Blood Transfusion:

New Guidelines

Joint Surgery and Anesthesiology Grand Rounds

July 2, 2009
Paul Picton MD
Lena M. Napolitano MD
Andrew Rosenberg MD

Perioperative Transfusion
Triggers

Paul Picton MD MRCP FRCA

Assistant Professor
Director, Transplant Anesthesia

Changes in cardiac output (A) oxygen extraction (B) oxygen delivery (C) and
oxygen consumption (D) as hemoglobin decreases in humans and animals
Klein HG, et al. Lancet 2007; 370:415-426

Anemia in Healthy Awake

Volunteers
Critical hemoglobin threshold unknown in
humans
At 5 g/dL - VO2 maintained but ST
changes (5%) and memory formation
impaired
At 6 g/dL - decline in cognitive function
Lieberman, et al. Anesthesiology 2000

Do Nothing Study
Retrospective study of 300 JW who underwent surgery
from 1981 - 1994
Even after adjusting for age, cardiovascular disease and
APACHE score, odds of death increased by 2.5 times for
each gram of Hb below 8 g/dL

Transfusion. 2002 Jul;42(7):812-8

Do Nothing Study
Retrospective study of 300 JW who underwent surgery
from 1981 - 1994
Even after adjusting for age, cardiovascular disease and
APACHE score, odds of death increased by 2.5 times for
each gram of Hb below 8 g/dL

Transfusion. 2002 Jul;42(7):812-8

The TRICC Study

Enrolled 838 euvolemic, anemic, critically ill pts
who were admitted to 1 of 25 Canadian ICUs
Patients were stratified according to center and
disease severity (APACHE II) and placed into
one of two groups
Restrictive group: Transfuse if Hb < 7 and maintain
between 7 and 9
Liberal group: Transfuse if Hb < 10 and maintain
between 10 and 12

The primary outcome measure was death from

all causes in the 30 days after randomization
Herbert PC, et al. NEJM 1999

TRICC - Design

The TRICC Study

No difference 30 day mortality
In healthy (APACHE II < 20) and young
(<55yrs) patients
Transfusion increased mortality

Herbert PC, et al. NEJM 1999

The TRICC Study

8.7% vs 16.1%

5.7% vs 13.0%
Herbert PC, et al. NEJM 1999

The TRICC Study

Average red cell units per patient:
2.6 4.1 vs. 5.6 5.3 (p < 0.01)
Average daily Hb concentrations:
8.5 0.7 g/dl vs. 10.7 0.7 g/dl (p < 0.01)

TRICC Sub Group Analyses

Trauma (n = 203)
McIntyre LA, et al. J Trauma 2004;57:563-568

Moderate to severe head injury (n = 67)

McIntyre LA, et al. Neurocrit Care 2006;05:4-9

Cardiovascular disease (n = 357)

Herbert PC, et al. Crit Care Med 2001; 29(2):227-234

Mechanical ventilation (n = 713)

Hebert PC, et al. Chest 2001 June;119(6):1851.

No difference in outcomes

A restrictive red blood cell transfusion

strategy generally appears to be safe in
most critically ill patients with cardiovascular
disease
with the possible exception of
patients with acute myocardial infarction and
unstable angina.

CRIT Study
Prospective, multiple center, observational
cohort study of 4,892 ICU pts in the US
Propensity score matched
Designed to examine the relationship of anemia
and RBC transfusion with clinical outcomes
Almost 95% of patients admitted to the ICU have
a Hb level below normal by day 3
In total, 11,391 RBC units were transfused.
Overall, 44% of pts admitted to the ICU received
one or more RBC units while in the ICU
Crit Care Med. 2004 Jan;32(1):39-52

CRIT Results
35% of Blood transfused in
patients with Hgb 9

The mean pre-transfusion Hb

was 8.6 1.7 g/dL

RBC transfusion was independently

associated with higher mortality (OR
1.65 CI 1.35-2.03). OR 2.62 if 3-4 units
transfused p < 0.0001

Hematocrit versus Postop Morbidity & Ischemia

ST

n = 27 high-risk pts
undergoing infra-inguinal
arterial bypass

Sx

Nelson A, Fleischer L, et al. Crit Care Med 1993

2001
Retrospective cohort
Cooperative Cardiovascular Project
78,974 patients 65 yrs acute MI
30 day mortality

Blood transfusion associated with mortality if Hct <

30%

 Analysis of 24,112 enrollees in 3 large

international trials of patients with acute
coronary syndromes
Association between transfusion and
outcome
Cox proportional hazards modeling
Main outcome = 30 day mortality

