Anecdotes as evidence
We need guidelines for reporting anecdotes of suspected adverse drug reactions
he short story” wrote William Trevor, a mas-
Reasons for publishing anecdotes
ter of the genre, “should be an explosion of
the truth.” Being interested in the clinical Reason Examples (see also bmj.com)
short stories that are anecdotal reports of adverse drug To describe a newly recognised adverse reaction Oculomucocutaneous syndrome and practolol5
reactions and interactions, I am concerned that they
To generate hypotheses Teratogenicity of antihistamines6 7
should explode the whole truth. To test hypotheses The loading dose of digoxin in severe renal insufficiency8
There are many such anecdotes. Of the 3252 To demonstrate diagnostic techniques Serum KL-6 in diagnosing amiodarone-induced lung
citations in the 24th volume of the Side Effects of Drugs damage9
Annual,1 in which the world literature on adverse drug To elucidate mechanisms Drug-induced torsade de pointes and QT interval
reactions and interactions for 2000 was critically
To elucidate or suggest methods of management Treatment of verapamil self poisoning11 12
reviewed, about a third (1075 citations) were anec-
To remind or educate The adverse effects of liquorice13
dotes; in contrast, there were only 45 systematic
To enable systematic review Venous thromboembolism and mestranol5
Now the hierarchy of clinical evidence emphasises
large randomised controlled trials and systematic The CONSORT group has developed standardised
reviews; in this scheme anecdotal reports are ill guidelines on reporting randomised controlled trials.15
favoured.3 Nevertheless, anecdotal reports of adverse However, no similar guidelines exist for anecdotal
reactions should be published, for they have different reports. I have therefore outlined a proposed set of
functions to randomised controlled trials (table), a fact guidelines (see bmj.com), which would encourage
that is not emphasised by the evidence hierarchy. clinicians to report cases uniformly (facilitating direct
A short story sometimes illuminates an aspect of comparisons of individual reports) and obtain the
life that a roman à thèse cannot. Likewise, trials typically information needed to assess whether an adverse event
show the sizes of benefits or adverse effects, but is actually an adverse drug reaction and to establish
anecdotes call attention to potential adverse reactions diagnostic techniques, mechanisms, and guidelines for
or interactions, mechanisms, diagnostic techniques, or management. Uniform presentation would also facili-
methods of management; anecdotes can generate or tate systematic review of suspected reactions.
test hypotheses, and remind or educate; and, like trials, “The short story tells only one thing,” wrote VS
they can be subjected to systematic review, but of a dif- Pritchett, another master of the genre, “and that
ferent kind.4 intensely.” The clinical anecdote should do likewise.
Pharmacovigilance demands at least 19 essential Jeffrey K Aronson consultant clinical pharmacologist
pieces of information about the patient and the
Department of Clinical Pharmacology, Radcliffe Infirmary, Oxford
adverse reaction, listed on bmj.com. They include 14 OX2 6HE (firstname.lastname@example.org)
items that are mentioned or implied on the Committee
on Safety of Medicines’ yellow card for reporting Competing interests: None declared.
suspected adverse drug reactions in the United
1 Aronson JK, ed. Side effects of drugs. Annual 24. Amsterdam: Elsevier, 2001.
Kingdom, and 14 that are mentioned or implied on the 2 Aronson JK, Derry S, Loke YK. Adverse drug reactions: keeping up to
MedWatch adverse event forms in the United States. date. Fundam Clin Pharmacol 2002;16:49-56.
3 Concato J, Shah N, Horwitz RI. Randomized, controlled trials,
Several of these are important in computing the likeli- observational studies, and the hierarchy of research designs. New Engl J
hood that the adverse event was an adverse reaction Med 2000;342:1887-92.
4 Jenicek M. Clinical case reporting in evidence-based medicine. 2nd ed.
(that is, an event that is caused by the drug) and in elu- London: Arnold, 1999.
cidating the mechanism.14 5 Inman WHW. Don’t tell the patient. Behind the drug safety net. Bishops
Waltham: Highland Park Productions, 1999.
However, anecdotal reports of suspected adverse 6 Orme ML’E. The Debendox saga. BMJ (Clin Res Ed) 1985;291:918-9.
drug reactions and interactions often do not contain all 7 Mazzotta P, Magee LA. A risk-benefit assessment of pharmacological and
nonpharmacological treatments for nausea and vomiting of pregnancy.
the information that they should. For example, I have Drugs 2000;59:781-800.
reviewed 35 reports about 48 patients that appeared in 8 Aronson JK, Grahame-Smith DG. Altered distribution of digoxin in renal
the BMJ from January 2000 to October 2002 under failure—a cause of digoxin toxicity? Br J Clin Pharmacol 1976;3:1045-51.
9 Endoh Y, Hanai R, Uto K, Uno M, Nagashima H, Takizawa T, et al. Diag-
“Drug points.” The median numbers of items nostic usefulness of KL-6 measurements in patients with pulmonary
mentioned in those reports were 9 (range 5-12) of all complications after administration of amiodarone. J Cardiol
the 19 essential items mentioned above, 9 (4-12) of the 10 Dessertenne F. La tachycardie ventriculaire à deux foyers opposés
yellow card items, and 8 (3-11) of the 14 MedWatch variables. Arch Mal Coeur Vaiss 1966;59:263-72.
11 Buckley CD, Aronson JK. Prolonged half-life of verapamil in a case of
items. And this analysis omits other desirable features overdose: implications for therapy. Br J Clin Pharmacol 1995;39:680-3.
of anecdotal reports, such as formal assessment of the 12 Gokel Y, Paydas S, Duru M. High-dose verapamil–trandolapril induced
rhabdomyolysis and acute renal failure. Am J Emerg Med 2000;18:738-9.
likelihood that the event was an adverse drug reaction, 13 De Klerk GJ, Nieuwenhuis MG, Beutler JJ. Hypokalaemia and Extra tables and a
possible mechanisms, and review of previous cases. hypertension associated with use of liquorice flavoured chewing gum. suggested protocol
Of course, it can be difficult to obtain high quality 14 Stephens MDB. The diagnosis of adverse medical events associated with
data when an event occurs, but often a more assiduous drug treatment. Adv Drug React Ac Pois Rev 1987;1:1-35. anecdotal reports
15 Altman DG, Schulz KF, Moher D, Egger M, Davidoff F, Elbourne D, et al. appear on bmj.com
approach would help. We need formal guidance on The revised CONSORT statement for reporting randomized trials:
what is required. explanation and elaboration. Ann Intern Med 2001;134:663-94. BMJ 2003;326:1346
1346 BMJ VOLUME 326 21 JUNE 2003 bmj.com