Background: The hip is the second most common large joint that is affected by osteoarthritis (OA), with prevalence ranging from 3% to 11% in patients aged ≥35 years. OA is often associated with significant pain, disability, and impaired quality of life. Treatment should be tailored according to the level of pain, disability, and handicap. Pharmacologic treatment options for hip OA include acetaminophen (recommended by the European League Against Rheumatism as a first-line treatment), NSAIDs such as diclofenac, and cyclooxygenase-2-selective NSAIDs such as celecoxib. Objective: The purpose of this study was to determine whether celecoxib 200 mg QD is noninferior to diclofenac 50 mg TID in the treatment of OA of the hip. Methods: This was a 12-week, randomized, double-blind, parallel-group, double-dummy, noninferiority study conducted at 40 centers in the United Kingdom. Patients with OA flare at baseline (determined by visual analog scale [VAS] measurement of ≥40 to <90 mm and patient's and physician's global assessments of arthritis ratings of “poor” or “very poor”) and awaiting joint replacement surgery were randomized to receive celecoxib QD or diclofenac TID. Patients were excluded if surgery was anticipated within 8 weeks. The United Kingdom National Health Service initiatives on waiting-list times caused a reduction in the number of potential patients available for participation. Therefore, the study protocol was amended such that change from baseline to week 6 (as opposed to week 12) in the patient's assessment of arthritis pain on walking, measured by VAS (0-100 mm), was the primary outcome. Primary analysis was carried out on the evaluable population (subjects with baseline and week 6 arthritis pain on walking VAS scores and no major protocol deviations). Celecoxib was declared noninferior to diclofenac if the upper limit of the 2-sided 95% CI of the treatment difference (celecoxib vs diclofenac) in the mean change from baseline in VAS did not exceed 10 mm. Tolerability was assessed by the documentation of observed and volunteered adverse events (AEs), physical examination findings, sitting blood pressure, and pulse at screening and at the end of the study (week 12 or early withdrawal). Results: A total of 249 patients aged ≥45 years were randomized to treatment. There were 126 patients in the celecoxib group and 123 patients in the diclofenac group. One patient in the celecoxib group did not receive any treatment and was excluded from analysis. Additionally, 54 patients in the celecoxib group and 45 patients in the diclofenac group discontinued treatment due to AEs and/or lack of treatment effectiveness. Therefore, 71 patients in the celecoxib group and 78 patients in the diclofenac group completed the study. No significant differences in demographic characteristics were observed between treatment groups. The mean (SD) age was 64.0 (9.0) years, 53.9% (76/141) of the patients were men and 46.1% (65/141) were women, and 99.3% (140/141) were white. At weeks 6 and 12, the patient's assessment of arthritis pain on walking (VAS) improved in both groups (-20.0 [23.6] mm in the celecoxib group and This work was presented in poster form at the European Federation of the International Association for the Study of Pain Chapters 5th Congress, September 13-16, 2006, Istanbul, Turkey. Accepted for publication November 28, 2007. -35 [27.0] mm in the diclofenac group [mean treatment difference, 14.4 mm; 95% CI, 6.1 to 22.7]). However, treatment differences in change from baseline favored diclofenac at week 6 (14.4 mm; 95% CI, 6.1 to 22.7) and week 12 (12.2 mm; 95% CI, 2.2 to 22.1). A post hoc analysis, performed after unblinding due to an imbalance in the numbers of patients previously receiving NSAIDs, found a greater treatment difference at week 6 between celecoxib and diclofenac in arthritis pain, favoring diclofenac, in previous nonusers of NSAIDs (n = 49, 18.6 mm; 95% CI, 4.5 to 32.8) compared with previous NSAID users (n = 92, 9.5 mm; 95% CI, -0.4 to 19.3). Celecoxib and diclofenac were generally well tolerated. A similar proportion of patients in both treatment groups experienced AEs (all causality): 67/125 of celecoxib-treated patients (53.6%) compared with 66/123 of diclofenac-treated patients (53.7%). Conclusion: This study did not demonstrate nonin-feriority of celecoxib 200 mg QD to diclofenac 50 mg TID in treating arthritis pain in patients with OA of the hip requiring joint replacement.