Answers are grounded in this paper and cite their sources.
A study of the ‘Current Problems’ publication of the Committee on Safety of Medicines in the UK has shown that many drug safety alerts or warnings are based upon a fairly small number of spontaneous adverse drug reaction reports. In terms of problems naming one drug and one serious reaction, or type of reaction, 24 (52%) of the 46 problems discussed in ‘Current Problems’ related to 10 or less yellow card reports. Nine (43%) of the 21 problems naming one drug and one non‐serious reaction, or type of reaction, were concerned with 10 or less reports. ‘Current Problems’ contain many entries other than this rather clear‐cut type of situation but, in respect of all other quantified entries in the publication that related to serious reactions, 13 (41%) of the 32 relevant problems were concerned with 10 or less yellow card reports. The situation seems to be different when non‐serious reactions are considered — as in general review articles or entries considering the whole range of reactions being reported in respect of a specific drug — when 9 (65%) of the 14 relevant ‘Current Problems’ entries were concerned with 51 or more yellow card reports. Of the problems that named one drug and one serious reaction, or type of reaction, 12 of the 24 relevant ‘Current Problems’ entries were concerned with 4 or less yellow card reports. The relevant alerts and warnings seem, in view of subsequent experience, to have been justified. Nevertheless, there would seem to be much merit in investigating health hazard signals on computerized databases before publication takes place that may lead to differential reporting and subsequently bias the relevant data.
No takes yet. Share an insight, caveat, or question.
Ronald D. Mann (1992) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: