Arthritis - rheumatoid and osteoarthritis - is one of the most prevalent chronic conditions affecting nearly 50% of persons over the age of 65. OBJECTIVES: The objective of this study was to compare cardiorenal events and costs between two COX-2 inhibitors - celecoxib and rofecoxib. METHODS: This was a retrospective analysis of medical and pharmacy claims. All patients newly started on a COX-2 inhibitor during 7/1/99 to 6/30/00 were identified and followed for 6 months before and after the initial COX-2 prescription. Incident cardiorenal events were attributable to a COX-2 if there was an ICD-9 diagnosis code within 45 days after the last days supply of the prescription. RESULTS: A total of 20,514 patients were newly prescribed celecoxib (n=12,487) or rofecoxib (n=8,027). Mean age was 65 (+/- 15) and 68% were female. Primary indication for COX-2 was pain (55%), followed by osteoarthritis (23%), other arthritis (17%), and rheumatoid arthritis (6%). There were no significant differences in baseline cardiorenal history and medication use (antiarthritic, antihypertensive, and GI-related) between the two cohorts. Among the baseline hypertensive patients, those on rofecoxib were 34% more likely to experience an incident cardiorenal event than patients on celecoxib, adjusting for age, gender, comorbidity, indication and dosage (OR=1.34;p=0.007). Results were similar for the baseline non-hypertensive patients, with those on rofecoxib reporting a higher risk of new cardiorenal events (OR=1.18; p=0.0009). Although not statistically significant, patients on rofecoxib incurred slightly higher total health care costs than those on celecoxib ($8,188 vs. $7,540; p=0.0867). CONCLUSION: The risk of cardiorenal events was significantly higher in rofecoxib-treated patients than celecoxib-treated patients. There were no statistically significant differences in total costs between the two COX-2 inhibitors, although celecoxib-treated patients incurred slightly lower total health care costs.