Psychological interventions for harmful gambling: Systematic review and network meta-analysis.

Background and aims: Gambling causes harms for the individual, their families and wider society. Psychological interventions have been shown to improve gambling severity, but their relative effectiveness is not well understood. Methods: Systematic review and network meta-analysis (NMA) of psychological interventions for harmful gambling. Outcomes were gambling symptom severity and gambling frequency (expressed as standardized mean differences [SMDs]). Results: We included 52 randomized controlled trials (RCTs), 9013 participants and 14 intervention classes. For gambling symptom severity, group CBT was the most effective intervention (SMD -1.09, 95% credible intervals [95%CrIs] -1.75 to -0.44) followed by individual CBT (SMD -0.56, 95%CrIs -1.07 to -0.07). Waitlist showed increased symptom severity compared with no treatment (SMD 0.96, 95%CrIs 0.47 to 1.46). All other intervention effects were characterized by uncertainty. For gambling frequency, all active interventions showed similar effects to treatment as usual. Discussion: Included studies were of low or very low quality and there was moderate-to-high heterogeneity for the gambling symptom severity NMA. However, threshold analysis suggested that conclusions based on the NMA were robust to potential biases and sampling variation in the included evidence. Bias adjustment analyses did not show evidence of bias associated with small studies or industry funding. Conclusions: Group CBT shows clinically important benefits and should be regarded as the best intervention for harmful gambling. For individuals who decline group CBT, or where this is not possible or suitable, individual CBT shows the next best evidence of benefit.