Fifteen trials were included in the review (number of participants unclear): seven had control groups, five had a crossover design and three did not have control groups. Four studies were classified as Class 1 evidence, one as Class 2 evidence, six as Class 3 evidence and four as Class 4 evidence. A full breakdown of the results was provided in the review paper.
Anticonvulsants (six trials: one with Class 1 evidence, two with Class 3 evidence and three with Class 4 evidence): A statistically significant greater reduction in pain scores was observed in anticonvulsant treatment groups versus comparator groups (pooled Cohen's d -1.88, 95% CI -3.13 to -0.64; four trials; 78 participants).
Cannabinoids (four trials: two with Class 1 evidence and two with Class 3 evidence): No statistically significant differences in pain scores were observed between cannabinoid treatment groups and comparator groups (pooled Cohen's d 0.08, 95% CI -0.74 to 0.89; three trials; 565 participants).
Other pharmacological interventions (five trials): Mixed results for pain reduction were found with use of antidepressants (two trials: one with Class 1 evidence and one with Class 3 evidence). One trial with Class 1 evidence found an improvement in pain reduction with the use of dextromethorphan/quinidine compared with placebo (Cohen's d-0.22). Two placebo-controlled crossover trials (one with Class 3 evidence and one with Class 4 evidence) each showed reductions in pain with the use of the opioid antagonist naltrexone (mean score difference between groups -2.13) and the opioid agonist morphine (Cohen's d -0.48).
In the four trials with Class 1 evidence, dizziness, nausea and somnolence were the most commonly reported adverse events. Further results (including data on adverse events per drug type) were reported in the review paper.