Rao SV et al. JAMA. 2004;292:1555-1562

Blood Transfusion and Clinical

Outcome in Acute Coronary Syndrome
Transfusion
Adjusted
hazard ratio
3.94
(3.26-4.75)

No Transfusion

Rao SV et al. JAMA. 2004;292:1555-1562

 Meta-analysis of observational studies

45 studies - 272,596 patients
Multivariate analysis correcting for age and
illness severity
Outcome measures:

Mortality
Infection
Multi-organ dysfunction
ARDS
Crit Care Med 2008;36(9):2667-74

Results

Association between blood

transfusion and the risk of
death (OR & 95% CI). Pooled
OR 1.7 (95% CI 1.4-1.9)
Crit Care Med 2008;36(9):2667-74

Association between blood

transfusion and the risk of
infectious complications (OR
& 95% CI). Pooled OR 1.8
(95% CI 1.5-2.2)

Results
Association
between blood
transfusion and
the risk of ARDS
(OR & 95% CI).
Pooled OR 2.5
(95% CI 1.6-3.3)
Crit Care Med 2008;36(9):2667-74

Millions

Financial Burden

Based on data from UMHHC cost accounting

system.
Cost includes cost of blood products and allocated

Summary
Post op Hct 15 - very high mortality
At Hct 18 - cognitive dysfunction in healthy
volunteers
Utilization of a transfusion trigger 21 (mean Hct
25) - confers survival benefit for those < 55 yrs
and those with an APACHE < 20
A liberal transfusion policy - trigger 30 (mean
Hct 32) does not benefit patients on critical care
At Hct 27 - ST changes in high risk patients.

Summary
Transfusion may benefit patients during
acute coronary syndromes if Hct < 25-29
There is only rarely an indication to
transfuse ANY patient with a Hct 30
Blood transfusions are not risk free
Decreasing transfusion may not only
decrease cost but also improve outcome

Closing Comments
Good prospective data limited to critical
care setting
Considerable scope for differences in
opinion
Concerning intra-operative transfusion best to come to some agreement pre op
and remain in communication
Give RBCs as single units when possible
Treat the patient not the Hct

Univ. Michigan Adult Blood

Transfusion Guidelines: 2009

Lena M. Napolitano MD, FACS, FCCP, FCCM

Professor of Surgery
Division Chief, Acute Care Surgery
Department of Surgery
University of Michigan
Ann Arbor, MI

Adult Blood Transfusion

Clinical Guidelines

Plan and Guideline endorsed by ECCA on March 24, 2009

Project Overview & Scope Of Work

Blood Utilization lean project work was commissioned by both OCA &
Hospital Administration, under the oversight of Dr. Skip Campbell
Team Make-Up
Dr. Tim Laing, Internal Medicine/OCA

Dr. Rob Davenport, Blood Bank

Lena Napolitano, MD-Surgeon/ICU

Bill Palazzolo, Dir. Pre-Op Clinic

Paul Picton, MD-Anest/Transplant

Shon Dwyer, AHD

Andrew Rosenburg, MD-Anest/Carelink

Vinita Bahl, SMT

Jeff Rohde, MD-Int. Medicine

Brendon Weil, Lean Coach

Ryen Fons, House Officer-Anest.

Gail Sinwell, Lean Coach

Russel Butler, Perfusion, CVC

Barb Chapman, CIDSS

Project Goal: To develop standard policies & practices leading to: improved
patient outcomes through the appropriate use of blood products and gain
process efficiencies by removing waste and delays in the blood dispensing &
administration process

Guidelines for Blood

Transfusion: PRBCs

These guidelines are intended to ensure that the most

appropriate, efficient and safe use of the blood supply is
achieved

To establish evidence-based criteria for the transfusion of

blood components

Every indication for the use of blood products cannot be

anticipated

These guidelines are not intended to override physician

judgement

Guidelines for Blood

Transfusion: PRBCs

Hemodynamically stable anemia without acute coronary syndrome:

hemoglobin trigger less than 7 g/dL, with a transfusion goal to
maintain hemoglobin 7 9 g/dL.

