# Feasibility and acceptability of an adapted WHO alcohol brief intervention: Pilot of a three-armed randomized trial in Sri Lanka

## TL;DR

This study tested a modified WHO alcohol intervention in Sri Lanka, finding it acceptable and feasible for future trials, though some adjustments are needed.

## Contribution

The study adapts and pilots a WHO alcohol brief intervention in Sri Lanka, evaluating its feasibility and acceptability in a local context.

## Key findings

- The adapted brief intervention was highly acceptable to patients and research assistants.
- Follow-up rates varied across the three study arms, with the highest for the ABI arm.
- Inpatient settings may not be ideal for future trials due to low usual drinking activity.

## Abstract

Risky drinking (RD) is a major health hazard in Sri Lanka. Alcohol brief intervention (BI) has been proven effective in minimizing RD but has not been utilised in Sri Lanka. We therefore aimed to adapt the WHO alcohol BI and targeted educational material to Sri Lanka, assess their feasibility and acceptability and evaluate appropriateness of methodology and measures for a future RCT.

A three-arm parallel-group pilot RCT.

The BI was adapted based on expert feedback. The study included male inpatients (with AUDIT-C screening score ≥5) of a tertiary hospital. The three study arms were: adapted brief intervention (ABI), education about unit of alcohol (UOA), and feedback on screening results (FOA). Trained research assistants (RAs) screened and implemented the interventions. We report on follow-up rates (feasibility), participant and RA feedback (acceptability), recruitment efficiency and data quality (methodological appropriateness), and appropriateness of outcome measures.

The ABI included a structured training manual for implementers, an alcohol information leaflet, and a personal information sheet. Patient follow-up rates were 69 %, 40 % and 71 % for FOA, UOA and ABI arms respectively. Family member recruitment was 31 %. Patient and RA feedback for ABI was overwhelmingly positive. Many patients were abstinent at baseline (37.5 %) and follow-up (75.9 %), mainly due to health concerns. FMQ revealed high ‘total family burden’. Patients struggled with TLFB recall. High childhood adversity prevalence (95.7 %) and low alcohol knowledge were observed.

The ABI demonstrated high acceptability among patients and RAs. All three interventions could be trialled in a future RCT. All measures except TLFB proved appropriate. Our innovative approach of evaluating outcomes from family members' perspectives proved feasible and valuable. The inpatient setting was not appropriate, rather a setting where patients continue their day-to-day activities, including usual drinking, should be considered in a future RCT.