Difficulties in assessing adverse drug reactions in clinical trials

1. The discovery of an adverse drug reaction (ADR) depends on: the relative frequencies of the drug-related and non-drug events; the mechanism of drug-induced toxicity; the number of patients exposed to the drug; and the methods used for detecting toxicity. Clinical trials are usually short-term studies conducted in a few hundred patients before marketing a drug. Therefore only the most common acute, dose-related ADRs are usually detected in the pre-marketing phase. 2. The detection and assessment of ADRs in clinical trials is still inadequate. Unstructured (open-ended questionnaires) and structured procedures (checklists) are the methods most commonly used. Since both methods have limitations, the discovery of ADRs in pre-marketing trials (phases I to III) should also rely on the complementary information provided by other methods: physiological and physical examinations, and pertinent laboratory tests. A more definite assessment of individual cases of ADRs should include the use of the adverse drug reaction probability scale (APS) or similar procedures. 3. Because the knowledge on the clinical toxicity of a drug will always be incomplete at the time of marketing, further investigation of the frequency and determinants of ADRs must be pursued in the post-marketing phase.