Gastrointestinal prophylactic therapy among patients with arthritis treated by rheumatology specialists

Objective. To determine rates of gastroprotective agent (GPA) use among patients with arthritis treated by rheumatologists, and to determine factors associated with GPA prescription. Methods. In a longitudinal outcome study, 11,451 patients with rheumatoid arthritis (RA) and osteoarthritis (OA) reported all medication use, ulcer history, functional status, and sociodemographic characteristics. Results. GPA were used in 21-24% of all patients with RA and OA and in about 35-40% of all high risk patients. In unadjusted analyses, GPA use was similar among NSAID users and non-users. In multivariable logistic regression analyses GPA use was associated with non-specific (NS) NSAID and COX-2 NSAID, prednisone, low dose aspirin, comorbidity, Health Assessment Questionnaire functional score, age < 65 years, increased income, not smoking, and being male. Despite numerous associations, the explanatory power for GPA use was poor (area under ROC curve = 0.680). Conclusion. GPA are used in 35% to 40% of patients with 4 risk factors for gastrointestinal ulceration. GPA use is not increased in NS NSAID users compared to COX-2 NSAID users, and was inversely associated with socioeconomic status. GPA use does not follow the model predicted by clinical trial results with respect to NS NSAID and age, reflecting a change in the pattern of NSAID use in patients with rheumatic disease. The major determinant of GPA use appears to be physician prescribing behavior.