You are on page 1of 2

Editorials

where they interfere with the reaction, leading to thousands of patients in the United Kingdom might be
falsely raised or lowered measurements. These affected. Furthermore, this problem is likely to worsen
interferences are specific to each patient, so only that owing to the wider use of biotechnologies such as
patient’s data will be affected, while quality assurance monoclonal antibodies and T cells for diagnostic and
criteria for the assay will have been passed. therapeutic purposes.13
Examples of this type of interference that has had Since these interferences are relatively uncommon,
serious clinical consequences include human chorionic clinicians need to be aware of them and alert to the
gonadotrophin assays.5 As a result of wrongly mismatch of clinical and biochemical data. They
interpreted results six young non-pregnant women should not discard clinical evidence in favour of a
were aggressively treated with chemotherapy and numerical value. Moreover, this form of interference is
surgery for non-existent “occult” trophoblastic disease.6 not specific to a single analyte and may affect other
In our experience at a national reference steroid immunoassays performed on the same patient in a dif-
laboratory, samples with consistent and substantial ferent clinical setting. Thus patients who have such
increases in steroids using routine direct immunoassays interference detected should have this fact docu-
have raised the possibility of disease but have mented in their clinical records, so that the results of
subsequently been found to be normal after reassay future immunoassays can be viewed with caution.
using more robust techniques involving extraction pro-
Adel AA Ismail consultant biochemist
cedures. In one case a raised oestradiol value led to a
(dr.ismail@panp-tr.northy.nhs.uk)
patient having a hysterectomy and bilateral oophorec-
tomy, and only when no fall in oestradiol was seen post- Julian H Barth director
operatively was the sample further analysed and the Leeds Supraregional Assay Service, Steroid Centre, General Infirmary,
original result found to be wrong because of immuno- Leeds LS1 3EX (j.h.barth@leeds.ac.uk)
assay interference. Similar problems are also noted in
other steroid assays, such as testosterone in women.7
False positive interference in troponin assays in patients 1 Berson SA, Yalow R, Bauman MA, Newerly K. I 131 insulin metabolism in
with chest pain due to acute coronary syndrome has led human subjects: demonstration of insulin binding globulin in the circula-
tion of insulin-treated subjects. J Clin Invest 1956;35:170-90.
to prolonged inpatient stays and invasive investiga- 2 Kricka LJ. Interference in immunoassays—still a threat. Clin Chem
tion.8 9 False negative results are equally important in 2000;46:1037-8.
3 Selby C. Interference in immunoassays. Ann Clin Biochem 1999;36:
that they lead to underinvestigation.10 704-21.
The presence of interfering antibodies is surpris- 4 Ward G, McKinnon, Badrick T, Hickman PE. Heterophilic antibodies
remain a problem for the immunoassay laboratory. Am J Clin Pathol
ingly common, affecting 30-40% of the population.3 1997;108:417-21.
They probably arise from mundane activities such as 5 Cole LA, Rinne KM, Shahabi S, Omrani A. False-positive hCG assay
results leading to unnecessary surgery and chemotherapy and needless
keeping pets, ingesting animal antigens, vaccination, occurrences of diabetes and coma. Clin Chem 1999;45:313-4.
infection, or even blood transfusion. Most analytical 6 Rotmensch S, Cole LA. False diagnosis and needless therapy of
presumed malignant disease in women with false-positive human chori-
assays currently in use can neutralise and block low con- onic gonadotrophin concentrations. Lancet 2000;355:712-5.
centrations of these interfering proteins (ìg to mg/l) 7 Torjesen PA, Bjøro T. Antibodies against I 125 testosterone in patient’s
serum: A problem for the laboratory and the patient. Clin Chem
with no or minimum impact on analytical accuracy. 1996;43:2047-8.
Larger amounts of interfering proteins, which may be as 8 Cavinsky M, Laterza O, Pfeifer JD, Farkas ST, Scott MG. IgM antibody to
Escherichia coli produces false positive results in multiple immunometric
high as grams per litre, or proteins with high binding assays. Clin Chem 2000;46:1157-61.
affinity can, however, overwhelm the analytical system, 9 Krahn J, Parry DM, Laroux M, Dalton J. High percentage of false positive
leading to assay interference and erroneous results. The cardiac troponin I results in patients with rheumatoid factor. Clin Biochem
1999;32:477-80.
number of these extreme cases is not known, though 10 Bohner J, von Pape KW, Hannes W, Stegmann T. False-negative
specific types of interference, such as heterophilic and immunoassay results for cardiac troponin I probably due to circulating
troponin I autoantibodies. Clin Chem 1996;42:2046-7.
anti-murine antibodies, in the order of 0.05% have been 11 Norden AGW, Jackson RA, Norden LE, Griffin AJ, Barnes MA, Little JA.
reported.4 11 Our experience suggests that interfering Misleading results for immunoassays of serum free thyroxine in the pres-
ence of rheumatoid factor. Clin Chem 1997;43:957-62.
antibodies of various types affect about 0.5% immu- 12 Kuroki M, Matsumoto Y, Arakawa F, Haruno M, Murakami M, Kuwahara
noassays (A Ismail, J Barth, unpublished data), though M, et al. Reducing interference from heterophilic antibodies in a two-site
immunoassay for carcinoembryonic antigen (CEA) by using human-
others have reported higher percentages.12 Even the mouse chimeric antibody to CEA as the tracer. J Immunol Methods
lowest prevalence quoted should be seen in the context 1995;180:81-91.
13 Weber TH, Kapyaho KI, Tanner P. Endogenous interference in
of the total number of immunoassays—many millions a immunoassays in clinical chemistry. A review. Scan J Clin Lab Invest
year in British hospital laboratories alone. Thus many 1990;50(suppl 201):77-82.

