Alcohol use in the year following approach bias modification during inpatient withdrawal: secondary outcomes from a double-blind, multi-site randomized controlled trial.

Abstract

Approach bias modification (ApBM) targeting alcohol approach bias has been previously shown to reduce likelihood of relapse during the first 2 weeks following inpatient withdrawal treatment (IWT). We tested whether ApBM's effects endure for a longer period by analysing alcohol use outcomes 3, 6 and 12 months post-discharge. A double-blind, sham-controlled randomized controlled trial. Four IWT units in Melbourne, Australia. Three hundred alcohol IWT patients (173 men, 126 women, 1 non-binary; mean age 43.5 years) were recruited between 4 June 2017 and 14 July 2019. Follow-up data collection was completed on 22 September 2020. Four ApBM sessions were delivered during IWT. ApBM trained participants (n = 147) to avoid alcohol and approach non-alcohol beverage cues. Controls (n = 153) responded to the same stimuli, but without approach/avoidance training. Date of first lapse was recorded for non-abstinent participants to determine time to first lapse. Time-line follow-back interviews assessed past-month alcohol consumption at each follow-up, with participants reporting no alcohol consumption classified as abstinent. In analyses of past-month abstinence, non-abstinence was assumed in participants lost to follow-up. Number of past-month drinking days, standard drinks and heavy drinking days (five or more standard drinks for women or non-binary; six or more standard drinks for men) were calculated for non-abstinent participants at each follow-up. ApBM significantly delayed time to first lapse [ApBM median: 53 days, 95% confidence interval (CI) = 21-61; controls = 12 days, 95% CI = 9-21, P = 0.045]. Past-month abstinence rates at 3-, 6- and 12-month follow-ups were 33/153 (21.6%), 30/153 (19.6%), and 24/153 (15.7%) in controls; and 51/147 (34.7%), 30/147 (20.4%) and 29/147 (19.7%) in the ApBM group, respectively. Past-month abstinence was significantly more likely in ApBM participants than controls at the 3-month follow-up [odds ratio (OR) = 1.93, 95% CI = 1.16-3.23, P = 0.012], but not at 6- or 12-month follow-ups (6-month OR = 1.05, 95% CI = 0.60-1.95, P = 0.862; 12-month OR = 1.32, 95% CI = 0.73-2.40, P = 0.360). No significant group differences were found for indices of alcohol consumption in non-abstinent participants. Approach bias modification for alcohol delivered during inpatient withdrawal treatment helps to prevent relapse, increasing rates of abstinence from alcohol for at least 3 months post-discharge.