Opioids for chronic osteoarthritis pain: An updated systematic review and meta-analysis of efficacy, tolerability and safety in randomized placebo-controlled studies of at least 4 weeks double-blind duration.

Abstract

This updated systematic review evaluated the efficacy and safety of opioids compared with placebo for chronic osteoarthritis pain. Clinicaltrials.gov, CENTRAL, MEDLINE and PsycINFO were searched from October 2013 to July 2019. Randomized controlled trials comparing opioids with placebo and at least 4weeks double-blinded duration were analysed. Primary outcomes were pain relief of 50% or greater, disability, tolerability and safety. Effects were summarized by a random effects model using risk differences or standardized mean differences with 95% confidence intervals. We added two new studies with 397 participants for a total of 22 studies with 8,942 participants. Study duration ranged between 4 and 24weeks. Studies with a parallel and cross-over design: Based on very low- to low-quality evidence, opioids provided no clinically relevant pain relief of 50% or greater and no clinically relevant reduction in disability compared with placebo. There was a clinically relevant harm related to the dropout rate due to adverse events. The frequency of serious adverse events did not differ from placebo. Enriched enrolment randomized withdrawal design: Based on very low- to low-quality evidence, opioids provided no clinically relevant pain relief of 50% or greater and no clinically relevant reduction in disability compared with placebo. Dropout rates due to adverse events and frequency of serious adverse events did not differ from placebo. Tolerability of opioids is low and efficacy is not clinically relevant in controlled studies from 4 to 24weeks for osteoarthritis pain. Within the context of randomized controlled trials (4-24weeks), opioids provided no clinically relevant pain relief and no clinically relevant reduction in disability compared with placebo in chronic osteoarthritis pain (hip, knee). Number needed to treat for an additional dropout due to side effects was 5 (95% confidence interval 4-7). Two studies found no signals of abuse and addiction. The frequency of serious adverse events including deaths did not differ from placebo.