Self-guided Internet-based cognitive behavior therapy (iCBT) for migraine interventions could improve access to care, but there is poor evidence of their efficacy.

A three-arm randomized controlled trial compared: iCBT focused on psychoeducation, self-monitoring and skills training (SPHERE), iCBT focused on identifying and managing personal headache triggers (PRISM) and a waitlist control. The primary treatment outcome was a ≥ 50% reduction in monthly headache days at 4 months post-randomization.

428 participants were randomized (mean age = 30.1). 240 participants (56.2%) provided outcome data at 4 months. Intention-to-treat (ITT) analysis with missing data imputed demonstrated that the proportion of responders with a ≥ 50% reduction was similar between combined iCBTs and waitlist (48.5/285, 17% vs. 16.6/143, 11.6%, p = 0.20), but analysis of completers showed both iCBT programs to be superior to the waitlist (24/108, 22.2% vs. 13/113, 11.5%, p = 0.047). ITT analysis with missing data imputed showed no difference between the two iCBTs (SPHERE: 24.8/143, 17.3% vs. PRISM: 23.7/142, 16.7%, p = 0.99). Uptake rates of the iCBTs were high (76.9% and 81.69% logged in at least once into SPHERE and PRISM, respectively), but adherence was low (out of those who logged in at least once, 19.01% [21/110] completed at least 50% modules in SPHERE and 7.76% [9/116] set a goal for trying out a given trigger-specific recommendation in PRISM). Acceptability ratings were intermediate.

Self-guided iCBTs were not found to be superior in our primary ITT analysis. Low adherence could explain the lack of effects as completer analysis showed effects for both interventions. Enhancement of adherence should be a focus of future research.