Acute hemorrhage with evidence of hemodynamic instability or

inadequate oxygen delivery

Symptomatic (tachycardia, tachypnea, postural hypotension) anemia (Hb

< 10 g/dL) not explained by other causes

Chronic Tx-dependent bone marrow syndromes: Hb < 10 g/dL.

Transfusion or exchange transfusion for severe sickle syndromes.

Hemodynamically stable anemia with ischemic heart disease: current

evidence does not support routine transfusion in non-ST segment
elevation acute coronary syndromes; although in ST-segment elevation
myocardial infarction Tx may be beneficial.

Guidelines for Blood

Transfusion: PRBCs

RBCs should be administered as single units for most operative

and inpatient indications (transfuse and reassess strategy) except
for ongoing blood loss with hemodynamic instability.

Tx decisions are clinical judgments that should be based on the

overall clinical assessment of the individual patient. Transfusion
decisions should not be based on laboratory parameters alone.

Routine premedication is not advised unless the patient has a

history of previous transfusion reactions. Premedication has not
been shown to reduce the risk of transfusion reactions.

EAST / SCCM Blood Tx Guidelines

CLINICAL PRACTICE GUIDELINE:
RED BLOOD CELL TRANSFUSION IN ADULT TRAUMA and CRITICAL CARE
Lena M. Napolitano MD
Stanley Kurek DO
Fred A. Luchette MD
For the EAST Practice Management Workgroup and
The American College of Critical Care Medicine Taskforce of the SCCM
The EAST Practice Management W orkgroup
Gary L. Anderson DO
Michael R. Bard MD
William Bromberg MD
William C. Chiu MD
Mark D. Cipolle MD, PhD
Keith D. Clancy MD
Lawrence Diebel MD
William S. Hoff MD
K. Michael Hughes DO
Imtiaz Munshi MD
Donna Nayduch RN, MSN, ACNP
Rovinder Sandhu MD
Jay A. Yelon MD

In press.
November 2009
Crit Care Med

The American College of Critical Care Medicine Taskforce of the SCCM

Howard L. Corwin MD
Philip S. Barie MD
Samuel A. Tisherman MD
Paul C. Hebert MD, MHSc

Risks of Blood Transfusion

Viral transmission

Acute transfusion reactions

Immunosuppression

Acute inflammatory response

Noninfectious Hazards

Immunosuppression

Infection

Risk of Infection per

Unit Transfused

Decline in HIV, HBV, HCV Risks

of Transmission via Blood Tx
1:100

HIV

HCV
HBV

1:1000
1:10,000
1:100,000
1:1,000,000
1:10,000,000

1983 1985
2001
Revised Donor
Deferral Criteria

1987

1989

Non-A, Non-B
Hepatitis
Surrogate Testing

HIV Antibody
Screening

Busch MP, et al. JAMA. 2003;289:959-62.

1991 1993

Year
HCV Antibody
Screening

1995

1997

p24 Antigen
Testing

1999
HCV and HIV
Nucleic Acid
Testing

Risks of Transfusion:
Infectious Disease

HIV = 1 in 1.8 million

HCV = 1 in 1.6 million

HBV = 1 in 220,000

HIV = human immunodeficiency virus.

HCV = hepatitis C virus.
HBV = hepatitis B virus.
Busch MP, et al. JAMA. 2003;289:959-62.

Serious Hazards of Transfusion

Post-transfusion
purpura
Acute lung injury
Graft vs host
disease

2%

Delayed
transfusion
reaction

3%

Transfusion-transmitted
infections

6%
8%

14%

53%

Incorrect blood/
component
transfused

15%
Acute
transfusion
reaction
Williamson LM, et al. BMJ. 1999;319:16-9.

Based on 366 spontaneously-reported

deaths/major complications between
October 1996 and September 1998
in the UK and Ireland.

Risks of Blood Transfusion

Minor allergic reactions
Bacterial infection (platelets)
Viral hepatitis
Hemolytic transfusion reaction
HTLV I/II infection
Acute lung injury
TRALI
Anaphylactic shock
Fatal hemolytic reaction
Graft-vs-host disease
Immunosuppression
HTLV = human T-cell leukemia-lymphoma virus.
Klein HG. Am J Surg. 1995. 170;6A(suppl):21S-26S.

