3.4-fold during that
profit institutions serving households. In countries,
time.
such as Ireland and Luxembourg, where a significant
Is this increase in R&D spending associated with a higher proportion of GDP refers to profits exported and not
output or productivity? In the United States, the worlds available for national consumption, GNI may be a
largest developer of pharmaceuticals, the annual number more meaningful measure than GDP.
of approved new drugs, formulations or indications has
more than doubled since 1970 (Figure 10.16). However,
when compared with R&D spending over that period
(adjusted for inflation), the number of approvals per
References
billion USD spent on R&D has reduced by a factor of 15
(Figure 10.16). Kezselheim, A., Y. Tan and J. Avorn (2015), The Roles of
Academia, Rare Diseases, and Repurposing in the Devel-
The reasons for this observation are likely to be complex.
opment of the Most Transformative Drugs, Health
Growing requirements to obtain regulatory approval have
Affairs, Vol. 34, pp. 286-293.
increased development costs. Higher failure rates and an
ever-increasing back catalogue of effective drugs may OECD (2015), Main Science and Technology Indicators Database,
also be a factor. More fundamental problems with the cur- online, available at: www.oecd.org/sti/msti.htm [accessed
rent R&D model and development pipeline have also been 8 July 2015].
suggested (Scannell et al., 2012). Risk-benefit decisions Scannell, J. et al. (2012), Diagnosing the Decline in Pharma-
made by industry regarding early R&D targets may also be ceutical R&D Efficiency, Nature Reviews Drug Discovery,
a function of the regulator, payer and the community pp. 191-200.
response to the eventual product. Of course, the downward

10.14. Business expenditure on R&D (BERD) in pharmaceutical industry as a proportion of GDP and of total BERD, 2011
(or nearest year)
% GDP % total BERD
% GDP % BERD
0.7 35
0.6 30
0.5 25
0.4 20
0.3 15
0.2 10
0.1 5
0 0
B e nd
Ic s
d
l
Sl m
Un De ni a
K ar k
Sw an
i t e dom
Hu en
Fr y
Ge c e
el
nd
Fi a
Po ds
y
a in
A d
th tr ia
ly
Ca e
da
Au ico
ec ia
N o p.
E y
ov ni a
Tu .
Po y
nd
a
ga
p
e
an
a
re
e
c
in
ar
an
an
ra
C z r al
It a
Re
Re
iu
at
rk
an
rw
ee
ed
n
p
na
la
la
la
e
Sp
r tu
Ch
Sl s to
ex
i t e nm
Ko
ng
Ne us
rm
el
la
Ja
nl
Is
lg
ov
St
Ir e
er
st
Un in g
Gr
ak
M
er
it z
d
Sw
Source: OECD Main Science and Technology Indicators Database.

1 2 http://dx.doi.org/10.1787/888933281362
10.15. Business expenditure on R&D in the pharmaceutical sector by region in 2000, 2005 and 2011 (or nearest years)
in 2005 USD PPP
Europe United States Other OECD Japan China
2005 USD PPP (billions)
90
80
70
60
50
40
30
20
10
0
2000 2005 2011
Source: OECD Main Science and Technology Indicators Database.

1 2 http://dx.doi.org/10.1787/888933281362
10.16. Annual FDA pharmaceutical approvals, per USD billion R&D spend (indexed to 2008 USD)
FDA approvals (3-year average) FDA approvals per USD billion R&D spend (3-year average)
FDA approvals FDA approvals per USD billion R&D spend
140 30
120 25
100
20
80
15
60
10
40
20 5
0 0
1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
Source: Pharmaceutical Research and Manufacturers of America (PhRMA); Food and Drug Administration (FDA); Scannell et al (2012).
1 2 http://dx.doi.org/10.1787/888933281362

11. AGEING AND LONG-TERM CARE
Demographic trends
Life expectancy and healthy life expectancy at age 65
Self-reported health and disability at age 65
Dementia prevalence
Recipients of long-term care
Informal carers
Long-term care workers
Long-term care beds in institutions and hospitals
Long-term care expenditure
international law.

Demographic trends
Longer life expectancies (see Life expectancy indicator in nearly 5% in 2010, and is expected to rise to 11% by 2050.
Chapter 3) and declining fertility rates mean that older The pace of population ageing has been slower in other
people make up an ever-increasing proportion of the popu- parts of the world, although it is expected to accelerate in
lations of OECD countries. coming decades. In large partner countries including Brazil,
On average across OECD countries, the share of the popula- China, India, Indonesia and South Africa, only 2% of the
tion aged over 65 years has increased from less than 9% in population was 80 years and over in 2010, but this share is
1960 to 15% in 2010 and is expected to nearly double in the expected to reach around 5% by 2050.
next four decades to reach 27% in 2050 (Figure 11.1, left Although the pressure that this growing proportion of peo-
panel). In about two thirds of OECD countries, at least one- ple aged 65 and 80 over will put on long-term care systems
quarter of the population will be over 65 years of age by will depend on the health status of people as they reach
2050. This proportion is expected to be especially large in these ages, population ageing is likely to lead to greater
Japan, Korea and Spain where nearly 40% of the population demand for elderly care. As the share of the economically
will be aged over 65 years by 2050. Population ageing will active population is expected to decline, it will also affect
also occur rapidly in China where the share of the popula- the financing of social protection systems and the potential
tion over 65 is expected to triple between 2010 and 2050, to supply of labour in the economy. On average across OECD
reach a level just below the OECD average. Conversely, countries, there were slightly more than four people of
Israel, the United States and Mexico will see a more gradual working age (15-64 years) for every person 65 years and
increase in the share of the elderly population due to sig- older in 2012. This rate is projected to halve from 4.2 in 2012
nificant inflows of migrants and higher fertility rates. to 2.1 on average across OECD countries over the next
The growth in the share of the population aged 80 years 40 years (OECD, 2013).
and over will be even more dramatic (Figure 11.1, right
panel). On average across OECD countries, 4% of the popu-
lation were 80 years old and over in 2010. By 2050, the per-
centage will increase to 10%. In Japan, Spain and Germany,
the proportion of the population aged over 80 is expected to Data on the population structure have been extracted
nearly triple between 2010 and 2050 (rising from 6% to 16% from the OECD Historical Population Data and Projec-
in Japan and from 5% to 15% in Spain and Germany). The tions (1950-2050). The projections are based on the
rise will be even faster in Korea where the share of the pop- most recent medium-variant population projections
ulation aged over 80 years will grow from 2% to 14% over from the United Nations, World Population Prospects
the next four decades. China will see similarly rapid ageing, 2012 Revision.
with the share of the population aged over 80 rising from
1% to 8%.
Population ageing is a phenomenon affecting most coun-
tries around the world, but the speed of the process varies References
(Figure 11.2). The speed of population ageing is particularly OECD (2013), OECD Pensions at a Glance 2013: OECD and G20
rapid in the European Union, where the share of the popu- Indicators, OECD Publishing, Paris,
lation aged 80 years and over increased from 1.5% in 1960 to http://dx.doi.org/10.1787/pension_glance-2013-en.

Demographic trends
11.1. Share of the population aged over 65 and 80 years, 2010 and 2050
2010 2050 2010 2050
Population aged 65 years and over Population aged 80 years and over
Japan 23 39 Japan 6 16
Korea 11 37 Spain 5 15
Spain 17 36 Germany 5 15
Germany 21 33 Korea 2 14
Italy 20 33 Italy 6 14
Greece 19 33 Switzerland 5 12
Czech Rep. 15 32 Austria 5 12
Portugal 18 32 Netherlands 4 11
Slovak Rep. 13 31 Finland 5 11
Slovenia 17 31 France 5 11
Poland 13 30 Czech Rep. 4 11
Estonia 17 29 Portugal 5 11
Switzerland 17 28 Slovenia 4 11
Hungary 17 28 New Zealand 3 11
Austria 18 28 Greece 5 11
OECD34 15 27 Estonia 4 10
Netherlands 15 27 United Kingdom 4 10
Finland 17 27 OECD34 4 10
Ireland 11 26 Canada 4 10
France 17 26 Belgium 5 10
New Zealand 13 26 Poland 3 10
Lithuania 16 26 Slovak Rep. 3 9
Canada 14 25 Sweden 5 9
Belgium 17 25 Denmark 4 9
Sweden 18 24 Norway 5 9
United Kingdom 16 24 Iceland 3 8
China 8 24 Luxembourg 4 8
Denmark 17 24 Lithuania 4 8
Iceland 12 23 Ireland 3 8
Norway 15 23 Hungary 4 8
Latvia 18 23 United States 4 8
Brazil 7 23 Australia 4 7
Luxembourg 14 22 Chile 2 7
Chile 9 22 Latvia 4 7
Russian Fed. 13 21 China 1 7
United States 13 21 Brazil 1 6
Australia 14 21 Israel 3 6
Turkey 7 21 Russian Fed. 3 6
Israel 10 17 Turkey 1 5
Mexico 6 16 Mexico 1 4
Indonesia 5 16 Indonesia 1 3
India 5 13 India 1 2
South Africa 5 11 South Africa 1 2
0 10 20 30 40 0 5 10 15 20
% %
Source: OECD Historical Population Data and Projections Database, 2015.

1 2 http://dx.doi.org/10.1787/888933281371
11.2. Trends in the share of the population aged over 80 years, 1960-2050
% EU28 Japan OECD Partner countries1 World
18
16
14
12
10
0
1960 1970 1980 1990 2000 2010 2020 2030 2040 2050
1. Partner countries include Brazil, China, India, Indonesia, Latvia, Lithuania, Russia and South Africa.
Source: OECD Historical Population Data and Projections Database, 2015.
1 2 http://dx.doi.org/10.1787/888933281371

