Prescriptions for Cyclooxygenase‐2 Inhibitors and Other Nonsteroidal Anti‐inflammatory Agents in a Medicaid Managed Care Population: African Americans Versus Caucasians

Objective.  To determine whether race is a predictor of a patient's likelihood of being prescribed selective cyclooxygenase‐2 inhibitors (COX‐2s) versus other nonsteroidal anti‐inflammatory agents (NSAIDs) in Medicaid managed care plans (MCO). Design.  All medical and prescription claims for Medicaid MCO enrollees receiving at least one prescription for a COX‐2 or NSAID between January 2000 and June 2002 were retrieved. Selected for study were adults claiming at least one COX‐2 prescription or NSAID prescription with a minimum 30 days of supply after June 2000; having 60 total days of supply or more over the study period was also required for study inclusion. The probability of being prescribed a COX‐2 was estimated as a logistic function of patient age, gender, race, city/suburban/rural residence, and history of rheumatoid arthritis, osteoarthritis, chronic back pain, acute pains, gastrointestinal problems, use of anticoagulants or corticosteroids, and comorbidities. Results.  Of the 16,868 enrollees meeting the selection criteria, 4,005 (24%) were prescribed a COX‐2 and 12,863 another NSAID. Half of those studied were African American, three‐quarters were female, and a third were 50–64 years old. After adjusting for confounders, odds of a COX‐2 prescription were a third less for African Americans and other races compared to Caucasians (OR, 0.67; 95% confidence intervals, 0.62–0.73). Conclusion.  Patient race is a significant predictor of COX‐2 prescriptions in the Medicaid population, even after adjusting for other demographic and clinical variables. Cost to the patient was not a factor, as the patient copayment was $1 for any prescription.