steoarthritis affects more than 500 million people around the world, and it is a leading cause of disability. 1 Non-pharmacological strategies, such as exercise and maintaining a healthy weight, are recommended for first line management, as are simple analgesics, such as nonsteroidal anti-inflammatory drugs (NSAIDs) and paracetamol (acetaminophen). 2 However, advice on using opioid analgesics to treat the pain of knee and hip osteoarthritis is inconsistent; 3,4 opioids are often prescribed, including for about 40% of people with knee osteoarthritis in the United States. 5 Systematic reviews of placebo-controlled trials of the effectiveness of opioids for treating osteoarthritis pain have been limited in scope. For example, one evaluated only opioid treatments of at least four weeks' duration, 6 while a second was restricted to oral opioid therapy. 7 GRADE ratings were not always reported in the abstract or conclusions, and the validity of opioid dose-response analyses were sometimes unclear. 8 One review excluded tramadol because it had been examined in a separate review. 9 We therefore undertook a systematic review to provide a comprehensive evaluation of the efficacy and safety of opioid analgesic therapy regimens for people with osteoarthritis, and to explore dose-effect relationships.
MethodsWe searched MEDLINE, EMBASE, the Cumulative Index to Nursing and Allied Health Literature (CINAHL), PsycINFO, the Allied and Complementary Medicine Database (AMED), the Cochrane Database of Systematic Reviews, and the Cochrane Central Register of Controlled Trials (CENTRAL) for eligible randomised controlled trials (RCTs) published in any language in peer-reviewed journals to 31 October 2020 (Supporting Information, table 1). To avoid overestimating treatment effects, 10 we also searched the World Health Organization International Clinical Trials Registry (https://trial search.who.int) for unpublished trials. We screened the reference lists of retrieved publications to identify further RCTs. Our systematic review was prospectively registered with PROSPERO (CRD420191142813; 16 October 2019).
Inclusion criteriaWe included RCTs in which the analgesic effect of an opioid was compared with that of placebo in people with osteoarthritis of any type (knee, hip, hand, spine) of any duration. All single ingredient and combination opioid-containing analgesic regimens were included, regardless of opioid dose and administration route. We did not include trials in which the effect of opioid therapy was compared with other treatments but not with placebo. Trials including a range of pain conditions were eligible if data for participants with osteoarthritis could be extracted.
Study selectionThree authors (CAS, WA, GZ) independently screened the titles and abstracts and read the full text of potentially eligible publications; disagreements were resolved by consensus.