| To Register SMDM Homepage | |
|
|
|
||||
Purpose: To compare side effects and perceived medication effectiveness among a cohort of osteoarthritis patients using non-steroidal anti-inflammatory drugs (NSAIDs) and cyclo-oxygenase Type II inhibitors (COX-2 inhibitors).
Methods: 4,386 self-reported osteoarthritis patients completed an internet survey providing comprehensive information on their disease. In addition to other characteristics, they reported medication use, perceived medication effectiveness, and frequency of six conditions possibly associated with osteoarthritis treatment--diarrhea, nausea/vomiting, heartburn, stomach pain, headache, and dizziness.
From this cohort we identified patients who had been regularly taking osteoarthritis medication (at least 3 times/week) for at least 7 months. Patients were classified as NSAID or COX-2 inhibitor users if they were using NSAID or COX-2 inhibitors without other concurrent osteoarthritis medications.
Potential confounding for treatment selection was adjusted for by developing a propensity score model of COX-2 inhibitor use (vs. NSAID use). The model included patient demographics, disease severity indicators, and clinically-plausible interactions and was estimated by logistic regression. We used the ‘greedy match’ approach to match patients sequentially by 5 to 1 decimal places. Using this new group of propensity-matched COX-2 inhibitor and NSAID users, we compared side effect burden and perceived medication effectiveness.
Results: Prior to matching COX-2 inhibitor users (n=410) were older, were more likely to be unemployed, experienced more pain, and missed more hours of work/activity due to their osteoarthritis compared to NSAID users (n=1409; all p<0.02). They had greater difficulty with most daily activities (p<0.05), but reported fewer side effects (p<0.01) and perceived their medication to be more effective than did NSAID users (p=0.01).
The propensity matching model yielded 396 matched pairs of COX-2 inhibitor and NSAID users which revealed no significant differences in demographics or disease severity variables. Propensity matched COX-2 inhibitor users experienced fewer side effects than NSAID users (p=0.04), particularly headaches (p<0.0001). This is significant because headaches were a strong predictor of lowered productivity in a separate regression (p<0.0001). COX-2 inhibitor users perceived their medications to be significantly more effective than did NSAID users (p=0.03).
Conclusion: Among a propensity score matched population of osteoarthritis patients, COX-2 inhibitor users report fewer side effects and perceive their medications to be more effective than NSAID users.
See more of Poster Session - Clinical Strategies; Judgment and Decison Making
See more of The 26th Annual Meeting of the Society for Medical Decision Making (October 17-20, 2004)