Screening, brief intervention, and referral to treatment compared with treatment as usual for alcohol use in an integrated refugee settlement in Zambia: a hybrid, type 1, randomised controlled trial.

Abstract

BACKGROUND: Alcohol and other drug use represent a leading cause of preventable death and disability globally, disproportionately burdening low-resource settings. Alcohol and other drug use interventions remain neglected in humanitarian health research and programming. We evaluated the effectiveness of a multicomponent intervention in reducing alcohol use in a refugee setting in Zambia.

METHODS: We conducted a parallel, individually randomised trial in the Mantapala refugee settlement in Zambia. Eligible participants were Congolese refugees and Zambian host community members, aged 15 years and older, with Alcohol Use Identification Test (AUDIT) scores of 8 or more for males and 4 or more for females. Participants were allocated (1:1) to screening, brief intervention, and referral to treatment (SBIRT) or treatment as usual via a computer-generated randomisation scheme with allocation concealment by opaque envelopes. SBIRT included screening, a brief intervention, and indicated referral to psychotherapy for alcohol and other drug use and mental health conditions delivered by trained non-specialists. Treatment as usual comprised referral to primary health centre staff trained in basic management and care for alcohol use problems. The primary outcome was AUDIT score evaluated at 6 months. Linear mixed models and an intention-to-treat approach were used following multiple imputation. Implementation was assessed qualitatively and through cost-effectiveness analysis. Due to the nature of the trial, participants counsellors, and research assistants were unmasked at enrolment, although data analysts remained masked via dummy coding of the study group. The trial was registered on ClinicalTrials.gov (NCT05471921) and is completed.

FINDINGS: Between April 26, 2023, and Feb 29, 2024, 443 participants were screened for eligibility, 43 were excluded, and 400 were enrolled and randomly assigned (199 to the SBIRT group and 201 to the treatment as usual group). 299 (75%) of the 400 participants completed the 6-month assessment and 310 (78%) completed the 12-month assessment. The mean age of participants was 36·9 years (SD 12·1) and 287 (72%) participants were male and 113 (28%) were female. 232 (58%) of the 400 participants were Congolese refugees. AUDIT scores reduced from baseline to 6 months among the SBIRT group (-12·6 [95% CI -14·3 to -11·0]) and the treatment as usual group (-8·1 [-9·8 to -6·3]). The difference in mean change between groups was significantly greater in the SBIRT group than in the treatment as usual group at 6 months (-4·6 [95% CI -6·9 to -2·2]; p<0·0001; d=0·56) and 12 months (-3·5 [95% CI -5·8 to -1·1]; p=0·0030; d=0·43). The incremental cost per improvement in AUDIT at 6 months was US$158. There were no study-related adverse events.

INTERPRETATION: To our knowledge, this is the first known trial in a humanitarian setting of an intervention finding a significant, sustained effect on reduced alcohol use. SBIRT is a feasible and effective strategy to reduce alcohol-related burden in humanitarian and low-resource settings, with potential for integration into national and global health policies and systems.

FUNDING: Elrha's Research for Health in Humanitarian Crises Programme.