1:5,000

1:100
1:2,500
1:5,000
1:6,000
1:200,000
1:500,000
1:500,000
1:600,000
Rare
Unknown

Immune Effects of Blood

Immunologic effects of autologous and

allogenic blood transfusions:
- Decreased T-cell proliferation
- Decreased CD3, CD4, CD8 T-cells
- Increased Soluble cytokine receptor
- sTNF-R, sIL-2R

Increased Serum neopterin

Increased Cell-mediated lympholysis
Increased TNF-alpha
- Increased suppressor T-cell activity
- Reduced natural killer cell activity
-

TRIM Transfusion-associated Immunomodulation

McAlister FA, et al, British Journal of Surgery 1998; 85: 171-178
Innerhofer et al. Transfusion 1999 Oct;39(10):1089

Blood Tx Increases Risk of

Postoperative Bacterial Infection

20 peer-reviewed studies, 1986-2000

N = 13,152 (Tx 5215, No-Tx 7937)
Association of Blood Tx to Infection
Common OR 3.45 (range 1.43-15.15)
17 of 20 studies with p < 0.05

Trauma subgroup
Common OR 5.26 (range 5.03-5.43)
All studies with p < 0.05 (0.005 0.0001)
Blood Tx associated with greater risk in trauma pts

Hill GE, Minei JP et al. J Trauma 2003;54:908-914

Prospective cohort study,

n=2085

Project Impact

Nosocomial Infections:
14.3% vs. 5.8%, p <
0.001

Taylor RW et al.
Crit Care Med 2006;
34:23022308

15,592 Cardiovascular operations

Infection endpoints bacteremia, SSI
55% of pts received PRBCs, 21% plts, 13%
FFP, 3% cryoprecipitate
Increased RBC tx associated with increased
infection (p < 0.0001), confirmed by
logistic regression analysis.

J Am Coll Surg 2006;202:131-138

Leukoreduction does not diminish tx-associated Microchimerism

Reed W, et al. Semin Hematol 2007:44:24-31

Utter G et al. Transfusion 2006 Nov;46(11):1863-9

Gould S et al. Am J Crit Care; Jan 2007;16(1):39-48

Why is blood transfusion

NOT associated with
improved outcome?

Stored RBCs

Decreased RBC deformability

Decreased 2,3, DPG

Metabolic acidosis

Altered oxygen carrying capacity

Increased red cell death with

increased age of blood (~30% dead)

No improvement in oxygen
utilization at the tissue level

Age of Blood

Poor Efficacy of Blood Tx

RBCs stored > 15 days lose deformability and ATP

Altered capillary lumen size (decreased cross-sectional

diameter) in critically ill patients

Increased stickiness (adherence) of RBCs to altered

endothelium in the microcirculation of critically ill pts.

Schechter, Gladwin, NEJM April 10, 2003

Distribution of Transfused Units

by Age of Blood CRIT Study

Percentage of Patients

60% of Blood transfused

is > 20 days old

0 - 10

10 - 20

20 - 30

30 - 40

Oldest Age of Blood in Days

In Trauma Subset, 68% of blood is > 20 days old

> 40

March 20, 2008

The median duration of storage

was 11 days for newer blood and
20 days for older blood.
Patients who were given older
units had higher rates of inhospital mortality (2.8% vs.
1.7%, P = 0.004), intubation
beyond 72 hours (9.7% vs.
5.6%, P<0.001), renal failure
(2.7% vs. 1.6%, P = 0.003), and
sepsis or septicemia (4.0% vs.
2.8%, P = 0.01).
A composite of complications
was more common in patients
given older blood (25.9% vs.
22.4%, P = 0.001).
Similarly, older blood was
associated with an increase in
the risk-adjusted rate of the
composite outcome (P = 0.03).
At 1 year, mortality was
significantly less in patients
given newer blood (7.4% vs.
11.0%, P<0.001).

Composite Outcome:
In-hospital mortality
And Complications
(STS)

Age of Blood Evaluation (ABLE)

ABLE Study-Hypothesis
The use of fresh red cells as compared to standard
issue red cells will lead to significant improvement
in morbidity and mortality

Age of Blood Evaluation (ABLE) in the

resuscitation of critically ill patients
International Study, CIHR, NIH, others
Projected n = 6800

ABLESomething about the design?

Study Design: Randomized doubleblind
controlled clinical trial.
Setting: 30 Canadian tertiary care intensive
care and trauma units. Additional study sites
in the US, UK, Europe and Australia
Study Population: 6800 critically ill or
trauma victims who require at least one red
cell unit within the first 72 hrs of acute care.