Life expectancy at age 65 has increased significantly for Life expectancy and healthy life expectancy at age 65 years
both men and women over the past few decades in OECD vary by educational status. For both men and women,
countries, rising by 5.5 years on average since 1970 highly educated people are likely to live longer and in bet-
(Figure 11.3). Some of the factors explaining these gains in ter health. Differences in life expectancy by education level
life expectancy at age 65 include advances in medical care are particularly large in Central and Eastern European
combined with greater access to health care, healthier life- countries, especially for men. In the Czech Republic,
styles and improved living conditions before and after peo- 65-year-old men with a high level of education could expect
ple reach age 65. to live seven years longer than those with a low education
Japan and Korea have achieved the highest gains in life level in 2012. By contrast, differences in life expectancy by
expectancy at age 65 since 1970, with an increase of almost education level are much smaller (less than two years) in
eight years. The gains have been much more modest in Nordic countries (Denmark, Finland, Norway and Sweden)
Hungary, the Slovak Republic and Mexico, with an increase and Portugal (Eurostat Database 2015).
of only about three years.
In 2013, people at age 65 in OECD countries could expect to
live another 19.5 years: 21 years for women and 18 years for
men (Figure 11.4). This gender gap of three years on aver- Definition and comparability
age across OECD countries has been fairly stable over time.
Life expectancy measures how long on average a per-
In 2013, life expectancy at age 65 was highest in Japan for
son of a given age can expect to live, if current death
women (24 years) and in Switzerland for men (nearly
rates do not change. However, the actual age-specific
20 years), followed by France in both cases. Among OECD
death rate of any particular birth cohort cannot be
countries, it was lowest in Hungary for both women and
known in advance. If rates are falling, as has been the
men.
case over the past decades in OECD countries, actual
Countries relative positions with respect to life expectancy life spans will be higher than life expectancy calcu-
at age 65 mirror closely their relative positions with regard lated using current death rates. The methodology
to life expectancy at age 80. Life expectancy at age 80 in used to calculate life expectancy can vary slightly
2013 was highest in France and Japan for women (who can between countries. This can change a countrys esti-
expect to live an additional 11.5 years) and highest in mates by a fraction of a year.
France and Spain for men (who can expect to live more
Disability-free life expectancy (or healthy life years)
than 9 years).
is defined as the number of years spent free of activity
Increased life expectancy at age 65 does not necessarily limitation. In Europe, this indicator is calculated
mean that the extra years lived are in good health. In annually by Eurostat for EU countries and some EFTA
Europe, an indicator of disability-free life expectancy countries. The disability measure is the Global Activ-
known as healthy life years is calculated regularly, based ity Limitation Indicator (GALI) which comes from the
on a general question about disability in the European European Union Statistics on Income and Living Con-
Union Survey of Income and Living Conditions (EU-SILC). ditions (EU-SILC) survey. The GALI measures limita-
Given that this indicator has only recently been developed, tion in usual activities due to health problems. While
long-time series are not yet available and efforts continue healthy life years is the most comparable indicator to
to improve its comparability. date, there are still problems with translation of the
Among European countries participating in the survey, the GALI question, although it does appear to satisfacto-
average number of healthy life years at age 65 was almost rily reflect other health and disability measures
the same for women and men, at 9.5 years for women and (Jagger et al., 2010).
9.4 years for men in 2013 (Figure 11.5). The absence of any
significant gender gap in healthy life years means that
many of the additional years of life that women experience
relative to men are lived with some type of activity limita- References
tion. Nordic countries (with the exception of Finland) had Jagger, C. et al. (2010), The Global Activity Limitation Indi-
the highest number of healthy life years at age 65 in 2013, cator (GALI) Measured Function and Disability Similarly
with women and men in Iceland and Norway expecting to across European Countries, Journal of Clinical Epidemiology,
live an additional 15 years free from disability on average. Vol. 63, pp. 892-899.

Years
Years
0
5
10
15
20
25
Years
0
2
4
6
8
10
12
14
16
0
5
10
15
20
25
Ic Ja Ja
el p p
an 15 Fr an 19 Fr an 21.5
d an 24 an
15 c 19 c
21.5
No
rw S w Sp e 24 S w Sp e
ay 15 i t z a in 19
i t z ain
er er 21.3
Sw 15 la 23 la
ed nd 19 nd
20.9
en 13 I 22 I
14
Au t al
19
De Au t al
nm Lu s t y 23
Lu st y 20.8
xe r a xe r a
ar
k 12 m li a 19 m li a 20.7
13
bo 22 bo
u u
19
Ir e Ca rg 20.5
Source: Eurostat Database 2015.

Ca rg
la
nd 11 na 22 na
d 19 d
12 20.3
Be 22 Is a

lg Is a ra
ra e
Lu iu
m 11 e 19 20.3
xe Ko l 21 Ko l
m 11 re re
Un 18 N e Gr a 20.2
22
bo N e Gr a
i te ur 11 w ee w ee
d g Ze ce 19 Ze ce 20.2
Ki
ng 11 al 22 al
do S w and 19 S w and 20.1
Sw m 11 ed 21 ed
it z 11 Ic en 19 Ic en 20.1
21
Men
er el el
la a a
nd 11 No nd 19 No nd 20.0
10 21 rw

rw
2013
Fr
an F i ay 19 F i ay 20.0
10 nl 21 nl
ce Un a Un a
11 18 i te Au nd 19.9
21
Po i te Au nd
r tu d st d st
Ki ri
18
Ki ri
ga 10 22 ng a 19.9
Men
ng a
l d d
9 P o P o
19
OE
CD Ne or t m 21 Ne or t m 19.8
Ne 24 9 th ug th ug
18

th 9
er al
l 22
er al
l 19.7
er
la Ge and 18
Ge and
nd 10 rm s 21 rm s 19.7
s
9 Be any 18 B e any
19.7
Sp lg 21 lg
a in 10 O E ium 18 O E ium
19.6
9 CD 21 CD
Au
18 Ir e 3 4 19.5
st 21
Ir e 3 4
Cz ria 9 la la
ec 9 Un Slo nd 18 Un Slo nd 19.5
21
Note: Countries are ranked in descending order of life expectancy for the whole population.
h i te ve i te ve
17
Re d ni a d ni a
9 19.3
Women
p. St 21 St
9
Women
Fi De a te 18
De a te
nl nm s 21 nm s 19.2
Note: Countries are ranked in descending order of healthy life expectancy for the whole population.
an 8 ar
d ar
k 18 k 19.1
Po 9 Ch 20 Ch
la il e 17 il e
18.3
nd 7 20
1970
Br Br
8 a a
E s z il 16 E s z il 18.0
20
Gr to
ee to n
8 n 15
11.5. Healthy life years at age 65, European countries, 2013

ce 17.8
11.4. Life expectancy at age 65 by sex, 2013 (or nearest year)
Po i a
20
Po ia
7 la la
11.3. Life expectancy at age 65, 1970 and 2013 (or nearest years)
M nd 16 M nd 17.7
It a
8 C z ex 20 C z ex
ly
Sl 7
ec ico 17 ec ico
17.7
ov Sl h R 19 Sl h R
ov e
ov e
en
ia 7 a k p. 16 a k p. 17.5
19 Re
Ge 8 Re
rm 15 Tu p. 16.8
an
Tu p. 19
y 7 Hu r ke y 15 Hu r ke y
16.7
Hu 7 ng 19 ng
ng ar ar
15 16.5
ar
y 6 La y 18 La y
6
Li t vi
14
Li t vi
Es Ru t h u a 19
Ru t h u a 16.2
to ss an ss an
Sl ni
a 5 14 16.1
19
ov So ian ia So ian ia
ak 6 ut Fe ut Fe
h d. 13 h d. 15.3
17
Re Af
4
p. Af
ric ric
4 a 12 a 13.9
16
195
1 2 http://dx.doi.org/10.1787/888933281383
1 2 http://dx.doi.org/10.1787/888933281383
1 2 http://dx.doi.org/10.1787/888933281383
Most OECD countries conduct regular health surveys which

allow respondents to report on different aspects of their Definition and comparability
health. These surveys often include a question on self-per-
ceived health status, along the lines of: How is your health Self-reported health reflects peoples overall percep-
in general?. Although these questions are subjective, indi- tion of their own health, including both physical and
cators of perceived general health have been found to be a psychological dimensions. Typically, survey respon-
good predictor of future health care use and mortality dents are asked a question such as: How is your
(DeSalvo, 2005; Bond et al., 2006). However, cross-country health in general? Very good, good, fair, poor, very
differences may be difficult to interpret, as survey ques- poor. OECD Health Statistics provides figures related to
tions may differ slightly and cultural factors can affect the proportion of people rating their health to be
responses. good/very good combined.
Keeping these limitations in mind, more than half of the Caution is required in making cross-country compari-
population aged 65 years and over report being in good sons of perceived health status, for at least two rea-
health in 13 of the 34 OECD countries (Figure 11.6). The sons. First, peoples assessment of their health is
highest rates are in New Zealand, Canada and the United subjective and can be affected by cultural factors. Sec-
States, where more than three-quarters of older people ond, there are variations in the question and answer
report good health, but the response categories offered to categories used to measure perceived health across
survey respondents in these three countries are different surveys/countries. In particular, the response scale
from those used in most other OECD countries, introducing used in Australia, Canada, New Zealand and the
an upward bias in the results (see box on Definition and United States is asymmetric (skewed on the positive
comparability below). Among European countries, older side), including the following response categories:
people in Sweden, Switzerland, Norway and Ireland report excellent, very good, good, fair, poor. The data
the best health status, with more than 60% assessing their reported in OECD Health Statistics refer to respondents
health to be good. answering one of the three positive responses (excel-
lent, very good or good). By contrast, in most other
At the other end of the scale, less than 20% of over-65s in
OECD countries, the response scale is symmetric, with
Portugal, Hungary, Estonia, Poland, Turkey, the Slovak
response categories being: very good, good, fair, poor,
Republic and Korea report being in good health. In nearly
very poor. The data reported from these countries
all countries, men over 65 were more likely than women to
refer only to the first two categories (very good,
rate their health to be good. On average across OECD coun-
good). Such difference in response categories biases
tries, 47% of men aged over 65 rated their health to be good
upward the results from those countries that are
or better, while 41% of women did so.
using an asymmetric scale.
The percentage of the population aged 65 years and over
Perceived general disability is measured in the EU-
who rate their health as being good or better has remained
SILC survey through the question: For at least the
fairly stable over the past 30 years in most countries where
past six months, have you been hampered because of
long time series are available. There has been significant
a health problem in activities people usually do? Yes,
improvement however in the United States, where the
strongly limited/Yes, limited/No, not limited. Persons
share has increased from 65% in 1982 to 77% in 2013.
in institutions are not surveyed, resulting in an under-
Measures of disability are not yet standardised across estimation of disability prevalence. Again, the mea-
countries, limiting the possibility for comparisons. In sure is subjective, and cultural factors may affect sur-
Europe, based on the EU Statistics on Income and Living vey responses.
Conditions survey, half of all over-65s reported that they
were limited either to some extent or severely in their usual
daily activities because of a health problem in 2013
(Figure 11.7). This ranged from a proportion of less than References
25% in Norway and Iceland up to nearly 75% in the Slovak Bond, J. et al. (2006), Self-rated Health Status as a Predictor
Republic and close to 70% in Estonia. On average across of Death, Functional and Cognitive Impairments: A Lon-
25 European OECD countries, most limitations reported gitudinal cohort Study, European Journal of Ageing, Vol. 3,
were moderate, 18% of the population aged 65 and over pp. 193-206.
reported severe limitations, which often correspond to
DeSalvo, K.B. et al. (2005), Predicting Mortality and Health-
needs for long-term care.
care Utilization with a Single Question, Health Services
Women were more likely than men to report severe activity Research, Vol. 40, pp. 1234-1246.
limitations due to a health problem in all European coun-
tries covered by this survey, with the exception of Poland.
The proportion of people aged 65 and over reporting some
severe activity limitations was highest in Greece and the
Slovak Republic, followed by Italy and Estonia (Figure 11.8).

