Temporal patterns of NSAID spontaneous adverse event reports: the Weber effect revisited.

OBJECTIVES: Determine whether recent US adverse event reports for several non-steroidal anti-inflammatory drugs (NSAIDs) conform to the temporal pattern observed by Dr JCP Weber in the UK in the early 1980s, i.e. a rising count in the first few years after launch presumably reflecting increased exposure, followed by a decline, presumably reflecting decreased enthusiasm for reporting as adverse events become well known. STUDY
SETTING: US adverse event report data available from the US Food and Drug Administration, reformatted by a commercial vendor.
METHODS: For the 5 NSAIDs launched in the US between 1987 and 1993 that had data suitable for this study, we tabulated by year from launch the number of reports and the reporting rate (number of reports per 1000 prescriptions).
RESULTS: The number of reports for 3 of the 5 NSAIDs showed a temporal pattern approximating that described by Weber. The number of reports for the other 2 NSAIDs showed temporal patterns markedly different from that described by Weber. For 4 of the 5 NSAIDs, reporting rates did not decline consistently with time from launch. DISCUSSION: The temporal patterns of adverse event reports are more complex than those described in Weber's classic report. The number of reports does not reliably rise and then fall after launch and the reporting rate does not reliably decrease with time from launch.