Conflict of interest in trials of internet-based interventions for depression: Systematic review and meta-regression.

Objective To investigate (a) the prevalence of industry sponsorship (InS), researchers' financial conflicts of interest (RFCOI) and researcher allegiance (RA) in clinical trials of internet-based interventions (IBI) for depression and (b) their possible relation with treatment effects. Method We conducted a systematic review and meta-analysis of 94 comparisons of therapist-guided or self-guided IBI with a passive control group. Results InS was present in 13 (38.24%) out of 34 comparisons of commercial IBI; RFCOI was present in 29 (30.85%), RA in 75 (79.79%) out of all 94 comparisons. In primary meta-regressions, presence of InS and RFCOI was significantly associated with larger treatment effects, while presence of RA was not. Overall, sensitivity analyses showed consistent results, but suggested differential results for trials of therapist-guided and self-guided IBI: While InS and RFCOI showed larger effects in trials of self-guided IBI, no effect was found for trials of therapist-guided IBI. Conclusion InS, RFCOI, and RA are prevalent and should receive more attention in IBI research. The role of RA on outcome warrants further investigation because this meta-analysis did not include comparative trials. Small effect sizes in trials without InS or RFCOI suggest that current meta-analytic efficacy estimates of self-guided IBI for depression are overestimated.