changes over time thereafter. In order to ensure
continued patient protection, it is therefore neces-
sary to monitor the safety profile of marketed drugs
continuously for new signals of concern that might
prompt revisions in prescribing information.
Sample sizes
Clinical trials designed to prove the safety and
efficacy of drugs are limited by sample size and
strict enrollment criteria. As such, ADRs occur-
ring at fairly low rates (e.g. 1 in 1000) or those
occurring in patient subpopulations not studied
during clinical investigations may not be identi-
fied during clinical trials and can only be identi-
fied post-marketing. New rare, serious events may
be reported only after large numbers of patients
take a new drug, often after several years of mar-
keting experience (Kessler, 1993). One rule of
thumb is that for a clinical development program
containing a known number of patients exposed at
appropriate doses and for appropriate periods of
time, there is 95% confidence level that at least
one specified type of adverse event will have been
observed if it has a frequency greater than three
times the reciprocal of the sample size. Thus, a
clinical development program with 3000 appro-
priately treated patients (perhaps larger than aver-
age) would be very likely to include patients with
adverse events occurring at a frequency of 1 in
1000 or greater.
Adverse events are sometimes termed type A
(usually pharmacologically predictable, rela-
tively frequent, seldom fataland usually identified
during clinical trials) or type B (unpredictable
idiosyncratic reactions which are usually infre-
quent but can be very serious or fatal) (Rawlins
and Thompson, 1977; Venning, 1983). Post-
marketing ADR monitoring usually identifies
the more serious, type B reactions. The sample
size needed in clinical trials to detect differences
between an incidence rate of 1/10 000 and
2/10 000 is about 306 000 patients (e.g. for a
placebo comparison of chloramphenicol-induced
aplastic anemia, which occurs in 1/30 000;
Lasagna, 1983). Clinical trials at this scale are
simply impractical.
Spontaneous or unsolicited ADRs reported post-
marketing may contain limited, unclear or imper-
fect information. It is the responsibility of the
manufacturer to try to obtain as much relevant
information as possible so they can be clinically
assessed, particularly those that are serious.Drug interactions
Potentially harmful drug interactions may not be
identified during controlled clinical trials, due to
the exclusion of patients taking concomitant medi-
cations, which are not allowed to be taken during
a study. For example, terfenadine, a novel non-
sedating antihistamine which was found to cause a
serious and potentially fatal cardiac arrhythmia,
torsades de pointes, when administered with keto-
conazole or erythromycin, and this could not realis-
tically have been expected to be identified in the
clinical trial setting. The mechanism of this adverse
drug interaction was found to be due to cumulation
of unmetabolized terfenadine, due to inhibition
of cytochrome P-450 (CYP) by ketoconazole or
erythromycin; the parent terfenadine molecule is
usually cleared very rapidly when there is no
concomitant CYP inhibitor.39.2 Council for International
Organizations of Medical
Sciences (CIOMS) initiativesRecognizing that drug surveillance was a global
problem, and that international standardization
would assist the assessment of large numbers of
patients, the CIOMS of the World Health Organi-
zation (WHO) began meeting in 1986 (CIOMS
Working Group I, 1990; Geraldet al., 1990). The
original CIOMS I ‘working party’ consisted of
representatives from six regulatory authorities
and seven multinational pharmaceutical manufac-
turers. This group had the goal of developing a
uniform adverse event reporting form (the
CIOMS I form) that would be acceptable interna-
tionally. A system of expedited reporting of serious
adverae events (SAEs) to regulatory authorities536 CH39 DRUG SURVEILLANCE