Effect of consumer reporting on signal detection: using disproportionality analysis.

Research paper by Isaac W IW Hammond, Donna S DS Rich, Trevor G TG Gibbs

Indexed on: 31 Oct '07Published on: 31 Oct '07Published in: Expert opinion on drug safety


Pharmacovigilance objectives and activities are designed to protect the health of consumers and are generally based on data acquisition from spontaneous adverse event reports (SADRs). SADRs come from different sources, including healthcare professionals, consumers, lawyers, other pharmaceutical companies, regulatory agencies and so on. Pharmacovigilance activities derived from SADRs include signal detection and description of the safety profile of the drug. Consumers are the most frequent source of most SADRs, even though the system was originally designed to receive reports from healthcare professionals. Most spontaneous adverse event reports are received from the US. GlaxoSmithKline (GSK) conducts monthly signal detection on all marketed compounds in its global database using disproportionality analysis, the empirical Bayesian algorithm known as a multiple-item gamma-Poisson shrinker. There are no systematic survey data or reviews of actual experiences within existing safety surveillance databases of how pharmaceutical companies handle consumer reports. Thus, a study was undertaken to determine the impact of consumer reports on signal detection using MGPS disproportionality analysis. Two data sets were created for four randomly selected GSK marketed compounds; one data set included reports from both consumer and healthcare providers and the second included only reports from healthcare providers. Disproportionality analysis was then used to evaluate the two data sets. A total of 23 signals were identified with a mean difference in time to signal detection of 1.8 years. The difference was in the range of -8-10 years. In 52.2% of events (12/23), the signal was identified earlier when consumer reports were included in the data. In 34.8% of events (8/23), the signal was identified in the same year in both data sets and, in 13% of the events (3/23), the signal was identified later when consumer reports were included in the data. It was concluded from this study that adverse event reports submitted directly to pharmaceutical companies by consumers can help significantly in the early detection of safety signals.