Background In randomised clinical trials, patients with osteoarthritis (OA) who received rofecoxib developed fewer clinically significant gastrointestinal (GI) adverse events than those received conventional NSAIDs.

Objectives We estimated the economic implications of these results in Switzerland.

Methods A model-based cost-effectiveness analysis was conducted, focusing on events related to serious GI complications (perforations, ulcers or bleeds or PUBs) and minor GI problems (e.g. dyspepsia) that imply healthcare resource use. The base case analysis had a one-year horizon. GI event probabilities were obtained from a pre-specified pooled analysis of the clinical trials of rofecoxib. The NSAID comparator was an aggregate of those commonly used in Switzerland. To extrapolate trial results to clinical practice, resource utilisation profiles for inpatient and outpatient care of GI events were developed using both the literature and expert opinions in UK, the United States and Canada. Weighted average drug costs of currently reimbursed medications were derived from IMS Health data. It was assumed that 20.6% of conventional NSAID users needed gastroprotective agents (GPAs) prophylactically based on published data in the UK and other countries. Efficacy of concomitant use of GPAs in preventing NSAID-associated PUBs was inferred from the literature. A 75% reduction in prophylactic GPA prescription rates with VIOXX was assumed for the base case. Sensitivity analyses explored the effects of varying drug costs, efficacy of GPAs and rates of concomitant GPA use.

Results Our base case analysis showed that in Switzerland, it would save 0.64 Swiss Francs per patient per day by switching conventional NSAIDs to rofecoxib for the treatment of OA. The iatrogenic cost factor for NSAIDs was 1.75. Results were robust over a range of model assumptions. The cost per PUB avoided ranged from an added cost of 5,951 Swiss Francs (assuming 0% of GPA use with NSAIDs) to a saving of 16,013 (assuming 50%), with a saving of 2,522 Swiss Francs at base case (20.6% GPA use with NSAIDs).

Conclusion Based upon this analysis of serious and minor GI events for OA patients in Switzerland, rofecoxib represents a cost-effective therapeutic option compared to NSAIDs due to cost savings in GI problems and reduction in use of GPAs.