11.6. Perceived health status in adults aged 65 years and over, 2013 (or nearest year)
Men Women Total
% of population aged 65 years and over reporting to be in good or very good health
100
90
80
70
60
50
40
30
20
10
0
Ca d 1
a1
Sw ia 1
OE el 1
il e 1
Au tes 1
No nd
th nd
De nds
l
ak a
i t z en
ay
K i nd
M 4
Au o
i t e Ic e k
Ge tr ia
ce
a in
Gr d
Sl c e
ia
ly
ec an
p.
Tu .
ey
Be m
xe ium
Is g
Es d
Hu i a
Po ar y
ga
p
an
ov r e
ar
an
n
ic
ur
en
n
It a
Re
i te ad
Re
do
an
rk
rw
an
ee
Sw ed
C z Jap
l
ra
la
la
la
CD
Sp
to
r tu
ex
ra
Ch
Sl Ko
nm
ng
s
bo
rm
la
N e Ir e l
nl
Lu lg
a
ov
Po
er
n
al
ng
Fr
st
h
St
er
Fi
m
Ze
d
w
d
Ne
Un
Un
1. Results not directly comparable with other countries due to methodological differences (resulting in an upward bias).
11.7. Limitations in daily activities in adults aged 65 years and over, European countries, 2013
Limited to some extent Limited strongly
80
70
60
50
40
30
20
10
0
l
en
nd
nd
ay
s
p.
ce
25
ria
a in
nd
ia
ce
ly
p.
ga
nd
ke
an
ni
ar
ar
an
an
ur
en
It a
Re
Re
iu
do
rw
an
ee
ed
la
la
la
CD
st
r tu
Sp
to
nm
ng
bo
rm
la
el
nl
lg
ov
Tu
Ir e
Po
er
ng
Au
Gr
Fr
Sw
No
Es
h
ak
Ic
er
Fi
OE
Be
Hu
m
Po
it z
De
Ge
ec
Sl
Ki
ov
th
xe
Cz
Sw
Ne
d
Sl
Lu
i te
Un
Source: Eurostat Database 2015. 1 2 http://dx.doi.org/10.1787/888933281398
11.8. Strong limitations in daily activities in adults aged 65 years and over, European countries, 2013
Men Women
40
35
30
25
20
15
10
5
0
nd
l
nd
k
ay
y
en
p.
a in
d
25
ce
ia
ria
nd
ly
p.
ce
m
a
ga
nd
an
ke
ar
ni
ar
an
an
ur
en
It a
Re
Re
iu
do
rw
an
ee
ed
la
la
la
CD
st
r tu
Sp
to
nm
ng
bo
r
rm
la
nl
el
lg
ov
Tu
Ir e
Po
er
ng
Au
Gr
Fr
Sw
No
Es
h
ak
er
Ic
Fi
OE
Be
Hu
m
Po
it z
De
Ge
ec
Sl
Ki
ov
th
xe
Cz
Sw
Ne
Sl
Lu
i te
Un
Note: Countries are ranked in ascending order of percentage with strong limitations in daily activities for the whole population.
Source: Eurostat Database 2015. 1 2 http://dx.doi.org/10.1787/888933281398

Dementia prevalence
Dementia describes a variety of brain disorders which pro- efforts are successful, the rise in prevalence may be less
gressively lead to brain damage and cause a gradual deteri- dramatic than these numbers suggest.
oration of the individuals functional capacity and social There has recently been a renewed international focus on
relations. Alzheimers disease is the most common form of tackling dementia and the OECD has been at the forefront
dementia, representing about 60% to 80% of cases. There is of this work, supporting countries to develop better poli-
currently no cure or disease modifying treatment, but bet- cies. Finding a cure must be the long-term goal, but this will
ter policies can improve the lives of people with dementia require greater investment and a more collaborative
by helping them and their families adjust to living with the approach to research, harnessing the potential of big data.
condition and ensuring that they have access to high qual- However, any cure is likely to take several years to develop
ity health and social care. and in the meantime countries need to act to improve the
According to WHO, 47.5 million people around the world lives of the millions of people living with dementia now.
live with dementia in 2015. With populations ageing and This must include promoting timely diagnosis, delivering
the effectiveness of preventive strategies still unclear, this high quality health and long-term care and providing sup-
number is expected to rise to 75.6 million by 2030 and port for families and carers (OECD, 2015).
almost triple by 2050, reaching 135.5 million (WHO, 2015). The
global cost of dementia was estimated at USD 604 billion in
2010 (Wimo et al., 2013) and as prevalence increases this
cost will grow. Definition and comparability
In 2015, there were an estimated 18 million people living
with dementia in OECD countries, equivalent to more than The prevalence estimates in Figure 11.9 are taken
one in every 70 people. Although some people develop from Prince et al. (2013), which is the latest and most
early-onset dementia, the vast majority of those with comprehensive systematic review of studies of
dementia are older people and across all OECD countries dementia prevalence around the world. Prevalence by
more than one in every 16 people aged over 60 are living country has been estimated by applying these age-
with the condition. Prevalence varies between countries: specific prevalence rates for the relevant region of the
Italy, Japan and Germany all have more than 20 people with world to population estimates from the UN (World
dementia per 1 000 population, while the Slovak Republic, Population Prospects: The 2012 Revision). Although
Korea, Mexico and Turkey have fewer than ten (Figure 11.9). gender-specific prevalence rates were available for some
regions, the overall rates were used in this analysis.
Much of the variation in prevalence is due to the age struc-
Prevalence rates are assumed to be constant over
tures of the populations in different countries, since demen-
time.
tia is strongly linked to age. Across all OECD countries, around
1.3% of people aged 60-64 have dementia, compared to nearly
45% of those aged over 90 (Figure 11.10). Age-specific prev-
alence is similar across most countries, although studies in References
Latin America have found higher rates than in other
regions (Prince et al., 2013). While this may be partly due to Matthews, F.E. et al. (2013), A Two-decade Comparison of
differences in study design, it has also been suggested that Prevalence of Dementia in Individuals Aged 65 Years and
low educational levels among older people and high vascu- Older from Three Geographical Areas of England: Results
lar risk could be contributing to increased rates of demen- of the Cognitive Function and Ageing Study I and II, The
tia (Rizzi et al., 2014). Lancet, Vol. 382, No. 9902.
If the age-specific prevalence of dementia remains the OECD (2015), Addressing Dementia: The OECD Response, OECD
same, ageing populations mean that it will become more Health Policy Studies, OECD Publishing, Paris,
common in the future. Prevalence will rise more quickly in http://dx.doi.org/10.1787/9789264231726-en.
countries that are ageing rapidly. For example, the next Prince, M. et al. (2013), The Global Prevalence of Dementia:
20 years will see prevalence in Japan rise from 21 to nearly A Systematic Review and Metaanalysis, Alzheimers &
37 per 1 000 people; and in Korea prevalence will more than Dementia, Vol. 9, No. 2, pp. 63-75.
double from 8 to 20 per 1 000 people (Figure 11.9). The over- Rizzi, L. et al. (2014), Global Epidemiology of Dementia:
all number of people living with dementia in OECD coun- Alzheimers and Vascular Types, BioMed Research Inter-
tries will rise from 18 million in 2015 to nearly 31 million in national, Vol. 2014, Article ID 908915, 8 pages.
2035, with the oldest people (aged over 90) accounting for
WHO (2015), Dementia, Fact Sheet No. 362,
an increasing share (Figure 11.11). However, there is some
www.who.int/mediacentre/factsheets/fs362/en/.
evidence that the age-specific prevalence of dementia may
be falling in some countries (Matthews et al., 2013) and it Wimo, A. et al. (2013), The Worldwide Economic Impact of
may be possible to reduce the risk of dementia through Dementia 2010, Alzheimer's & Dementia, Vol. 9, No. 1,
healthier lifestyles and preventive interventions. If such pp. 1-11.

Dementia prevalence
11.9. Estimated prevalence of dementia per 1 000 population, 2015 and 2035
2015 2035
40
35
30
25
20
15
10
a
A gal
C z c ela s
Ir e i l e
Po nd
Br a
M z il
do a
Ru I n d
Sl an el
Ko .
Tu co
ey
a
h ia
F i en
P o ium
ng d
d ry
ec nd
Ch .
No ds
Ne Es ia
Ze ia
Un Hun nd
a k d.
Ja l y
rm n
Fr ny
Gr c e
Sp e
S w a in
B e and
i te ze ria
De dom
er k
OE ay
Lu Ca 3 4
m da
st g
Sl r a li a
p
I te
In C h i n
si
re
ric
c
th ar
Ki n
Ge pa
A u o ur
u t In d
s s sr a
w ton
en
It a
Re
Re
ov F e
i te ga
rk
rw
an
ee
ed
xe n a
i
a
la
la
d rla
a
Un S w i us t
CD
ne
r tu
ex
N e nm
la
al
nl
lg
Af
ov
St
b
i
t
So
Source: OECD analysis of data from Prince et al. (2013) and the United Nations.
1 2 http://dx.doi.org/10.1787/888933281401
11.10. Age-specific prevalence of dementia across all OECD countries, 2015

% of population
50
40
30
20
10
0
60-64 65-69 70-74 75-79 80-84 85-89 90+
1 2 http://dx.doi.org/10.1787/888933281401
11.11. Estimated number of people with dementia in all OECD countries, by age, 1995, 2015 and 2035
Million population
10
2035
6
2015
4
1995
2
0
60-64 65-69 70-74 75-79 80-84 85-89 90+
1 2 http://dx.doi.org/10.1787/888933281401

As people age, they are more likely to develop disabilities where older people move into specially adapted houses
and need support from family, friends and long-term care where 24/7 care is available. This model of care allows peo-
(LTC) services. As a result, while LTC services are delivered ple with relatively severe needs to retain more indepen-
to younger disabled groups, the majority of LTC recipients dence and autonomy than they would in a traditional care
are older people. On average across the OECD, more than institution.
half of all LTC recipients are aged over 80 and nearly four in
five are aged over 65 (Figure 11.12). Rising life expectancies
mean that older people make up an increasing proportion
of the populations of OECD countries. The risk of dementia Definition and comparability
(see indicator on Dementia prevalence) and other debili-
LTC recipients are defined as persons receiving long-
tating conditions increases with age, so demand for LTC
term care by paid providers, including non-profes-
services is likely to increase although this effect may be
sionals receiving cash payments under a social pro-
partially offset by improving health in old age. As a result,
gramme. They also include recipients of cash benefits
the average proportion of the population receiving LTC in
such as consumer-choice programmes, care allow-
OECD countries has risen from 1.9% in 2000 to 2.3% in 2013.
ances or other social benefits which are granted with
While population ageing is a significant driver of the the primary goal of supporting people with long-term
growth in LTC users over time, it explains relatively little of care needs. LTC institutions refer to nursing and resi-
the cross-country variation. For example, Portugal has a dential care facilities which provide accommodation
relatively old population but only a small proportion receiv- and long-term care as a package. LTC at home is
ing formal LTC. By contrast, Israel has one of the youngest defined as people with functional restrictions who
populations in the OECD but a greater than average propor- receive most of their care at home. Home care also
tion receiving LTC. A more important driver is the availabil- applies to the use of institutions on a temporary basis,
ity of publicly funded LTC services. Countries with strong community care and day-care centres and specially
public provision, such as the Netherlands and Nordic coun- designed living arrangements. Data for Iceland and
tries, report the greatest number of LTC recipients as a Canada are only available for people receiving care in
share of their populations, while countries with limited institutions, so the total number of recipients will be
public provision, such as the United States, Portugal and underestimated.
Poland, report much smaller numbers. However, data for
Concerning the number of people receiving LTC in
people receiving care outside of public systems are more
institutions, the estimate for Ireland is under-
difficult to collect and may be underreported, meaning that
reported. Data for Japan underestimate the number of
figures for countries that rely more heavily on privately-
recipients in institutions because hospitals also pro-
funded care may be artificially low. Cultural norms around
vide LTC. In the Czech Republic, LTC recipients refer to
the degree to which families look after older people may
recipients of the care allowance (i.e., cash allowance
also be an important driver of the utilisation of formal ser-
paid to eligible dependent persons). Data for Poland
vices (see indicator on Informal carers).
only refer to services in nursing homes. Data in Spain
In response to most peoples preference to receive LTC ser- only refer to a partial coverage of facilities or services.
vices at home, many OECD countries have over the past In Australia, the data do not include recipients who
decade implemented programmes and benefits to support access the Veterans Home Care Program and those
home-based care, in particular for older people. In most who access services under the National Disability
countries for which trend data are available, the propor- Agreement, as it is currently unknown how many of
tion of LTC recipients aged 65 and over receiving long- these people could be included in LTC recipients.
term care at home has increased over the past ten years Australia collects data on users of aged care, but this
(Figure 11.13), with particularly large increases in Sweden, does not distinguish those using services on a long-
France and Korea. Often this is the result of specific poli- term basis, so the figures presented here are esti-
cies: for example, Sweden has reduced its institutional care mated. With regard to the age threshold in chart 11.13,
capacity in an effort to encourage community care; while data for France refer to people aged over 60.
France has adopted a multi-year plan to increase home
nursing care capacity to 230 000 by 2025 (Colombo et al.,
2011).
While the proportion of LTC recipients living at home has
increased over the past decade in most OECD countries, it References
has declined from 69% to 60% in Finland. However, this Colombo, F. et al. (2011), Help Wanted? Providing and Paying
does not represent an increase in the use of traditional for Long-Term Care, OECD Publishing, Paris,
institutions, but an increase in the use of service housing http://dx.doi.org/10.1787/9789264097759-en.