Monitoring the safety of over the counter drugs


We need a better way than spontaneous reports

L
ast year the Food and Drug Administration in in response to a case-control study by the Yale
the United States recommended that phenyl- haemorrhagic stroke project investigators designed to
propanolamine be removed from non- determine the risk of haemorrhagic stroke in people
prescription and prescription medicines and that exposed to phenylpropanolamine.1 Though this action
pharmaceutical companies voluntarily discontinue was not followed in the United Kingdom, because the
products containing phenylpropanolamine. This was market conditions for phenylpropanolamine are BMJ 2001;323:706–7

706 BMJ VOLUME 323 29 SEPTEMBER 2001 bmj.com


Editorials

different, the findings of the haemorrhagic stroke cists. Although records of dispensed prescriptions are
project highlight the need for continued monitoring of used in everyday practice to draw attention to
the safety of non-prescription medicines. suspected drug interactions and adverse reactions, it is
In the United States formulations of phenylpropa- rare for pharmacists to record details of over the coun-
nolamine contain up to 150 mg of a racemic mixture of ter purchases. It may be time for that to change. In the
two enantiomers of (±)-norephedrine,2 and withdrawal interests of public health, pharmacists should be
of appetite suppressants and cough and cold prepara- encouraged to use available computer facilities to
tions containing phenylpropanolamine may well have record over the counter purchases when appropriate,
been an appropriate response to the haemorrhagic with demographic details and the reason for purchase.
stroke project data. In Europe the licensed maximum In this way, as well as providing additional data for
daily dose in non-prescription cough and cold remedies adverse drug reaction reporting, a further degree of
is lower (100 mg), and appetite suppressants containing control of the sale would be exercised to help ensure
phenylpropanolamine are not available in the United that the products are not used inappropriately.
Kingdom. The UK’s Committee on Safety of Medicines Population based pharmacy databases could be
reviewed the data but concluded that products contain- used for modified prescription event monitoring11
ing phenylpropanolamine did not need to be with- studies. Pharmacy prescription event monitoring
drawn.3 It considered that risk factors for haemorrhagic surveillance of selected pharmacy status drugs could
stroke were well known,4 and patient information leaflets be undertaken by sending questionnaires to pharma-
for products containing phenylpropanolamine clearly cists and asking them to fill in details from their
state that they should not be used by people with high database. Outcome data could be obtained from
blood pressure, hyperthyroidism, coronary heart dis- patients and their doctors. Telephone interviews for
ease, and other risk factors because of the drug’s follow up have been used successfully to facilitate
sympathomimetic effects.5 improved response.12 Though pharmacoepidemiologi-
Non-prescription medicines are often perceived as cal investigations such as cohort or case-control studies
safe by the consumer because they are freely available are feasible using cohorts of patients obtained from
for self selection. The haemorrhagic stroke project databases maintained by pharmacists, the practical
results show, however, that over the counter products difficulties need to be addressed in pilot studies.
may be far from innocuous. The current emphasis in The healthcare system must establish a practical yet
healthcare is to shift some responsibility on to the con- scientifically robust method of surveillance of adverse
sumer through expansion of self medication.6 Over the reactions associated with over the counter medicines.
counter products in the United Kingdom have either a The part of the pharmaceutical industry that deals with
