Comparative effectiveness of oral pharmacologic interventions for knee osteoarthritis: A network meta-analysis

Objectives: To explore the relative efficacy of oral pharmacologic interventions in the treatment of knee OA. Methods: A systematic literature review was conducted using the MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trials databases to identify trials conducted in patients with knee OA with a minimum 6 weeks of follow-up. The standardized mean differences of the change from baseline to week 6 in Western Ontario and McMaster Universities Arthritis Index (WOMAC) pain between the treatment groups were estimated using Bayesian random-effects network meta-analyses. Subgroup analyses of baseline pain status (high, pain score ≥60 mm; low, pain score <60 mm) were performed. Results: Of 4067 manuscripts, 44 were included in the evidence synthesis. Etoricoxib had the highest ranking for improving WOMAC pain (probability of being top ranked, p (best) = .43) followed by naproxen (p (best) = .12), acetaminophen (AAP) (p (best) = .04), and celecoxib (p (best) = .02). The top three ranked interventions were etoricoxib, celecoxib and aceclofenac in the higher pain group, and tramadol, celecoxib, and diclofenac in the lower pain group. Conclusion: In the overall analysis, etoricoxib, celecoxib, and aceclofenac had the highest rankings for improving WOMAC pain. The ability to improve knee OA symptoms may differ depending on baseline pain and radiologic features.