11.12. Proportion of population receiving long-term care, 2013 (or nearest year)
0-64 65-79 80+ Total for all ages Total for all ages in 2000 (or nearest year)
% of total population
5
4.5
4.3
4.2
4 3.8
3.3
3.2 3.0
3 2.9 2.9
2.8
2.7 2.7 2.7
2.5
2.3
2.2 2.2
2.0 2.0
2
1.6
1.5
1.2
1 0.8
0.7 0.7
0.5 0.4
0.4
0.3
l
s
nd
da
en
ay
nd
es
nd
k
el
g
d
ia
p.
21
li a
p.
ria
ly
ain
ga
nd
an
ni
re
ar
ar
an
pa
an
an
ur
ra
en
It a
Re
Re
iu
at
rw
ra
ed
na
la
la
la
CD
st
r tu
Sp
to
Ko
nm
ng
bo
rm
la
el
Ja
nl
al
Is
lg
ov
St
Po
Ir e
er
st
Au
Sw
Ca
No
Es
h
ak
Ic
er
Ze
Fi
OE
Be
Hu
Po
Au
it z
Ge
De
ec
Sl
d
ov
th
xe
i te
w
Cz
Sw
Ne
Sl
Ne
Lu
Un
1 2 http://dx.doi.org/10.1787/888933281419
11.13. Share of long-term care recipients aged 65 years and over receiving care at home, 2000 and 2013 (or nearest year)
2000 2013
% of total LTC recipients aged 65 years and over
90
80 77.5 76.6
71.2 70.8 70.1 70.0

69.4 69.3
70 67.9 69.1 67.8
65.7 66.3
65.0 64.9
63.0
61.3 60.0
60 58.7 58.9
55.4 55.9
51.1 52.2
50 48.7
40.8 42.1
40.0
40
30
20
10
0
nd
es
y
ay
13
en
ce
li a
nd
an
re
ar
pa
an
ur
at
rw
an
ra
ed
CD
la
Ko
ng
bo
rm
la
Ja
nl
St
er
st
Fr
Sw
No
er
OE
Fi
Hu
Au
it z
Ge
d
th
xe
i te
Sw
Ne
Lu
Un

1 2 http://dx.doi.org/10.1787/888933281419

Informal carers
Family and friends are the most important source of care higher poverty rates, and a higher prevalence of mental
for people with LTC needs in OECD countries. Because of health problems. Many OECD countries have implemented
the informal nature of care that they provide, it is not easy policies to support family carers with a view to mitigating
to get comparable data on the number of people caring for these negative impacts. These include paid care leave (e.g.,
family and friends across countries, nor on the frequency Belgium), flexible work schedules (e.g., Australia and the
of their caregiving. The data presented in this section come United States), respite care (e.g., Austria, Denmark and
from national or international health surveys and refer to Germany) and counselling/training services (e.g., Sweden).
people aged 50 years and over who report providing care Moreover, a number of OECD countries provide cash bene-
and assistance to family members and friends. fits to family caregivers or cash-for-care allowances for
On average across OECD countries, around 15% of people recipients which can be used to pay informal caregivers
aged 50 and over provided care for a dependent relative or (Colombo et al., 2011).
friend in 2013 (Figure 11.14). There is significant variation Declining family size, increased geographical mobility and
between countries, with nearly 20% of over-50s in Belgium rising participation rates of women in the labour market
and Estonia providing informal care, compared to just over mean that there is a risk that fewer people will be willing
10% in Israel and Australia. Rates of informal care are neg- and able to provide informal care in the future. This could
atively correlated with the proportion of older people have two consequences. Firstly, those that do provide infor-
receiving formal services (see indicator on Recipients of mal care may be required to provide higher-intensity care.
informal care) and the density of LTC workers (see indica- This will make the support that they receive even more
tor on Long-term care workers). Countries such as Estonia important if negative health and employment outcomes
and the Czech Republic, with relatively few LTC workers and are to be avoided. Secondly, a reduction in the supply of
recipients of formal services, have higher rates of informal informal care would put increasing pressure on public LTC
care; while countries such as Israel and Sweden, with large systems. These systems will need adequate funding and
number of LTC workers and many older people receiving LTC infrastructure in place to cope with increased demand, oth-
services, have lower rates of informal care. The causality erwise people could be left without access to the services
here is not clear: it could be that strong public provision they need.
means families do not have to care for older people with LTC
needs, or it could be that a strong tradition of family support
reduces the need for extensive public provision.
The majority of informal carers are women in all OECD Definition and comparability
countries and on average more than 60% of carers are
women. This ranges from a high of 70% in Slovenia to a low Family carers are defined as people providing daily or
of 55% in Sweden (Figure 11.15). weekly help to family members, friends and people in
their social network living in their household or out-
On average across OECD countries, 74% of informal carers
side of the household who require help for Activities
provide care on a daily basis, while the remaining 26% pro-
of Daily Living (ADL) and Instrumental Activities of
vide care only on a weekly basis. However, there is wide
Daily Living (IADL). The data relate only to the popula-
variation across countries in the intensity in caregiving
tion aged 50 and over, and are based on national or
(Figure 11.16). In countries with comprehensive public
international health surveys. Survey results may be
LTC systems, such as the Netherlands, Switzerland and
affected by reporting biases or recall problems. Data
Nordic countries, family and friends provide less intensive
for Australia are limited to those providing assistance
care. The highest intensity of care is reported in Spain,
with mobility, self-care, and communication, so may
Slovenia and Israel although these countries actually
be underestimated relative to other countries.
have relatively few people providing informal care
(Figure 11.14). Taking the total number of carers into
account, Estonia, Belgium, the Czech Republic and France
have greatest proportion of over-50s providing daily care to
family or friends, suggesting that informal care is particu- References
larly important in these countries.
Colombo, F. et al. (2011), Help Wanted? Providing and Paying
Intensive caregiving is associated with a reduction in for Long-Term Care, OECD Publishing, Paris,
labour force attachment for caregivers of working age, http://dx.doi.org/10.1787/9789264097759-en.

Informal carers
11.14. Population aged 50 and over reporting to be informal carers, 2013 (or nearest year)
%
30
25
19.7 19.0
20 18.1 17.5
16.5 15.9 15.7 15.4 14.9 14.7 14.4
15 13.5 12.9 12.3
11.4 11.0 10.8 10.4
10
0
m
nd
p.
ce
ria
ly
17
a in
en
el
ia
li a
nd
an
ni
ar
ur
ra
en
It a
Re
iu
do
an
ra
ed
CD
la
st
to
Sp
nm
bo
rm
la
Is
lg
ov
er
st
ng
Au
Fr
Sw
Es
er
OE
Be
Au
it z
Ge
De
ec
Sl
Ki
th
xe
Cz
Sw
Ne
d
Lu
i te
Un
Source: OECD estimates based on 2013 HILDA survey for Australia, 2012-13 Understanding Society survey for the United Kingdom and 2013 SHARE
survey for other European countries.
1 2 http://dx.doi.org/10.1787/888933281423
11.15. Share of women among all informal carers aged 50 and over, 2013 (or nearest year)
%
100
90
80
70.1
70 64.8 64.5 63.5 63.3 63.2 61.9 61.7 61.5 61.1 61.0 60.0 59.0 58.5
60 57.9 56.1 55.9 54.8
50
40
30
20
10
0
en
nd
a in
el
m
y
ria
k
ia
li a
ce
ly
p.
17
nd
an
ni
ar
ur
ra
en
It a
Re
iu
do
an
ra
ed
CD
la
st
to
Sp
nm
bo
rm
la
Is
lg
ov
er
st
ng
Au
Fr
Sw
Es
er
OE
Be
m
Au
it z
Ge
De
ec
Sl
Ki
th
xe
Cz
Sw
Ne
d
Lu
i te
Un
Source: OECD estimates based on 2013 HILDA survey for Australia, 2012-13 Understanding Society survey for the United Kingdom and 2013 SHARE
survey for other European countries.
1 2 http://dx.doi.org/10.1787/888933281423
11.16. Frequency of care provided by informal carers, 2013

Daily Weekly
% of carers
100
90 14 14 16 21 21 22 23 23 24 26 28 29 33
80 38 39 41
70
60
50
40 86 86 84 79 79 78 77 77 76 74 72 71 67
30 62 61 59
20
10
0
m
ain
el
ia
ria
ly
15
ce
p.
nd
en
nd
an
ni
ar
ur
ra
en
It a
Re
iu
an
ed
CD
la
st
Sp
to
nm
bo
rm
la
Is
lg
ov
er
Au
Fr
Sw
Es
er
OE
Be
m
it z
Ge
De
ec
Sl
th
xe
Cz
Sw
Ne
Lu
Source: OECD estimates based on 2013 SHARE survey.


