Background
NSAIDs are not considered disease modifying in osteoarthritis (OA), though there has been debate that some may be chondroprotective and others chondrodestructive. We performed a case control study comparing NSAID use in subjects with hip or knee OA who had been treated surgically vs. medically. The purpose was to determine the relationship between NSAIDs and OA severity and if NSAIDs had a chondroprotective or chondrodestructive effect in our cohort.
Material and Method
The study population was subjects who were referred to orthopedic surgeons or rheumatologists and had no inflammatory arthritis or secondary OA. A questionnaire was mailed asking about current/ever use of specific NSAIDs. Between groups analyses were conducted for surgically treated cases vs. medically treated controls with mild, moderate and severe OA.
Results
Of 618 subjects, 73 had radiographic mild, 133 moderate and 412 severe OA. Surgically treated cases were less likely to have taken any NSAID (OR 0.44, 95% CI [0.23, 0.81], p<0.01. There was an inverse dose response with use of > or = 3 NSAIDs being taken in those with mild vs severe OA (OR 3.7), mild vs moderate OA (OR 2.5), and moderate vs severe OA (OR 1.5), all with p<0.0002; the same was found with > or = 2 NSAIDs, p<0.0002.
Conclusions
NSAID use and number of NSAIDs taken was greater in mild radiographic OA. We cannot support that some NSAIDs have negative cartilage effects from this study, and in fact, most NSAIDs were used in patients with less severe radiographic OA of the hip or knee.

Keywords: Osteoarthritis - surgery, Severity of Illness Index, Knee Joint - physiopathology, Hip Joint - physiopathology, Drug Utilization Review, Case-Control Studies, Arthroplasty, Replacement, Knee, Arthroplasty, Replacement, Hip, Anti-Inflammatory Agents, Non-Steroidal - therapeutic use