The efficacy of an internet-based cognitive behavioral program added to treatment-as-usual for alcohol-dependent patients in primary care: a randomized controlled trial.

Abstract

Most alcohol-dependent people have a moderate level of dependence. General practitioners (GPs) hesitate to engage in this area, and need to have access to treatment they find applicable and feasible to use. The aim of this present study was to test if an open-ended internet-based cognitive-behavioral therapy (iCBT) program added to treatment-as-usual (TAU) is more effective than TAU-only for alcohol-dependent patients in primary care. The present study was a two-group, parallel, randomized controlled superiority trial comparing iCBT+TAU versus TAU-only at 3- and 12-month follow-ups. TAU was delivered at 14 primary care centers in Stockholm, Sweden. A total of 264 patients (mean age 51 years, of whom 148 were female and 116 were male) with alcohol dependence and hazardous alcohol consumption were enrolled between September 2017 and November 2019. Participants were randomized at a ratio of 1:1 to iCBT, as a self-help intervention added to TAU (n = 132) or to TAU-only (n = 132). The GPs gave participants in both treatment arms feedback on the assessments and biomarkers and offered TAU at the primary care center. Primary outcome was weekly alcohol consumption in g/week at 12-month follow-up, analyzed according to intention-to-treat (n = 132 + 132). The per-protocol analysis included participants who completed at least one module of iCBT (n = 102 + 132). There was no significant difference in weekly alcohol consumption between iCBT+TAU and TAU in the intention-to-treat (ITT) analysis at 12-month follow-up [iCBT+TAU = 133.56 (95% confidence interval, CI = 100.94-166.19) and TAU = 176.20 (95% CI = 144.04-208.35), P = 0.068, d = 0.23]. In the per-protocol analysis, including only those who initiated iCBT, the iCBT+TAU group showed lower mean weekly alcohol consumption compared with TAU [iCBT+TAU = 107.46 (95% CI = 71.17-143.74), TAU = 176.00 (95% CI = 144.21-207.80), P =0.010, d = 0.42]. In Sweden, an internet-based cognitive-behavioral program added to treatment-as-usual to reduce alcohol consumption showed weak evidence of abenefit at 12 months in the intention-to-treat analysis and good evidence of a benefit in the per-protocol analysis.