in a particular patient. Occasionally, a case report,
even from a patient, will describe fully his/her
adverse event, including positive rechallenge, and
this is very important information in relation to the
safety profile of the drug.
Spontaneous case reports can reassure a com-
pany if a report describes a large accidental over-
dose with no serious adverse effects. They can also
provide reassurance, when reviewed in aggregate,
when no reports for drugxcausing eventyover
time periodzhave been received. Clusters of simi-
lar spontaneous reports should be meaningfully
analyzed for consistency in time to onset post-
dose, pattern of presentation, rechallenge and
dechallenge, to identify a signal and to get a feel
for its significance.
The main advantage of spontaneous case reports
is that they can provide important signals when
reviewed collectively. Although it would be
wrong to underestimate their occasional individual
importance, it is the consistency of time to onset
and pattern of presentation that is important. The
spontaneous case report database cannot be used to
give an accurate incidence rate of even the Type B
adverse reactions, because not all cases are
reported (Fletcher, 1991; Kessler, 1993). Nor do
spontaneous case reports lend themselves to mean-
ingful comparisons between different drugs. Not
only are all cases not reported for either drug but
also the reporting pattern varies with the time from
launch (the reporting rate generally peaks one to
two years after marketing) (Weber, 1984; Sachs
and Bortnichak, 1986), and also the reporting rate
for a particular adverse reaction tends to increase
after publication of a signal.
Pharmaceutical companies, individual regula-
tory authorities and the WHO have databases
which facilitate this overview. The use of a
standard coding dictionary of adverse event
terms is essential for this sort of analysis, and
one, MedDRA (Medical Dictionary for Regu-
latory Activities) has been accepted as the ‘gold
standard’ to be used. Nevertheless, routine
review of individual cases by responsible, experi-
enced reviewers is the most essential factor in
identifying new signals and ensuring patient
protection.
39.5 Causality assessment
It is often difficult to assess causality or treatment
association. For individual patients, factors such as
polypharmacy and multiple events occurring dur-
ing therapy can interfere with the causality assess-
ment of ADRs. In one study, three clinical
pharmacologists independently evaluated 500
untoward clinical events. There were broad differ-
ences in interpretation in causality of adverse
events (Koch-Weseret al., 1977).
The factors influencing causality assessments
are as follows:What is the background incidence of the event
independent of any treatment?Is there evidence that the incidence in users of the
drug is greater than the background incidence?What is the chronology of the occurrence of the
reaction?Is the chronology consistent between reports?Is the reaction biologically plausible, based on
what is known about the pharmacodynamics and
pharmacokinetics of the drug?Is there evidence of a drug–drug interaction?Is there an alternative or more plausible explana-
tion (e.g. natural history of disease, concurrent
conditions, other therapies, other exposures)?Isthereaction knownto occur withother drugsin
the same class or with similar structure?Is the reaction commonly associated with drugs
in general?Is there any supporting evidence from clinical
trials, post-marketing surveillance studies or
animal studies?Are there any cases which reoccurred on rechal-
lenge?540 CH39 DRUG SURVEILLANCE