The Study Intervention

Leukoreduced RBCs
Fresh RBCs defined as 8 days or less
Primarily

for feasibility as limited biological

rationale for cut-off

Control groupstandard-issue RBCs

(average age of 21 days)
Local transfusion guidelines/practices

ABLEWhat Outcomes will we measure?

Primary outcome: 30-day all cause mortality.
Secondary outcomes:
1) Other mortality rates
2) Organ failure
3) Nosocomial infections
4) Quality of life using the SF-36 and costs
of care.

CRIT Study
(USA) [2]

Trauma
patients from
CRIT Study
(USA) [3]

TRICC
Investigators
(Canada) [4]

North Thames
Blood Interest
Group (UK)
[5]

ABA
Multicenter
Trials Group
(US, Canada)
[6]

3534

4892

576

5298

1247

666

Mean admission
hemoglobin (g/dL)

11.3 2.3

11.0 2.4

11.1 2.4

9.9 2.2

--

--

Percentage of
patients transfused
in ICU

37.0%

44.1%

55.4%

25.0%

53.4%

74.7%

Mean transfusions
per patient (units)

4.8 5.2

4.6 4.9

5.8 5.5

4.6 6.7

5.7 5.2

13.7 1.1

Mean pretransfusion
hemoglobin (g/dL)

8.4 1.3

8.6 1.7

8.9 1.8

8.6 1.3

--

9.3 0.1

Mean ICU length of

stay (days)

4.5

7.4 7.3

9.4 8.6

4.8 12.6

--

--

ICU mortality

13.5%

13.0%

--

22.0%

21.5%

--

Hospital mortality

20.2%

17.6%

9.9%

--

--

21.0%

ABC Trial
(Western
Europe) [1]

[1] Vincent JL, Baron JF, Reinhart K, et al. ABC (Anemia and Blood Transfusion in Critical Care ) Investigators. Anemia and blood transfusion in critically ill patients.
JAMA 2002;288:1499-1507.
[2] Corwin HL, Gettinger A, Pearl RG, et al. The CRIT Study: Anemia and blood transfusion in the critically ill current clinical practice in the United States. Crit Care
Med 2004;32:39-52.
[3] Shapiro MJ, Gettinger A, Corwin H, Napolitano LM, Levy M, Abraham E, Fink MP, MacIntyre N, Pearl RG, Shabot MM. Anemia and blood transfusion in trauma
patients admitted to the intensive care unit. J Trauma 2003;55:269-274.
[4] Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion Requireemtns in
Critical Care investigators, Canadian Critical Care Trials Group. N Engl J Med 1999;340:409-417.
[5] Rao MP, Boralessa H, Morgan C, et al and the North Thames Blood Interest Group. Blood component use in critically ill patients. Anaesthesia 2002 Jun;57(6):530-4.
[6] Palmieri TL, Caruso DM, Foster KN, et al and the American Burn Association (ABA) Multicenter Trials Group. Effect of blood transfusion on outcome after major burn
injury: A multicenter study. Crit Care Med 2006 Jun;34(6):1602-7.

Guidelines for Transfusion in Trauma

Studies on RBC transfusion and outcome in ischemic heart disease.

Year

Hebert

Hebert

Wu

Blood Transfusion and Clinical

Study
Design

Patients

Primary Results

1997

Retrospective

Critically ill patients with

cardiac disease, as part of a
retrospective assessment of
transfusion practices in
Canadian ICUs

Increased survival with

transfusion when Hb < 9.5
g/dL

2001

Prospective,
subgroup
analysis

357

Subgroup of patients with

cardiac disease from the
TRICC trial

No difference in mortality
Increased organ dysfunction
with transfusion

Approx
79,000

Patients aged 64 years who

had been hospitalized with a
disgnosis of acute MI,
Medicare database

Increased survival with

transfusion

Approx
24,000

Meta-analysis of data that

had been collected as part of
the GUSTO IIb, PURSUIT and
PARAGON B trials of
patients with ACS

Increased mortality,
combined death or MI

Data from 16 ACS studies

Decreased mortality in STEMI

Increased mortality in nonST-elevation ACS

Patients with non-STsegment elevation acute

coronary syndromes

Increased mortality,
combined death or MI

2001

Retrospective

Rao

2004

Retrospective

Sabatine

2005

Retrospective

Yang

2005

Retrospective

85,111
total
cohort;
74,271 no
CABG

Adapted in part from: Gerber DR. Crit Care Med 2008;36(4):1068-1074.

