Alcohol interventions for persons with HIV: Meta-analysis of randomized controlled trials using phosphatidylethanol and self-report.

BACKGROUND

Interventions are needed to reduce alcohol use for people living with HIV (PWH), but prior randomized controlled trials (RCT) evaluated efficacy by self-reported alcohol use, potentially hampering validity. We aimed to determine alcohol intervention efficacy using the alcohol biomarker, phosphatidylethanol (PEth), combined with self-report, and compare results to each analysed alone.

METHODS

We conducted a systematic review of alcohol intervention RCTs to April 2023, followed by an individual participant data meta-analysis (IPD-MA) using two-step random effects modeling. Our primary outcome was unhealthy alcohol use defined as PEth/self-report (binary), i.e., PEth≥ 50ng/mL or Alcohol Use Disorders Identification Test - Consumption (AUDIT-C) ≥ 3 (women) and ≥ 4 (men). We also evaluated PEth and self-report alone.

RESULTS

We screened 280 studies, found 20 eligible, and obtained IPD for 16 (N = 3559, median age=41, 72.8 % male). Participants receiving an alcohol intervention had significantly lower odds of follow-up unhealthy alcohol use by PEth/self-report (OR=0.69, 95 % CI: 0.55-0.86; Cohen's d=0.21; I2=29.4 %, 95 % CI: 0.0 %-62.8 %). Risk-of-bias assessment indicated low or moderate risk. The certainty of evidence was moderate. Findings for PEth alone (OR=0.81, 95 % CI: 0.69-0.97; Cohen's d=0.12; I2=5.7 %, 95 % CI: 0.0 %-49.5 %) and self-report alone (OR=0.67, 95 % CI: 0.50-0.89; Cohen's d=0.22; I2=68.7 %, 95 % CI: 6.6 %-84.5 %) outcomes were similar, but heterogeneity was greater for self-report alone. Findings were robust to higher PEth/self-report cutoffs and continuous measures.

CONCLUSIONS

Results confirm prior findings of significant efficacy of alcohol interventions for PWH. Effect sizes were small across measurements, while heterogeneity was high when using self-report alone. Combined PEth/self-report is a useful primary outcome for alcohol intervention studies.