Long-term care (LTC) is a labour-intensive service. Formal Increasing demand for LTC services and a possible decline
LTC workers are defined as paid staff, typically nurses and in the availability of family caregivers mean that demand
personal carers, providing care and/or assistance to people for LTC workers is likely to rise. Responding to increasing
limited in their daily activities at home or in institutions, demand will require policies to improve recruitment (e.g.
excluding hospitals. Formal care is complemented by infor- encouraging more unemployed people to consider training
mal, usually unpaid, support from family and friends, and working in the LTC sector); improve retention (e.g.
which accounts for a large part of care for older people in enhancing pay and work conditions); and increase produc-
all OECD countries (see indicator on Informal carers). tivity (e.g. through reorganisation of work processes and
Relative to the population aged 65 and over, Sweden and more effective use of new technologies) (Colombo et al.,
the United States have the most LTC workers and Turkey 2011; European Commission, 2013).
and Portugal the least (Figure 11.17). In all countries except
for Israel, Japan, Estonia and Korea, the majority of LTC
staff work in institutions, even though the majority of
recipients usually receive care at home (see indicator on Definition and comparability
Care recipients). This reflects the fact that those in insti-
tutions often have more severe needs and require more Long-term care workers are defined as paid workers
intensive care. who provide care at home or in institutions (outside
hospitals). They include qualified nurses and personal
Most LTC workers are women and work part-time. Over
care workers providing assistance with ADL and other
90% of LTC workers are women in Canada, Denmark, the
personal support. Personal care workers include dif-
Czech Republic, Ireland, Korea, New Zealand, the Slovak
ferent categories of workers who may be called under
Republic, the Netherlands, Norway and Sweden. Foreign-
different names in different countries. They may have
born workers also play an important role in LTC provision,
some recognised qualification or not. Because per-
although their presence is uneven across OECD countries.
sonal care workers may not be part of recognised
While Germany has very few foreign-born LTC workers,
occupations, it is more difficult to collect comparable
nearly one in four care workers in the United States is for-
data for this category of LTC workers across countries.
eign-born (Colombo et al., 2011). The recruitment of for-
LTC workers also include family members or friends
eign-born workers can help respond to growing demand for
who are employed under a formal contract either by
LTC, but growing migrant inflows have raised issues around
the care recipient, an agency, or public and private
the management of irregular migration, and paid work
care service companies. They exclude nurses working
which is undeclared for tax and social security purposes.
in administration. The numbers are expressed as
The LTC sector represents a small but growing share of head counts, not full-time equivalent.
total employment in OECD countries, averaging just over
The data for Italy exclude workers in semi-residential
2%. This share has increased over the past decade in many
long-term care facilities. The data for Japan involve
countries, with the broadening of public provision and
double-counting (as some workers may work in more
increased demand for services. In Japan, the number of LTC
than one home). The data for Ireland refer only to the
workers has more than doubled since 2001, following the
public sector. The data for Australia are estimates
implementation of a universal LTC insurance programme
drawn from the 2011 National Aged Care Workforce
in 2000 and government policies to professionalise LTC
Census and Survey, and underrepresent the numbers
work, while there was a slight decrease in total employ-
of people who could be considered LTC workers.
ment over this period. Similarly, LTC employment in
Germany has outstripped the growth in total employment
since 2001. In contrast, LTC employment in Sweden and the
Netherlands countries which already had comprehensive
LTC systems and high employment in the sector in the References
early 2000s has roughly followed trends in overall employ-
Colombo, F. et al. (2011), Help Wanted? Providing and Paying
ment (Figure 11.18).
for Long-Term Care, OECD Publishing, Paris,
On average, around 30% of LTC workers are nurses and the http://dx.doi.org/10.1787/9789264097759-en.
other 70% are personal care workers (also referred to
European Commission (2013), Long-term Care in Ageing
as nursing aides, health assistants in institutions or home-
Societies Challenges and Policy Options Commission
based care assistants) with less formal training. Since qual-
Staff Working Document, SWD 41, Brussels.
ity of care depends on all staff having appropriate skills,
many OECD countries have set educational and training OECD and European Commission (2013), A Good Life in Old
requirements for personal care workers, although these Age? Monitoring and Improving Quality in Long-term Care,
vary substantially, especially where home-based care is OECD Health Policy Studies, OECD Publishing, Paris,
concerned (OECD/European Commission, 2013). http://dx.doi.org/10.1787/9789264194564-en.

11.17. Long-term care workers per 100 people aged 65 and over, 2013 (or nearest year)
Institutions Home Institutions + home
Workers per 100 people aged 65+
14
12
2.6
10
8
3.9 2.7
3.3
6 12.3
9.9 2.7
9.4 2.9 1.1
4 4.5 1.7 1.6
5.8 5.2
5.5 5.2
2 4.3 4.2 4.2 2.1 0.8 1
3.2 3 2.8
1.4 1.6 1.4 1.3 0.1
0.8 1.1 0.1
0
en 1
l
el
nd
y
k
li a
14
nd
a
ria
p.
p.
es
ga
ain
nd
an
ke
ni
re
ar
ar
pa
ra
Re
Re
ra
at
CD
la
la
st
r tu
to
Ko
ed
nm
ng
r
rm
la
Ja
Is
Sp
Tu
Ir e
er
st
St
Au
Es
ak
er
OE
Sw
Hu
Po
Au
it z
De
Ge
ec
d
ov
th
i te
Cz
Sw
Ne
Sl
Un
1. In Sweden, Spain and the Slovak Republic, it is not possible to distinguish LTC workers in institutions and at home.
1 2 http://dx.doi.org/10.1787/888933281433
11.18. Trends in long-term care employment and total employment, selected OECD countries, 2001-13
Long-term care Total employment
Japan Germany
Index (2000 = 100) Index (2001 = 100)
250 250
200 200
150 150
100 100
50 50
0 0
2001 2003 2005 2007 2009 2011 2013 2001 2003 2005 2007 2009 2011 2013
Netherlands Sweden
Index (2000 = 100) Index (2003 = 100)
250 250
200 200
150 150
100 100
50 50
0 0
2001 2003 2005 2007 2009 2011 2013 2001 2003 2005 2007 2009 2011 2013

1 2 http://dx.doi.org/10.1787/888933281433

The number of beds in long-term care (LTC) institutions need to travel to each person separately. However, from the
and in LTC departments in hospitals provides a measure of point of view of public budgets, it often costs more, since
the resources available for delivering LTC services to indi- informal carers make less of a contribution and LTC sys-
viduals outside of their home. tems often pick up board and lodging costs as well as care
On average across OECD countries, there were 45 beds in costs. Moreover, LTC users generally prefer to remain at
LTC institutions and five beds in LTC departments in hospi- home and most countries have taken steps in recent years
tals per 1 000 people aged 65 and over in 2013 (Figure 11.19). to support this preference and promote community care
Belgium had the highest number of LTC beds in 2013, with (see Figure 11.13). However, depending on individual cir-
around 72 beds per 1 000 people aged 65 and over in LTC cumstances, a move to LTC institutions may be the most
institutions. On the other hand, there were fewer than appropriate option, for example for people living alone and
20 beds per 1 000 people aged 65 and over in LTC institu- requiring round the clock care and supervision (Wiener et
tions or in hospitals in Italy and Poland. al., 2009) or people living in remote areas with limited
home-care support. It is therefore important that countries
On average across all OECD countries, there has been a slight
retain an appropriate level of residential LTC capacity, and
increase in the number of LTC beds per 1 000 population
that care institutions develop and apply models of care that
over 65 since 2000 (Figure 11.20). This increase consists
promote dignity and autonomy.
entirely of beds in LTC institutions, with the number of
hospital beds remaining constant on average. However, this
masks a lot of variation. At one extreme, some countries
with well-established, comprehensive LTC systems have Definition and comparability
been reducing residential LTC capacity. Sweden has
reduced the number of LTC beds by 2.4 per year per 1 000 Long-term care institutions refer to nursing and resi-
population over 65, as part of a drive to move LTC out of dential care facilities which provide accommodation
residential facilities and into the community (Colombo et and long-term care as a package. They include spe-
al., 2011). The Netherlands, Denmark and Norway have also cially designed institutions or hospital-like settings
made significant reductions in the number of beds avail- where the predominant service component is long-
able. At the other end of the scale, Korea has seen a mas- term care for people with moderate to severe functional
sive increase in capacity since 2000, adding 4.5 beds per restrictions. Beds in adapted living arrangements for
year per 1 000 population over 65, with the increase partic- persons who require help while guaranteeing a high
ularly marked since the introduction of a public LTC insur- degree of autonomy and self-control are not included.
ance scheme in 2008. In contrast to many other countries, a For international comparisons, they should not
significant proportion of the LTC beds added in Korea are in include beds in rehabilitation centers.
hospitals. Spain has also increased its number of LTC beds However, there are variations in data coverage across
significantly, although all of the additional beds are in LTC countries. Several countries only include beds in pub-
institutions rather than hospitals. licly-funded LTC institutions, while others also
While most countries allocate very few beds for LTC in hos- include private institutions (both profit and non-for-
pitals, others still use hospital beds quite extensively for profit). Some countries also include beds in treatment
LTC purposes. Despite recent increases in the number of centers for addicted people, psychiatric units of gen-
beds in LTC institutions in Korea, the majority of LTC beds eral or specialised hospitals, and rehabilitation cen-
are still in hospitals. In Japan many hospital beds are used ters. Australia does not collect data on the numbers of
for long-term care, but recently the number has been beds provided for LTC. Data on Australian LTC beds in
decreasing. Some European countries, such as Finland, institutions are estimated from aged care database.
Hungary and Estonia, still have a significant number of LTC
beds in hospitals, but in general there has been a move
towards replacing hospital beds with institutional facilities,
References
which are often cheaper and provide a better living envi-
ronment for people with LTC needs. Finland, France and Colombo, F. et al. (2011), Help Wanted? Providing and Paying
Iceland have all seen significant increases in LTC beds in for Long-Term Care, OECD Publishing, Paris,
institutions and decreases in hospital LTC beds since 2000 http://dx.doi.org/10.1787/9789264097759-en.
although in the case of Iceland, this is partly due to Wiener, J. et al. (2009), Why Are Nursing Home Utilization
changes in how beds are categorised. Rates Declining?, Real Choice System Change Grant
Providing LTC in institutions can be more efficient than Program, US Department of Health and Human Services,
community care for people with intensive needs, due to Centres for Medicare and Medicaid Services, available at
economies of scale and the fact that care workers do not www.hcbs.org/files/160/7990/SCGNursing.pdf.

11.19. Long-term care beds in institutions and hospitals, 2013 (or nearest year)
Institutions Hospitals
Per 1 000 population aged 65 and over
80
72.1 71.1
70.6
70 67.6 67.4
65.5
59.5 59.0
60 58.1
56.6 55.8
54.5 54.3 54.2 54.0 53.7
53.1
50.2 49.7 49.5
50 48.9 48.9 47.9
45.6 45.0
40 38.8
36.6
35.1
30 28.8
23.4
20 18.9 18.0
10
Li es
m
d
en
nd
ak s
Un O nd
ia
Po y
d
p.
ce
m d
Hu g
Sl r y
ay
a
No ia
li a
Ge da
Ir e y
Ki 21
Es k
a
a in
ec ia
d p.
el
nd
ia
De om
nd
an
ni
e
ar
l
an
an
pa
n
ur
ra
an
en
C z s tr
tv
It a
Re
e
iu
at
rw
n
r
ra
ed
na
la
la
la
Lu eala
to
Sp
Un h R
Ko
nm
d
ng
bo
rm
Ne Fr a
la
el
nl
Ja
La
Is
lg
ov
i te EC
tu
St
er
st
ng
Au
Sw
Ca
Ic
er
Fi
Be
Au
it z
Z
ov
th
xe
i te
w
Sw
Ne
d
Sl

1 2 http://dx.doi.org/10.1787/888933281448
11.20. Trends in long-term care beds in institutions and in hospitals, 2000-13 (or nearest year)
Hospitals Institutions Net annual change
Average annual change in beds per 1 000 population aged 65 and over
5
4.5
4 3.7
2
1.4 1.3
1.2
0.9
1 0.6
0.5 0.4 0.4 0.3 0.2 0.2
-0.1 -0.1 -0.1
0 -0.3 -0.3 -0.5 -0.5 -0.5
-0.7
-0.8 -0.9
-1.1
-1
-2 -2.3
-3
es
nd
s
a
a in
li a
ly
ia
el
p.
ia
g
17
ce
nd
da
ay
en
nd
an
re
ni
ar
ar
an
an
pa
ur
ra
an
It a
tv
Re
iu
at
rw
an
ra
ed
na
CD
la
la
to
Sp
Ko
nm
ng
bo
rm
la
el
nl
Ja
La
Is
lg
tu
St
Po
er
st
Fr
Sw
Ca
No
Es
h
Ic
er
OE
Fi
Be
Hu
m
Li
Au
it z
Ge
De
ec
th
xe
i te
Cz
Sw
Ne
Lu
Un
Note: The OECD average includes only countries with data for both institutions and hospitals.
1. Australia, Germany, Luxembourg, the Netherlands, Norway and Switzerland do not report any long-term care beds in hospital.
1 2 http://dx.doi.org/10.1787/888933281448