Studies on RBC transfusion and outcome in ischemic heart disease.

Year

Singla

Aronson

Alexander

2007

2008

2008

Study
Design

Prospective
database

Prospective
database

Primary Results

Patients with anemia and

suspected ACS receiving
transfusion, using data
prospectively collected as
part of an ongoing registry

Increased mortality, recurrent

MI

Increased mortality in
patients with nadir Hb >
8g/dL
2358

Patients with acute MI

Decreased mortality in
patients with nadir Hb <
8g/dL

Prospective
database
44242
CRUSADE
Initiative

Patients

Patients with non-STsegment elevation acute

coronary syndromes

Increased mortality in
patients with nadir
Hematocrit > 30%
Decreased mortality in
patients with nadir
Hematocrit 24%

Adapted in part from: Gerber DR. Crit Care Med 2008;36(4):1068-1074.

FOCUS

NHLBI
Transfusion
Trigger for
Functional
Outcomes in
Cardiovascular
Patients
Undergoing
Surgical Hip
Fracture Repair
N=2600
25 Med Ctrs
US, Canada
J.L. Carson MD

FOCUS

Inclusion criteria:
Undergo surgery for hip fracture
Have a history of cardiovascular disease
Have a postoperative Hgb < 10 g/dL

Randomized to keep Hgb > 10 g/dL or not

Tx permitted but not required if Hgb < 8 g/dL
Primary outcome is ability to walk 10 feet without human
assistance at 60 days
Negative outcome is postoperative unstable angina, myocardial
infarction or death
MI diagnosis based on 4 blood tests, 3 EKGs, medical history
Telephoned at 30 and 60 days to determine functional capacity
and vital status.
Long-term mortality by searching vital statistics registries in U.S.
and Canada

State of the Science Symposium in Transfusion

Medicine and Hemostasis/Thrombosis

SURGERY Committee

Jeff Carson (Chair) Clinical trials and Transfusion Medicine,

Robert Wood Johnson Medical School

Darrell Triulizi (CoChair) Transfusion Medicine, University of

Pittsburgh

John Marshall: General Surgery, Univ. Toronto

Lena Napolitano: General Surgery, Univ. Michigan

Chris Stowell: Transfusion Medicine, Mass General

Richard Weiskopf: Anesthesia, UCSF

Transfusion Triggers in CAD, Elective Cardiac Surgery

Effect of Blood
Transfusion on LongTerm Survival
After Cardiac
Operation

1915 CABG pts

After correction for
comorbidities and
other factors, tx was
still associated with
a 70% increase in
mortality (RR 1.7;
95% CI 1.4 to 2.0; p
0.001).

Engoren MC et al. (MCO, Toledo)

Ann Thorac Surg 2002;74:11806

10,289 CABG pts, 1995 2002

Perioperative RBC tx is
associated with adverse outcome.
Attention should be directed
toward blood conservation
methods and a more judicious
use of PRBC.

Cleveland Clinic, OH

0
1
2
3-5
6

Ann Thorac Surg 2006;81:1650 7

 Institution-specific protocols should screen for patients at high risk for blood
transfusion. Available evidence-based blood conservation techniques include:
(1) drugs that increase preoperative blood volume (eg, erythropoietin) or decrease
postoperative bleeding (eg, antifibrinolytics)
(2) devices that conserve blood (eg, intraoperative blood salvage and blood sparing
interventions)
(3) interventions that protect the patients own blood from the stress of operation (eg,
autologous predonation and normovolemic hemodilution)
(4) consensus, institution-specific blood transfusion algorithms supplemented with pointof-care testing, and most importantly
(5) a multimodality approach to blood conservation combining all of the above
Society of Thoracic Surgeons Blood Conservation Guideline Task Force; Society of Cardiovascular
Anesthesiologists Special task Force on Blood Transfusion. Ann Thorac Surg 2007;83:S27-86.

Do Blood Transfusions
Improve Outcome
in Sepsis?

