Observational evidence for determining drug safetyBMJ 2008; 336 doi: https://doi.org/10.1136/bmj.39491.493252.80 (Published 20 March 2008) Cite this as: BMJ 2008;336:627
- Nick Freemantle, professor of clinical epidemiology and biostatistics1,
- Alar Irs, lecturer in clinical pharmacology2
- 1Health Care Evaluation Group, University of Birmingham, Edgbaston, Birmingham B15 2TT
- 2Institute of Pharmacology, University of Tartu, 50411 Tartu, Estonia
Drug regulators are constantly criticised for their handling of drug safety. Comment on the activities of the US Food and Drug Administration with regard to drug safety has become particularly loud and damning. A recent article that proposed sweeping changes to the system criticised the FDA for lacking relevant expertise, being too close to the drug industry, and failing to ensure that companies deliver on their safety commitments with new drugs.1 The main practical recommendation was that systems of monitoring and reporting adverse events should be enhanced, highlighting a perceived role for observational studies. Drug safety is also an important concern in Europe, and the European Commission Enterprise and Industry Directorate General is consulting with stakeholders on changes of legislation to strengthen and rationalise pharmacovigilance in Europe.2 It describes a strategy that also focuses on observational studies of drug safety. But can observational studies alone improve judgments on drug safety?
Using adverse event reporting and observational studies to examine safety might seem attractive, but for several reasons observational studies alone cannot provide reliable estimates of treatment effects. For example, safety outcomes can be confounded with patients’ prognosis, drugs …