To the Editor: Dr Lasser and colleagues1 found that 10.2% of all new drugs approved between
1975 and 1999 either acquired a black box warning or were withdrawn. Thus,
they conclude that postapproval adverse drug reactions (ADRs) are common.
I would like to address 2 additional issues in postmarketing surveillance.
First, premarketing clinical trials are consistently designed with a
narrow spectrum of clinical disorders and participating patient profiles to
achieve a required statistical result. Rarely do these patients reflect the
exact clinical profiles, including use of concomitant medications, of the
patient population to whom the drugs will be administered after marketing.
Second, serious ADRs are often rare, albeit extremely important. Therefore,
the number of patient exposures required to identify such events may be large,
sometimes numbering in the millions. If only a fraction of such events is
reported to either the manufacturer or the Food and Drug Administration (FDA),
that number becomes enormous.