Long-term care (LTC) expenditure has risen over the past Many OECD countries have expanded the availability of
few decades in most OECD countries and is expected to rise home care services in order to allow people receiving LTC to
further in the coming years, with population ageing leading remain more independent and part of their community.
to more people needing ongoing health and social care, ris- Between 2005 and 2013, the annual growth rate of public
ing incomes leading to higher expectations of quality of life spending on home care matched spending growth for care in
in old age, the supply of informal care potentially shrinking institutional care settings at 4.3% per year (Figure 11.23).
and productivity gains difficult to achieve in such a labour- However, there were significant increases in home care
intensive sector (De La Maisonneuve and Oliveira Martins, spending of more than 7% per year in Korea, Estonia, Japan
2013). and France.
A significant share of LTC services is funded from public Projection scenarios suggest that public resources allocated
sources. Total public spending on LTC (including both the to LTC as a share of GDP could double or more by 2060
health and social care components) accounted for 1.7% (Colombo et al., 2011; De La Maisonneuve and Oliveira
of GDP on averag e across OECD countries in 2013 Martins, 2013). One of the main challenges in many OECD
(Figure 11.21). The highest spender was the Netherlands, countries in the future will be to strike the right balance
where public expenditure on long-term care was two and a between providing appropriate social protection to people
half times greater than the OECD average, at 4.3% of GDP. At with LTC needs and ensuring that this protection is fiscally
the other end of the scale, the Slovak Republic, Greece, sustainable.
Estonia, Hungary, the Czech Republic, Poland and Israel
allocated less than 0.5% of their GDP to public provision of
long-term care. This variation partly reflects differences in
population structure, but mostly the development of for- Definition and comparability
mal LTC systems, as opposed to more informal arrange-
ments based mainly on care provided by unpaid family LTC spending comprises both health and social sup-
members. Despite the problems of underreporting, pri- port services to people with chronic conditions and
vately-funded LTC expenditure plays a relatively large role disabilities needing care on an on-going basis. Based
in Switzerland (0.6% of GDP), Germany (0.6%) and Belgium on the System of Health Accounts (SHA), the health
(0.4%). As a share of total spending on LTC (including pri- component of LTC spending relates to nursing and
vate and public health and social components), private personal care services (i.e. assistance with activities of
spending accounts for more than a third in the United daily living (ADL)). It covers palliative care and care
States (43%), Germany (37%) and Spain (36%). Most private provided in LTC institutions or at home. LTC social
spending is out-of-pocket, since private LTC insurance does expenditure primarily covers assistance with instru-
not play an important role in any country. mental activities of daily living (IADL). Countries
The boundaries between health and social LTC spending reporting practices between the health and social
are still not fully consistent across countries, with some components of LTC spending may differ. In addition,
reporting particular components of LTC as health care, publicly-funded LTC expenditure is more suitable for
while others view it as social spending. The Netherlands, international comparisons as there is significant vari-
Sweden, Norway and Denmark spend over 2% of GDP on ation in the reporting of privately-funded LTC expen-
the health part of LTC, which is double the OECD average. diture across OECD countries.
Finland has the highest level of public spending on social Data for the United States refer to institutional care
LTC, reaching 1.6% of GDP, much higher than the OECD only, so underestimate the total amount of public
average of 0.5%. The Netherlands and Japan spend more spending on long-term care services.
than 1% of GDP on social LTC, but this accounts for less
than 0.1% of GDP in Korea, Spain and Luxembourg.
Public spending on LTC has grown rapidly in recent years in
some countries (Figure 11.22). The annual growth rate in
References
public expenditures on LTC was 4.0% between 2005 and
2013 across OECD countries, which is above the growth in Colombo, F. et al. (2011), Help Wanted? Providing and Paying
health care expenditures over the same period. Countries for Long-Term Care, OECD Publishing, Paris,
such as Korea and Portugal have implemented measures to http://dx.doi.org/10.1787/9789264097759-en.
expand the comprehensiveness of their LTC systems in De La Maisonneuve, C. and J.O. Martins (2013), Public
recent years and so have among the highest public spend- Spending on Health and Long-term Care: A New Set of Pro-
ing growth rates since 2005, although spending in both jections, OECD Economic Policy Papers, No. 6, OECD Publish-
countries remains relatively low as a share of GDP. ing, Paris, http://dx.doi.org/10.1787/5k44t7jwwr9x-en.

11.21. Long-term care public expenditure (health and social components), as share of GDP, 2013 (or nearest year)
Health LTC Social LTC
% GDP
5
4.3
3.2
3
2.4 2.3 2.2
2.1
1.9 1.9
2 1.8 1.8
1.7
1.3
1.2 1.2
1.0 0.9
1 0.7 0.7
0.5 0.5 0.4 0.4
0.3 0.2 0.2
0.0 0.0
0
es 1
nd
l
s
ic
d
el
en
ay
n
d
ce
11
a
g
ria
da
ia
nd
ic
ce
ga
nd
an
ni
re
ar
ar
ai
an
an
pa
ur
bl
ra
bl
en
iu
rw
an
ee
ed
na
CD
la
la
st
r tu
to
Sp
at
Ko
nm
ng
bo
rm
pu
pu
la
nl
Ja
el
Is
lg
ov
Po
er
Au
Gr
Fr
Sw
Ca
No
St
Es
Ic
er
OE
Fi
Be
Hu
m
Po
Re
Re
it z
De
Ge
Sl
th
d
xe
Sw
i te
ak
Ne
Lu
ec
Un
ov
Cz
Sl
Note: The OECD average only includes the eleven countries that report health and social LTC.
1. Figures for the United States refer only to institutional care.
1 2 http://dx.doi.org/10.1787/888933281455
11.22. Annual growth rate in public expenditure 11.23. Annual growth rate in public expenditure on
on long-term care (health and social), in real terms, long-term care (health), by setting, in real terms, 2005-13
2005-13 (or nearest year) (or nearest year)
Korea 36.1 Institution LTC Home LTC
Switzerland 13.8
Korea 36.3
Portugal 10.8 57.2
Estonia 7.6
Estonia 8.1 15.6
Belgium 6.1 Japan 2.2
8.7
Japan 5.6 France 4.3
7.0
Czech Rep. 5.0
Finland -3.4
6.4
Norway 4.5 3.7
Norway 5.7
OECD22 4.0
3.5
Spain 5.4
Spain 3.9
Belgium 6.6
Netherlands 3.9 5.1
OECD18 4.3
Luxembourg 3.8 4.3
Switzerland 12.8
Finland 3.8 3.8
Austria 3.4 Germany 1.4
3.7
Germany 2.4 Canada 1.6
3.2
Poland 2.3
Czech Rep. 5.6
2.8
Slovenia 2.2
Austria 4.7
2.6
Canada 1.8 -1.4
Hungary 2.1
United States 1.8
2.0
Denmark 1.6
Denmark 1.8
3.8
Iceland 1.4 Netherlands 1.4
2.3
Sweden 1.2 Slovenia 0.6
5.3
Hungary -0.3 Luxembourg 0.5
Slovak Rep. -0.5 14.3
Poland 0.5
-10 0 10 20 30 40 -10 0 10 20
Average annual growth rate (%) Average annual growth rate (%)
Note: The OECD average excludes Korea (due to the extremely high Note: The OECD average excludes Korea (due to the extremely high
growth rate). growth rate).

OECD 2015
ANNEX A
Additional information on demographic and economic

context, and health expenditure and financing
211
ANNEX A. ADDITIONAL INFORMATION ON DEMOGRAPHIC AND ECONOMIC CONTEXT, AND HEALTH EXPENDITURE AND FINANCING
Table A.1. Total population, mid-year, 1970 to 2014

Thousands
1970 1980 1990 2000 2010 2011 2012 2013 2014
Australia 12 507 14 695 17 065 19 028 22 032 22 340 22 724 23 132 23 524
Austria 7 467 7 549 7 678 8 012 8 363 8 392 8 430 8 479 8 503
Belgium 9 656 9 859 9 967 10 251 10 896 11 048 11 128 11 183 11 284
Canada 21 745 24 518 27 691 30 687 34 127 34 484 34 880 35 317 35 540
Chile 9 570 11 174 13 179 15 398 17 094 17 248 17 403 17 557 17 819
Czech Republic 9 858 10 304 10 333 10 255 10 474 10 496 10 511 10 514 10 527
Denmark 4 929 5 123 5 141 5 340 5 548 5 571 5 592 5 615 5 597
Estonia 1 360 1 477 1 569 1 397 1 331 1 327 1 323 1 318 1 316
Finland 4 606 4 780 4 986 5 176 5 363 5 388 5 414 5 439 5 460
France 50 772 53 880 56 709 59 062 62 918 63 223 63 514 63 790 64 360
Germany1 61 098 61 549 63 202 82 212 81 777 81 798 80 426 80 646 80 925
Greece 8 793 9 643 10 157 10 917 11 153 11 103 11 037 10 948 11 381
Hungary 10 338 10 711 10 374 10 211 10 000 9 972 9 920 9 893 9 843
Iceland 204 228 255 281 318 319 321 324 327
Ireland 2 957 3 413 3 514 3 805 4 560 4 577 4 587 4 598 4 610
Israel 2 958 3 878 4 660 6 289 7 624 7 766 7 910 8 057 8 186
Italy 53 822 56 434 56 719 56 942 59 277 59 379 59 540 60 234 60 789
Japan 103 721 117 061 123 613 126 927 128 058 127 799 127 515 127 296 127 083
Korea 32 241 38 124 42 869 47 008 49 410 49 779 50 004 50 220 50 424
Luxembourg 339 364 382 436 507 518 531 543 556
Mexico 50 628 66 737 87 065 100 896 114 256 115 683 117 054 118 395 119 713
Netherlands 13 039 14 150 14 952 15 926 16 615 16 693 16 755 16 804 16 858
New Zealand 2 828 3 170 3 390 3 858 4 366 4 404 4 433 4 472 4 388
Norway 3 876 4 086 4 241 4 491 4 889 4 953 5 019 5 080 5 137
Poland 32 664 35 574 38 111 38 259 38 043 38 063 38 063 38 040 38 037
Portugal 8 680 9 766 9 983 10 290 10 573 10 558 10 515 10 457 10 375
Slovak Republic 4 538 4 980 5 299 5 389 5 391 5 398 5 408 5 413 5 416
Slovenia 1 725 1901 1998 1989 2049 2053 2057 2060 2062
Spain 33 815 37 439 38 850 40 263 46 577 46 743 46 773 46 620 45 943
Sweden 8 043 8 311 8 559 8 872 9 378 9 449 9 519 9 600 9 699
Switzerland 6 181 6 319 6 716 7 184 7 825 7 912 7 997 8 089 8 188
Turkey 35 294 44 522 56 104 67 393 73 142 74 224 75 176 76 148 76 903
United Kingdom 55 663 56 314 57 248 58 893 62 766 63 259 63 700 64 107 64 091
United States 205 052 227 225 249 623 282 162 309 326 311 583 313 874 316 129 318 892
OECD (total) 870 967 965 259 1 052 204 1 155 498 1 236 028 1 243 502 1 249 052 1 256 518 1 264 123
Partners
Brazil 96 078 118 563 146 593 171 280 193 253 194 933 196 526 198 043 199 492
China (Peoples Rep.) 814 423 984 122 1 165 429 1 280 429 1 359 822 1 368 440 1 377 065 1 385 567 1 393 784
Colombia .. .. 34 130 40 296 45 510 46 045 46 582 47 121 47 662
Latvia 2 359 2 512 2 663 2 368 2 098 2 060 2 034 2 013 1 994
Lithuania 3 140 3 413 3 698 3 500 3 097 3 028 2 988 2 958 3 163
India 555 064 698 721 868 891 1 042 262 1 205 625 1 221 156 1 236 687 1 252 140 1 267 402
Indonesia 114 080 145 510 178 633 208 939 240 677 243 802 246 864 249 866 252 812
Russian Federation 130 392 138 655 147 969 146 597 142 849 142 961 143 207 143 507 143 787
South Africa 22 502 29 077 36 793 44 846 51 452 51 949 52 386 52 776 53 140
1. Population figures for Germany prior to 1991 refer to West Germany.

1 2 http://dx.doi.org/10.1787/888933281510

Table A.2. Share of the population aged 65 and over, 1970 to 2014
1970 1980 1990 2000 2010 2011 2012 2013 2014
Australia 8.3 9.6 11.1 12.4 13.6 13.8 14.2 14.4 14.7
Austria 14.0 15.5 14.8 15.4 17.6 17.6 17.8 18.0 18.3
Belgium 13.3 14.3 14.8 16.7 17.1 17.0 17.3 17.5 17.7
Canada 7.9 9.4 11.3 12.6 14.2 14.5 14.9 15.2 15.6
Chile 5.0 5.5 6.1 7.2 9.0 9.3 9.5 9.8 10.0
Czech Republic 12.0 13.6 12.5 13.8 15.3 15.6 16.2 16.8 17.3
Denmark 12.1 14.3 15.6 14.8 16.3 16.8 17.3 17.8 18.3
Estonia 11.6 12.5 11.6 14.9 17.5 17.5 17.7 18.1 18.4
Finland 9.0 11.9 13.3 14.8 17.0 17.5 18.1 18.7 19.3
France 12.8 14.0 13.9 16.0 16.8 16.9 17.3 17.7 ..
Germany 13.0 15.5 15.2 16.2 20.7 20.6 21.0 21.1 20.8
Greece 11.1 13.0 13.6 16.4 19.0 19.3 19.8 20.2 19.7
Hungary 11.5 13.5 13.2 15.0 16.6 16.8 16.9 17.2 17.6
Iceland 8.7 9.8 10.5 11.5 12.0 12.3 12.6 12.9 13.1
Ireland 11.1 10.7 11.4 11.1 11.2 11.5 11.9 12.2 12.6
Israel 6.7 8.6 9.1 9.8 9.9 10.0 10.3 10.7 10.9
Italy 10.7 13.1 14.7 18.1 20.4 20.5 20.8 21.0 21.4
Japan 7.1 9.1 12.1 17.4 23.0 23.3 24.1 25.1 26.0
Korea 3.1 3.8 5.1 7.2 11.0 11.4 11.8 12.2 12.7
Luxembourg 12.4 13.6 13.3 14.2 13.8 13.7 13.8 13.8 14.1
Mexico 4.6 4.3 4.3 5.2 6.2 6.3 6.4 6.5 6.7
Netherlands 10.1 11.4 12.7 13.5 15.3 15.5 16.2 16.8 17.3
New Zealand 8.4 9.7 11.2 11.8 13.0 13.3 13.8 14.2 14.8
Norway 12.8 14.6 16.3 15.2 14.8 15.0 15.3 15.6 15.8
Poland 8.2 10.1 9.9 12.1 13.6 13.6 14.0 14.4 14.9
Portugal 9.2 11.1 13.2 16.0 18.3 18.7 19.1 19.4 19.3
Slovak Republic 9.1 10.5 10.2 11.4 12.4 12.6 12.8 13.1 13.5
Slovenia 9.5 11.3 10.6 13.8 16.5 16.5 16.8 17.1 17.5
Spain 9.5 10.8 13.4 16.7 16.8 17.1 17.4 17.7 18.4
Sweden 13.5 16.2 17.7 17.3 18.0 18.4 18.7 19.0 19.3
Switzerland 11.2 13.8 14.5 15.2 16.7 16.8 17.1 17.3 17.5
Turkey 4.3 4.7 4.2 5.3 7.0 7.2 7.3 7.5 7.7
United Kingdom 12.9 14.9 15.7 15.8 16.2 16.4 16.7 17.1 17.6
United States 9.8 11.3 12.5 12.4 13.1 13.3 13.7 14.1 14.5
OECD34 9.8 11.4 12.0 13.4 15.0 15.2 15.5 15.9 16.2
Partners
Brazil 3.5 4.0 4.4 5.4 6.8 7.0 7.2 7.4 7.6
China (Peoples Rep.) 4.0 5.1 5.8 6.9 8.4 8.5 8.7 8.9 9.1
Colombia .. .. 5.0 5.7 6.7 6.9 7.0 7.2 7.3
India 3.3 3.6 3.9 4.4 5.1 5.1 5.2 5.3 5.4
Indonesia 3.3 3.6 3.8 4.7 5.0 5.1 5.1 5.2 5.3
Latvia 11.9 13.0 11.8 14.9 18.3 18.5 18.7 18.9 18.7
Lithuania 9.9 11.3 10.8 13.8 17.6 18.0 18.2 18.3 17.2
Russia 7.7 10.2 10.0 12.4 12.8 12.7 12.9 13.0 13.3
South Africa 3.4 3.1 3.2 3.4 5.2 5.3 5.4 5.5 5.6

1 2 http://dx.doi.org/10.1787/888933281523

Table A.3. GDP per capita in 2013 and average annual growth rates, 1970 to 2013
GDP per capita in
Average annual growth rate per capita, in real terms
USD PPP
2013 1970-80 1980-90 1990-2000 2000-10 2010-2013
Australia 44 976 1.3 1.5 2.4 1.6 1.3

Austria 45 082 3.5 2.1 2.2 1.1 0.9
Belgium 41 573 3.2 1.9 2.0 0.9 -0.2
Canada 42 839 2.8 1.4 1.8 0.8 1.1
Chile 22 178 .. .. 5.0 3.1 4.5
Czech Republic 28 739 .. .. 0.6 3.0 0.0
Denmark 43 782 1.9 2.0 2.3 0.3 -0.4
Estonia 25 823 .. .. .. 3.9 5.2
Finland 39 869 3.4 2.7 1.9 1.3 -0.5
France 37 671 3.0 2.0 1.7 0.6 0.4
Germany1 43 887 2.8 2.1 1.3 1.0 1.8
Greece 25 854 3.6 0.2 1.7 1.5 -5.9
Hungary 23 336 .. .. .. 2.2 1.0
Iceland 42 035 5.2 1.6 1.6 1.5 1.8
Ireland 45 677 3.2 3.3 6.3 0.6 0.6
Israel 32 502 .. 1.9 2.9 1.4 1.6
Italy 35 075 3.3 2.3 1.6 -0.1 -1.8
Japan 36 236 3.2 4.1 0.9 0.7 1.2
Korea 33 089 7.4 8.6 6.0 3.9 2.4
Luxembourg 91 048 1.9 4.5 3.6 1.1 -0.8
Mexico 16 891 3.7 -0.9 2.0 0.6 1.9
Netherlands 46 162 2.3 1.7 2.5 0.9 -0.6
New Zealand 34 899 1.0 1.2 1.7 1.3 1.5
Norway 65 640 4.1 1.2 4.0 0.9 1.6
Poland 23 985 .. .. 3.7 4.0 2.7
Portugal 27 509 3.5 3.0 2.6 0.5 -2.1
Slovak Republic 26 497 .. .. .. 4.8 1.8
Slovenia 28 859 .. .. 1.9 2.4 -1.2
Spain 33 092 2.6 2.6 2.4 0.7 -1.3
Sweden 44 646 1.6 1.9 1.8 1.5 0.4
Switzerland 56 940 1.0 1.6 0.5 1.0 0.5
Turkey 18 508 .. .. 1.8 3.0 3.6
United Kingdom 38 255 2.0 2.7 2.1 1.1 0.6
United States 53 042 2.1 2.4 2.2 0.7 1.3
OECD 38 123 2.9 2.3 2.4 1.6 0.7
Partners
Brazil 16 192 .. -0.6 0.8 2.4 1.9
China (People's Rep.) 11 661 .. 7.7 9.3 9.9 7.7
Colombia 12 695 .. 1.5 1.0 2.8 4.0
Costa Rica 13 872 .. .. .. 2.6 3.0
India 5 406 .. 3.3 3.5 5.9 4.8
Indonesia 10 023 .. 3.4 2.6 3.9 4.4
Latvia 22 958 .. .. .. 5.2 6.1
Lithuania 25 715 .. .. . 5.4 6.0
Russian Federation 25 247 .. .. .. 5.1 2.8
South Africa 12 553 .. -0.8 -0.1 1.9 1.0
1. Data prior to 1991 refers to Western Germany.

Source: OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en. International Monetary Fund, World
Economic Outlook Database, April 2015.
1 2 http://dx.doi.org/10.1787/888933281533

Table A.4. Health expenditure per capita in 2013, average annual growth rates,
2009 to 2013
Health expenditure
per capita in USD Annual growth rate per capita in real terms1
PPP
2013 2009/10 2010/11 2011/12 2012/13 2005-13
Australia2 3 866 -0.6 3.8 2.9 .. 2.5

