NSAID use profiles derived from reimbursement data in France.

Because extrapolation of adverse event rates from clinical trials to the general population relies on the assumption of similar drug-usage patterns, we have studied usage patterns for nonsteroidal anti-inflammatory drugs (NSAIDs) from the National Health Insurance claims database. Reimbursement patterns of NSAIDs and analgesics were studied in a cohort of 232 users, half selected from authorisations of reimbursed spa cures, and the other half - strictly matched according to age and gender - who had redeemed at study onset at least one prescription for NSAIDs, analgesics or other antirheumatic drugs. Drug utilisation was measured as the total number of defined daily doses (DDD) bought over the 9-month study period. Over this time, the patients bought a median 45 DDD of NSAIDs out of the 270 needed to cover the whole period. Only two patients bought 270 DDD or more. Median paracetamol use (alone or combined) was 26 days. This study shows that patients use NSAIDs either intermittently or at lower than recommended doses, which is consistent with a primarily symptomatic use. Intermittent or on-demand use of NSAIDs, or at doses lower than full, could partly explain the lower rates of gastrointestinal bleeding and other ulcer complications in postmarketing epidemiological studies than could be expected from the findings of clinical trials.