Efficacy of Blood Tx in Sepsis

Zimmerman JL. Use of blood products in sepsis: An evidence-based review. Crit Care
Med 2004;32[Suppl]S542-547

Changes in measurements of post-transfusion

Author and Year

Study population

Amount transfused
(units)

Ronco et al 1990

PCP pneumonia

1.5 Units

Yes

Yes

Yes

NA

Fenwick et al 1990

ARDS

24

1.5 Units

Yes

Yes

No

No

Ronco et al 1991

ARDS

17

1.5 Units

Yes

Yes

No

NA

Shah et al 1982

Post-trauma

1 or 2 Units

Yes

No

No

NA

Steffes et al 1991

Postoperative and Post-trauma

21

1-2 Units

Yes

Yes

Yes

No

Babineau et al 1992

Postoperative

31

328 9 mL

Yes

Yes

No

No

Gilbert et al 1988

Septic

17

20 g/L

Yes

Yes

No

No

Dietrich et al 1990

Medical shock (septic/cardiac)

32

577 mL

Yes

Yes

No

No

Conrad et al 1990

Septic shock

19

30 g/L

Yes

Yes

No

No

Marik et al 1993

Septic

23

3 Units

Yes

Yes

No

No

Lorento et al 1993

Septic

16

2 Units

Yes

Yes

No

NA

Mink et al 1990

Septic shock
2 mo 6 y

8-10 mL/kg x 1-2 h

Yes

Yes

No

NA

Lucking et al 1990

Septic shock
4 mo 15 y

10-15 mL/kg x 1-3 h

Yes

Yes

Yes

NA

Silverman et al 1992

Septic shock
21 88 y

21

2 Units

Yes

Yes

No

No

Gramm et al 1996

Septic shock
46 3 y

19

2 Units

Yes

No

No

NA

Fernandes et al 2001

Septic shock
18-80y

10

1 Units

Yes

No

No

No

Kahn et al 1986

Acute respiratory failure

15

7-10 mL/kg

Yes

No

No

NA

Casutt et al 1999

Postoperative
32-81y

67

368 10 mL

Yes

Yes

No

NA

Walsh et al 2004

Euvolemic anemic critically ill

patients without ongoing
hemorrhage

22

2 Units

Yes

NA

NA

No

Hb

DO2

VO2

Lactate

Early Goal-directed Therapy in the Rx of

Severe Sepsis and Septic Shock

Severe sepsis and septic shock patients (n=263)

SIRS and SBP < 90mm Hg or lactate > 4mmol/L

Prospective, randomized controlled trial

Goal-directed therapy vs. control (standard of care)

Goal-directed therapy performed in ER prior to ICU

Placement of oximetric CVP line, CVP goal 8-12, ScVO2 > 70%

Guidelines for pressor and vasodilators, dobutamine, blood tx

Maintained for at least 6 hours

Rivers E et al. NEJM 345(19) November 8, 2001:1368-77

Early Goal-directed Therapy in the Rx of

Severe Sepsis and Septic Shock

Early Goal-directed Therapy resulted in:

Reduced In-hospital mortality, 30.5% vs 46.5%
(p=0.0009)

Higher ScVO2, lower lactate, lower base deficit

Early goal-directed therapy provides significant
benefits in outcome in patients with severe sepsis
and septic shock.

Rivers E et al. NEJM 345(19) November 8, 2001:1368-77

Validation Study
Multicenter Trial
20 sites
Derek Angus et al.
Univ. of Pittsburgh
ProCESS
Protocolized Care for
Early Septic Shock
NIH-sponsored
$8.4 Million

EAST/SCCM Blood Tx Guidelines

Recommendations Regarding RBC

Transfusion in Sepsis

Level 1

Level 2

There are insufficient data to support Level 1

recommendations on this topic.

The transfusion needs for each septic patient must

be assessed individually since optimal transfusion
triggers in sepsis patients are not known and there
is no clear evidence that blood transfusion
increases tissue oxygenation.

Anemia of
Chronic
Disease or
Anemia of
Inflammation

Dysregulation of iron
homeostasis
Impaired proliferation
of erythroid progenitor
cells
Blunted EPO response

Weiss and Goodnough.

N Engl J Med.
2005;352(10):1011-1023.

Blunted Epo Response in Critically Ill

Inhibition of EPO gene transcription in
renal juxtaglomerular cells
Inflammatory Cytokines
(IL-1, IFN, TNF, TGF)

Direct inhibition of RBC

production by bone marrow

Direct inhibition of the erythroid

precursor cell response to
erythropoietin

Indirect limitation of iron availability by

increasing iron sequestration in macrophages.

