Objective
The objective of this study is to evaluate the comparative benefits and harms of opioids and cannabis for medical use for chronic non-cancer pain.

Design
Systematic review and network meta-analysis.

Data sources
EMBASE, MEDLINE, CINAHL, AMED, PsycINFO, PubMed, Web of Science, Cannabis-Med, Epistemonikos and the Cochrane Library (CENTRAL) from inception to March 2021.

Study selection
Randomised trials comparing any type of cannabis for medical use or opioids, against each other or placebo, with patient follow-up ≥4 weeks.

Data extraction and synthesis
Paired reviewers independently extracted data. We used Bayesian random-effects network meta-analyses to summarise the evidence and the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) approach to evaluate the certainty of evidence and communicate our findings.

Results
Ninety trials involving 22 028 patients were eligible for review, among which the length of follow-up ranged from 28 to 180 days. Moderate certainty evidence showed that opioids provide small improvements in pain, physical functioning and sleep quality versus placebo; low to moderate certainty evidence supported similar effects for cannabis versus placebo. Neither was more effective than placebo for role, social or emotional functioning (all high to moderate certainty evidence). Moderate certainty evidence showed there is probably little to no difference between cannabis for medical use and opioids for physical functioning (weighted mean difference (WMD) 0.47 on the 100-point 36-item Short Form Survey physical component summary score, 95% credible interval (CrI) −1.97 to 2.99), and cannabis resulted in fewer discontinuations due to adverse events versus opioids (OR 0.55, 95% CrI 0.36 to 0.83). Low certainty evidence suggested little to no difference between cannabis and opioids for pain relief (WMD 0.23 cm on a 10 cm Visual Analogue Scale (VAS), 95% CrI −0.06 to 0.53) or sleep quality (WMD 0.49 mm on a 100 mm VAS, 95% CrI −4.72 to 5.59).

Conclusions
Cannabis for medical use may be similarly effective and result in fewer discontinuations than opioids for chronic non-cancer pain.