Abstract

Predispositions to prescribe a pure analgesic, a cyclooxygenase inhibiting nonsteroidal antiinflammatory drug (NSAID) or a nonacetylated salicylate for a fictitious, uncomplicated case of osteoarthritis (OA) were surveyed in a sample of 152 community based primary care practitioners who had been characterized with respect to medical education and practice characteristics. Only 2 respondents (1%) prescribed a pure analgesic; 35% prescribed subantiinflammatory doses of cyclooxygenase inhibiting NSAID. The remainder (64%) recommended nonacetylated salicylates and cyclooxygenase inhibiting NSAID in doses large enough to achieve an antiinflammatory effect. Past participants in postgraduate rheumatology electives prescribed more costly regimens than those who had not participated (p = 0.05). When the case was altered to include a history of previous peptic ulcer, 44% chose cyclooxygenase inhibiting NSAID with adjunctive prophylaxis against NSAID induced gastropathy (e.g., misoprostol). When the complication was changed to renal insufficiency, recommendations for sulindac increased 3-fold over those for the uncomplicated case (34 vs 11%), and were most common among more recent medical school graduates and past participants in rheumatology electives (p less than 0.05 for both). The potential effects of educational and practice variables on the therapeutic strategies and costs of OA care in the community merit further systematic study.