Abstract
The early detection and improved understanding of adverse reactions caused by therapeutic drugs is vital to the safety of medical practice. In his recent valuable survey of the problems of pharmacoepidemiology, and after comment on the importance of metanalyses of clinical trials, Rawlins (1) stated that comparable: “overviews of safety problems … have yet to be secured by a robust statistical methodology.” As in other disciplines, many established biometric techniques are available for use, but these usually rest upon conditions relating to the source and ascertainment of data that are not fulfilled by informally acquired and spontaneously submitted records. Has not the time arrived for renewed efforts to test alternative methods of data acquisition, and to incorporate the best of these into national measures for drug safety?
Get full access to this article
View all access options for this article.