SICU - Patient Characteristics

2004

2005

2006

2007

2008

1491

1361

1353

1354

1275

APACHE III Score-Day 1

48.2

48.3

49.1

50.5

55.8

Hospital LOS

14.1

14

13.9

12.9

13.5

ICU-LOS

4.1

4.76

4.77

4.22

4.49

Readmissions Rates

6.2

7.9

7.1

8.4

7.4

Active Treatment

56%

51%

57%

63%

64%

Low-Risk Monitor

34%

38%

33%

27%

24%

Level of Therapy on Admission

Anemia Management Protocol

40% Reduction in Blood Tx in SICU

Oct-Dec 2004

Jul-Sep 2006

SICU Blood Utilization

Added to Keystone ICU Reports

Oct-Dec 2004

Jul-Sep 2006

SICU Blood Utilization

Added to Keystone ICU Reports

Oct-Dec 2004

Jul-Sep 2006

ICU Mortality

2007
3.36%
6.60%
7.21%

O/E

0.71

0.54

0.47

0.41

Hospital Mortality

2007
5.35%
10.89%
9.67%

O/E

0.74

0.59

0.55

0.56

Blood Dashboard for Clinical Services - DRAFT

Trend Report of Percent of RBC

Transfusions by Pre-Transfusion Hct

Current Month Snapshot of

Percent of RBC Transfusions by PreTransfusion Hct with drill-down to
Patient-Level Detail

Summary

Anemia is common

No evidence that blood tx for treatment of

anemia improves outcome

Critically ill patients can tolerate Hb levels

as low as 7 mg/dL

Blood should be transfused for physiologic

indications

New UMich Blood Tx Guidelines

UM Carelink Support for Improved

Transfusion Practice
Andrew Rosenberg MD
Medical Director, UM Carelink
Chief, Critical Care Division Anesthesiology

Clinical IT supports good decisions, best

practices & institution policies
For emergency transfusion call Blood Bank
Pre-op requests for PRBCs on standby & OR
transfusion NOT part of this process.
UMCL (UM Carelink);
Is the primary method to order blood.
Provides Clinical Decision Support {Alerts}
Serves as a useful clinical database {Queries}

Clinician feedback needed (6-2222, light bulb

icon in UMCL)

Transfusion Alert Rule Logic

Based on ECCA Transfusion Guidelines
Hemodynamically stable anemia w/out CAD

Transfusion trigger= Hg < 7g/dL

Maintain Hg 7-9g/dL

For PRBC Order set only

1 or 2 units ordered (alert will NOT fire for 3 or more units)

And Hemoglobin > 7g/dL
And/or Hg result >48 hrs old/ Or no Hg result available
And Pt age > 17 yo.

Alert Box Information

Four Alert Messages

I. Hgb <7 g/dL but last Hgb result > 48 hours.
Request does not meet ECCA Guidelines when ordering 1 or 2 units PRBC
Last HGB is over 48 hours old
HGB: ## g/dL DATE
Confirm HGB before ordering or select override reason to complete order.
II. Hgb> 7g/dL and HGB result < 48 hrs.
Transfusion may not be advised if the HGB is > 7g/dL
III. Hbg > 7 g/dL but HGB result > 48 hrs.
HGB is greater than 7 g/dL and is over 48 hours old.
IV. No Hgb result available
No HGB result on file

Override Reasons
1.
2.
3.
4.
5.
6.
7.
8.

Active Bleeding
Cardiovascular disease
Hemoglobinopathy
Hemolysis
Oxygen carrying deficit
Refractory Hypotension
Symptomatic anemia
Attending Physician deems necessary

UMHS Blood Transfusion Guideline

Order & Compliance Monitoring (Future State Map)
Blood Transfusion Guideline
*RBC Transfusion Trigger = Hemoglobin < 7

Annual
Review

Evidence
Based

Compliance
Report Capture
(If Order Is
Beyond Trigger)

Blood
Transfusion
Ordered

House Wide
Communication/
Education

Carelink Order
Checkpoints

*Redefines role/scope of
Transfusion Committee to
act as oversight body

*Reports Track Compliance To

Guideline & Transfusion Volume

Email Sent
To Service
Chief
*Email includes link to patient
level data to assist review

Response
Explanation
Submitted To
Transfusion
Committee

Response
Review By
Transfusion
Committee