Discontinuing a non-steroidal anti-inflammatory drug (NSAID) in patients with knee osteoarthritis: Design and protocol of a placebo-controlled, noninferiority, randomized withdrawal trial

Abstract

BackgroundKnee osteoarthritis (OA) is the most common cause of knee pain in older adults. Despite the limited data supporting their use, non-steroidal anti-inflammatory drugs (NSAID) are among the most commonly prescribed medications for knee OA. The use of NSAIDs for knee pain warrants careful examination because of toxicity associated with this class of medications. MethodsWe describe the design of a placebo-controlled, noninferiority, randomized withdrawal trial to examine discontinuation of an NSAID in patients with painful knee OA. Participants will be veterans enrolled in the VA Healthcare System with knee OA pain despite NSAID use and/or relatively higher risk of NSAID toxicity. After a two-week run-in period where eligible subjects will replace their current NSAID with the study NSAID (meloxicam), those remaining eligible (target N=544) will be randomized to receive four weeks of either placebo or continued meloxicam. The primary outcome is knee pain (Western Ontario and McMaster Universities Osteoarthritis Index pain subscale, range 0—20) at four weeks post-randomization. The primary hypothesis is that placebo will be noninferior to (that is, not much worse than) meloxicam within a noninferiority margin of 1. Secondary outcomes include lower extremity disability, global impression of change, adherence to study medication and use of co-therapies. DiscussionThis study is the first clinical trial to date examining the effects of withdrawing an NSAID for OA knee pain. If successful, this trial will provide evidence against the continued use of NSAIDs in patients with OA knee pain. Trial registrationClinicalTrials.gov: NCT01799213. Registered February 22, 2013.