Austria 4 553 1.5 0.5 2.3 -0.3 1.6
Belgium 4 256 -0.8 2.7 0.1 0.1 1.8
Canada 4 351 2.0 -1.3 0.3 0.1 1.9
Chile3 1 606 6.0 5.1 6.1 8.3 5.9
Czech Republic 2 040 -3.1 2.5 -0.1 -0.2 2.5
Denmark 4 553 -1.4 -1.4 0.2 -0.5 1.3
Estonia 1 542 -4.3 0.8 .. 4.4 3.9
Finland 3 442 1.6 2.3 0.8 0.2 1.5
France 4 124 0.8 .. 0.6 1.2 1.2
Germany 4 819 3.0 0.8 2.7 1.7 2.4
Greece 2 366 -10.9 -2.8 -12.2 -2.5 -2.3
Hungary 1 719 4.4 1.9 -2.2 -0.6 -0.8
Iceland 3 677 -6.1 0.1 1.3 3.4 0.0
Ireland2 3 663 -8.7 -4.1 1.1 .. 1.2
Israel 2 428 3.1 2.9 5.7 2.8 2.7
Italy 3 077 1.1 -0.9 -3.0 -3.5 -0.6
Japan 3 713 5.2 4.9 3.0 .. 3.7
Korea 2 275 8.1 4.0 4.4 5.3 7.2
Luxembourg2 4 371 -2.2 -5.8 -5.0 .. -2.1
Mexico 1 048 1.3 -2.1 5.9 2.0 1.7
Netherlands 5 131 2.3 1.7 3.2 -0.3 2.5
New Zealand 3 328 0.4 0.8 2.7 -1.3 2.4
Norway4 5 862 -0.1 2.6 1.9 0.6 1.6
Poland 1 530 .. 2.0 1.2 3.8 5.8
Portugal 2 514 1.1 -4.8 -5.0 -3.2 -0.9
Slovak Republic 2 010 .. -2.4 4.4 0.0 6.7
Slovenia 2 511 0.9 0.1 -0.8 -1.4 1.4
Spain 2 898 -0.1 -0.6 -2.4 -3.8 1.0
Sweden 4 904 -0.3 .. 1.4 2.0 1.4
Switzerland 6 325 .. 2.1 3.5 1.9 1.9
Turkey 941 -1.2 1.2 -0.7 5.4 3.0
United Kingdom 3 235 -1.3 -0.1 0.3 0.6 1.7
United States 8 713 1.9 1.0 1.6 1.5 1.9
OECD 3 453 0.1 0.6 0.8 0.9 2.0
Partners
Brazil5 1 471 7.7 2.4 .. .. 4.2
China (People's Rep.)5 649 6.1 12.3 12.5 .. 12.0
Colombia5 864 -1.0 1.9 7.4 .. 5.9
Costa Rica5 1 380 .. .. .. .. ..
India5 215 .. .. .. .. ..
Indonesia5 293 9.2 3.8 11.8 5.3 6.2
Latvia 1 216 -1.8 -1.8 2.5 3.7 1.7
Lithuania 1 573 -3.7 3.4 1.9 1.3 4.9
Russian Federation 5 1 653 -4.0 1.4 -0.3 1.8 6.3
South Africa5 1 121 1.9 2.0 5.3 1.3 2.1
1. Using national currency units at 2005 GDP price level.

2. Latest year 2012.
3. CPI is used as deflator.
4. GDP deflator refers to Mainland Norway.
5. Including investment.
1 2 http://dx.doi.org/10.1787/888933281543

Table A.5. Expenditure on health, percentage of GDP, 1980-2013

1980 1990 2000 2010 2011 2012 2013
Australia 5.8 6.5 7.6 8.5 8.6 8.8 e ..

Austria 7.0 | 7.7 9.2 10.1 9.9 10.1 10.1
Belgium 6.2 7.1 | 8.0 | 9.9 10.1 10.2 10.2 e
Canada 6.6 8.4 8.3 10.6 10.3 10.2 10.2
Chile .. .. 6.4 6.7 6.7 7.0 7.3
Czech Republic .. 3.8 | 5.7 | 6.9 7.0 7.1 7.1
Denmark 8.4 8.0 8.1 | 10.4 10.2 10.4 10.4
Estonia .. .. 5.2 6.1 5.7 | 5.8 6.0
Finland 5.9 7.2 | 6.7 8.2 8.2 8.5 8.6
France 6.7 8.0 | 9.5 | 10.8 | 10.7 10.8 10.9
Germany 8.1 8.0 | 9.8 11.0 10.7 10.8 11.0
Greece .. 6.0 7.2 9.2 e 9.7 e 9.1 e 9.2 e
Hungary .. .. | 6.8 | 7.7 7.6 7.5 7.4
Iceland 5.8 7.4 9.0 8.8 8.6 8.7 8.7
Ireland 7.5 5.6 5.6 8.5 8.0 8.1 ..
Israel 7.0 6.6 6.8 7.0 7.0 7.4 e 7.5 e
Italy .. 7.0 | 7.6 8.9 8.8 8.8 8.8
Japan 6.4 5.8 7.4 9.5 10.0 10.1 10.2 e
Korea 3.5 3.7 4.0 6.5 6.5 6.7 6.9
Luxembourg .. .. 5.9 7.2 6.8 6.6 ..
Mexico .. 4.3 | 4.9 | 6.2 5.9 6.1 6.2
Netherlands 6.6 7.1 | 7.0 | 10.4 10.5 11.0 11.1
New Zealand 5.7 6.7 7.5 | 9.7 e 9.7 e 9.8 e 9.5 e
Norway 5.4 7.1 | 7.7 | 8.9 8.8 8.8 8.9
Poland .. 4.3 5.3 | 6.5 6.3 6.3 6.4
Portugal 4.8 5.5 | 8.3 9.8 9.5 9.3 9.1
Slovak Republic .. .. 5.3 | 7.8 7.5 7.7 7.6
Slovenia .. .. 8.1 | 8.6 8.5 8.7 8.7
Spain 5.0 6.1 | 6.8 | 9.0 9.1 9.0 8.8
Sweden .. 7.3 | 7.4 | 8.5 | 10.6 10.8 11.0
Switzerland 6.6 7.4 | 9.3 | 10.5 10.6 11.0 11.1
Turkey 2.4 2.5 | 4.7 5.3 5.0 5.0 5.1
United Kingdom 5.1 5.1 6.3 8.6 8.5 8.5 8.5
United States 8.2 11.3 12.5 16.4 16.4 16.4 16.4
OECD 6.1 6.5 7.2 8.8 8.8 8.9 8.9
Partners
Brazil1 .. .. 7.0 8.7 8.7 8.9 9.1
China (People's Rep.)1 .. .. 4.6 5.0 5.1 5.4 5.6
Colombia1 .. .. 5.9 6.8 6.5 6.8 6.8
Costa Rica1 .. .. 7.1 9.7 10.2 10.1 9.9
India1 .. .. 4.3 3.8 3.9 3.9 4.0
Indonesia1 .. .. 1.8 2.7 2.7 2.9 2.9
Latvia .. .. .. 6.1 5.6 5.4 5.3
Lithuania .. .. .. 6.8 6.5 6.3 6.1
Russian Federation1 .. .. 5.4 6.9 6.7 6.5 6.5
South Africa1 .. .. 8.3 8.7 8.6 8.9 8.9
| Break in series.
e: Preliminary estimate.
1. Including investment.
1 2 http://dx.doi.org/10.1787/888933281551

ORGANISATION FOR ECONOMIC CO-OPERATION
AND DEVELOPMENT
The OECD is a unique forum where governments work together to address the economic,
social and environmental challenges of globalisation. The OECD is also at the forefront of efforts to
understand and to help governments respond to new developments and concerns, such as
corporate governance, the information economy and the challenges of an ageing population. The
Organisation provides a setting where governments can compare policy experiences, seek answers
to common problems, identify good practice and work to co-ordinate domestic and international
policies.
The OECD member countries are: Australia, Austria, Belgium, Canada, Chile, the Czech
Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Israel,
Italy, Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal,
the Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the
United States. The European Union takes part in the work of the OECD.
OECD Publishing disseminates widely the results of the Organisations statistics gathering and
research on economic, social and environmental issues, as well as the conventions, guidelines and
standards agreed by its members.
OECD PUBLISHING, 2, rue Andr-Pascal, 75775 PARIS CEDEX 16

(81 2015 07 1 P) ISBN 978-92-64-23257-0 2015
OECD INDICATORS
This new edition of Health at a Glance presents the most recent comparable data on the performance of health
systems in OECD countries. Where possible, it also reports data for partner countries (Brazil, China, Colombia,
Costa Rica, India, Indonesia, Latvia, Lithuania, Russian Federation and South Africa). Compared with the
previous edition, this new edition includes a new set of dashboards of health indicators to summarise in a clear
and user-friendly way the relative strengths and weaknesses of OECD countries on different key indicators of
health and health system performance, and also a special focus on the pharmaceutical sector. This edition also
contains new indicators on health workforce migration and on the quality of health care.
Contents
Chapter 1. Dashboards of health indicators
Chapter 2. Special focus: Pharmaceutical spending trends and future challenges
Chapter 3. Health status
Chapter 4. Non-medical determinants of health
Chapter 5. Health workforce
Chapter 6. Health care activities
Chapter 7. Access to care
Chapter 8. Quality of care
Chapter 9. Health expenditure and financing
Chapter 10. Pharmaceutical sector
Chapter 11. Ageing and long-term care
Consult this publication on line at http://dx.doi.org/10.1787/health_glance-2015-en.

This work is published on the OECD iLibrary, which gathers all OECD books, periodicals and statistical databases.
Visit www.oecd-ilibrary.org for more information.
ISBN 978-92-64-23257-0
81 2015 07 1 P
9HSTCQE*cdcfha+

Health at A Glance 2015 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Health at A Glance 2015 PDF

Uploaded by

Copyright:

Available Formats

Health at a Glance 2015

Please cite this publication as:

ISBN 978-92-64-23257-0 (print)

Series: Health at a Glance

Photo credits: Cover Oleksiy Mark/Shutterstock.com; Chapter 3: Comstock/Jupiterimages; Chapter 4: Comstock/

Corrigenda to OECD publications may be found on line at: www.oecd.org/about/publishing/corrigenda.htm.

HEALTH AT A GLANCE 2015 OECD 2015 3

Chapter 1. Dashboards of health indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Chapter 2. Pharmaceutical spending trends and future challenges . . . . . . . . . . . . . . . . 29

Chapter 3. Health status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

Chapter 4. Non-medical determinants of health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

HEALTH AT A GLANCE 2015 OECD 2015 5

Chapter 5. Health workforce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Chapter 6. Health care activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

Chapter 7. Access to care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Chapter 8. Quality of care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

Chapter 9. Health expenditure and financing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

6 HEALTH AT A GLANCE 2015 OECD 2015

Expenditure by disease and age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

Chapter 10. Pharmaceutical sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177

Chapter 11. Ageing and long-term care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Annex A. Additional information on demographic and economic context,

Follow OECD Publications on:

This book has... StatLinks2

HEALTH AT A GLANCE 2015 OECD 2015 7

H ealth at a Glance 2015 presents cross-country comparisons of the health status of

New drugs will push up pharmaceutical spending unless policy adapts

Life expectancy continues to rise, but widespread differences persist across

HEALTH AT A GLANCE 2015 OECD 2015 9

Out-of-pocket spending remains a barrier to accessing care

10 HEALTH AT A GLANCE 2015 OECD 2015

HEALTH AT A GLANCE 2015 OECD 2015 11

Content of the publication

HEALTH AT A GLANCE 2015 OECD 2015 13

Non-medical determinants of health

Health care system performance

Health care resources and activities

Health workforce Health care activities

Demographic and economic context, and health expenditure and financing

Source: Adapted from Kelley, E. and J. Hurst (2006).

Chapter 8 examines quality of care or the degree to which care is delivered in

HEALTH AT A GLANCE 2015 OECD 2015 15

then choose Health). More information on OECD Health Statistics is available at

OECD country ISO codes

Australia AUS Japan JPN

Partner country ISO codes

Brazil BRA Indonesia IDN

16 HEALTH AT A GLANCE 2015 OECD 2015

Dashboards of health indicators

18 HEALTH AT A GLANCE 2015 OECD 2015

HEALTH AT A GLANCE 2015 OECD 2015 19

Risk factors to health

20 HEALTH AT A GLANCE 2015 OECD 2015

Health care resources

HEALTH AT A GLANCE 2015 OECD 2015 21

22 HEALTH AT A GLANCE 2015 OECD 2015

Table 1.1. Health status

HEALTH AT A GLANCE 2015 OECD 2015 23

Table 1.2. Risk factors

Alcohol Overweight and

24 HEALTH AT A GLANCE 2015 OECD 2015

Table 1.3. Access to care

HEALTH AT A GLANCE 2015 OECD 2015 25