Do drug-related safety warnings have the expected impact on drug therapy? A systematic review

摘要

Purpose: The need for drug-related safety warnings is undisputed, and their impact should also be evaluated. This systematic review investigates and assesses the impact of safety warnings on drug therapy. Methods: Studies published in English between January 1998 and December 2018 were searched in EMBASE and MEDLINE, complemented by manual search. Randomised controlled trials, cohort studies with a before/after component, and case-control studies were included, selected to predefined criteria, and assessed for their reporting and methodological quality. Results: Out of 7454 references identified, 72 studies were included. A total of 28/72 (39%) studies described the impact of safety warnings on drug therapy as being effective, whereas 12/72 (17%) studies did not. Further, 26/72 (36%) studies described a partial implementation of the warnings (one part of the warning had an impact on drug therapy and another did not). Unintended effects were investigated in 6/72 (8%) studies. While 34 (47%) studies examined safety warnings on psychotropic drugs using an interrupted time series (ITS) design (53%), a before/after (26%), and a time series design (21%), 38 (53%) studied other substances using an ITS design (34%), a before/ after (40%), and a time series design (26%). The proportion of an effective impact on drug therapy was lower in the "psychotropic drugs" group (23%) than in the "others" group (53%). Conclusion: Drug-related safety warnings induce intended and unintended effects. The included studies are of broadly varying methodological quality. To better compare their effectiveness, studies should be conducted using standardised procedures.