Eighteen RCTs were identified (925 patients, range 13 to 160, from Table 1); 766 completed participants met the inclusion criteria. The trials had a mean Oxford Score of 6.1. There were 13 cross-over trials and the rest had parallel groups.
Overall efficacy: Fifteen of 18 trials showed a significant analgesic effect. Four of six trials reported significantly improved sleep; the effects were modest. Adverse events were described as transient or mild to moderate, and well tolerated, not leading to withdrawal from the trial. There were no serious adverse events according to the review definition. Withdrawals due to side effects ranged from 0% to 18%. Results for the level of function were presented (five trials).
Cannabis (four RCTs): All trials showed a significant improvement in neuropathic pain for smoked cannabis versus placebo with no serious adverse events.
Oromucosal extracts of cannabis based medicine (seven RCTs): Six trials showed a significant improvement in chronic pain versus placebo. Five of seven trials reported no serious adverse events. One trial of rheumatoid arthritis found a significant decrease in disease activity.
Nabilone (four RCTs): Three trials found a significant analgesic effect compared with placebo; the fourth trial found that both nabilone and dihydrocodeine reduced pain to a similar extent. One of four trials reported no serious adverse events.
Dronabinol (two RCTs): Both trials found a significant reduction in pain compared with placebo.
THC-11-oic acid analogue (CT-3 or ajulemic acid; one RCT): There was a significant improvement in pain intensity after three hours, but not after eight hours compared with placebo, with no serious adverse events.