Background: Cyclo-oxygenase (COX)-2 inhibitors were introduced to world markets with claims of improved gastrointestinal safety compared with traditional NSAIDs. Randomized clinical trials had demonstrated fewer adverse gastrointestinal events with COX-2 inhibitors, but no difference with other adverse events, including adverse renal events. There was a rapid uptake of these medicines.
Objective: To compare uptake rates of NSAIDs, including COX-2 inhibitors, in a reference population with those in two high-risk populations: a population taking medicines affecting the renin-angiotensin system and loop diuretics, and a population taking medicines for diabetes mellitus.
Method: An observational study was undertaken in which the Department of Veterans' Affairs claims dataset was used to identify: veterans dispensed ACE inhibitors (ACEIs) or angiotensin II type 1 receptor antagonists (angiotensin receptor blockers [ARBs]) and furosemide (ACEI-ARB/furosemide cohort); veterans dispensed medicines for diabetes (diabetes cohort); and all other veterans (reference cohort) from July 1999 to July 2007. Concurrent dispensing of NSAIDs was assessed.
Results: Prior to celecoxib becoming subsidized in Australia, the baseline level of NSAID use was 19.5% in the reference cohort, 15.3% in the diabetes cohort and 15.6% in the ACEI-ARB/furosemide cohort. After the listing of celecoxib, utilization of NSAIDs increased by 42.2% in the reference cohort, with similar increases in the diabetes cohort (40.8%; p = 0.88 compared with the reference cohort) and the ACEI-ARB/furosemide cohort (49.6%; p = 0.09 compared with the reference cohort). With the withdrawal of rofecoxib, utilization of NSAIDs in the reference cohort fell by 25.3%, with similar falls in the diabetes cohort (24%; p = 0.28 compared with the reference cohort) and the ACEI-ARB/furosemide cohort (26.1%; p = 0.43 compared with the reference cohort).
Conclusions: Despite the increased vulnerability of veterans receiving ACEI-ARB/furosemide or diabetes medicines to adverse events of NSAIDs, uptake rates of COX inhibitors were equivalent to the rest of the veteran population. This suggests the gastrointestinal safety messages were interpreted broadly by prescribers and the adverse renal effects were not considered.