pharmacy only authorisation, which requires supervi- over the counter products seems commercially driven,
sion of sale by a pharmacist in a pharmacy, or may be and the funding of such pharmacovigilance is unlikely
licensed for general sale in non-pharmacy retail to come voluntarily from industry unless an element of
outlets. In Britain the introduction of national statutory compulsion becomes a condition of the over
guidelines has encouraged applications for reclassifica- the counter licence. As the queue of drugs waiting to
tion of drugs from prescription only, which require the gain over the counter status grows, now is the time to
direct involvement of a medical practitioner, to implement such controls.
pharmacy status. With further such switches scheduled,
an increased number of powerful medications are David Clark senior lecturer
likely to become available over the counter. Evidence Department of Pharmacology, University of Otago, Dunedin, New
Zealand
from observational studies suggests that such products
are not always used in accordance with the manufac- Deborah Layton research pharmacist
turers’ recommendations,7 so the use of over the coun- Saad A W Shakir director
ter drugs needs to be monitored and quantified. Drug Safety Research Unit, Bursledon Hall, Southampton SO31 1AA
In New Zealand pharmacists are actively encouraged
to report adverse drug reactions involving prescribed or
over the counter medicines to the Center for Adverse 1 Kernan, WN, Viscoli CM, Brass LM, Broderick JP, Brott T, Feldmann E, et
Reactions to Medicines. These spontaneous reports are al. Phenylpropanolamine and risk of hemorrhagic stroke. N Engl J Med
2000;343:1826-32.
followed up by requests for further clinical details from 2 Moffat AC .Phenylpropanolamine. Putting the record straight [letter].
the patient’s general practitioner. In 1999 the yellow Pharmaceut J 2000;265:817.
3 Committee on Safety of Medicines. Phenylpropanolamine and heamor-
card scheme for reporting adverse effects in the United rhagic stroke. Current Problems in Pharmacovigilance 2001;27:5-6.
Kingdom was extended to community pharmacists. 4 Jick H, Aselton P, Hunter JR. Phenylpropanolamine and cerebral haem-
orrhage. Lancet 1984;i:1017.
Notifications from pharmacists will increase the number 5 Lake CR. Adverse drug effects attributed to phenylpropanolamine: a
of notifications of adverse drug reactions and may gen- review of 142 case reports. Am J Med 1990;89:195-208.
erate new safety signals. Collaboration between pharma- 6 Bradley C, Blenkinsopp A. Over the counter drugs. The future for
self-medication. BMJ 1996;312:835-7.
cists and physicians may also improve reporting,8 9 7 Sinclair HK, Bond CM, Hannaford PC. NSAIDs: improving patient care
especially as many physicians receive little formal teach- [letter]. Pharmaceut J 2000;265:718.
8 Meyboom RHB, Egberts ACG, Gribnau FWJ, Hekster YA. Pharmaco-
ing about over the counter medicines. However, the vigilance in perspective. Drug Safety 1999;6:429-47.
limitations of spontaneous reporting are well recog- 9 Thomas DHV, Noyce PR. Over the counter drugs: the interface between
self medication and the NHS. BMJ 1996;312:688-91.
nised. The reasons for doctors under-reporting adverse 10 Inman WHW. Monitoring for drug safety. Lancaster: MTP Press, 1980:36-7.
drug reactions include fear of litigation, diffidence, and 11 Mann RD, Wilton LV, Pearce GL, Mackay FJ, Dunn NR. Prescription
event monitoring (PEM) in 1996: a method of non-interventional obser-
complacency,10 and pharmacists are likely to hampered vational cohort pharmacovigilance. Pharmacoepidemiol Drug Safety 1997;6
by similar constraints. (suppl 3):S5-11.
12 Gauld NJ, Shaw JP, Emmerton LE, Pethica BD. Surveillance of a recently
Patient medication records are a requirement for switched non-prescription medicine using a pharmacy-based approach.
professional standards of good practice for pharma- Pharmacoepidemiol Drug Safety 2000;9:207-14.

BMJ VOLUME 323 29 SEPTEMBER 2001 bmj.com 707

You might also like