Background
Trauma exposure is associated with hazardous and/or harmful alcohol use. Occupational groups frequently exposed to trauma may be at risk of alcohol harm. This meta-analysis determined the prevalence of hazardous and harmful alcohol use across trauma-exposed occupations and meta-regressions explored the impact of pre-defined covariates on the variance in prevalence estimates.

Method
Literature was searched from 2000 to March 2020, using Scopus, Web of Science and PsycINFO. Studies were included in the meta-analysis if they used a standardized measure of alcohol use (e.g., Alcohol Use Disorder Identification Test (AUDIT)). Studies were excluded if they measured alcohol use following an isolated sentinel event (e.g., 9/11). The following occupations were included: first responders, health care workers, Armed Forces, war journalists and train drivers.

Results
1882 studies were identified; 55 studies were eligible. The pooled prevalence of hazardous use was 22% (95% Confidence Intervals [CI]: 17%–27%) and 11% (95% CI: 8%–14%) for harmful use. Hazardous alcohol use was significantly lower in health care workers (13%; 95% CI: 10%–16%) than first responders (26%; 95% CI: 20%–32%) and Armed Forces (34%; 95% CI: 18%–52%). There was marked heterogeneity across studies and higher prevalence rates in low-quality studies. The meta-regression identified higher proportion of males and younger mean age as predictors of variance.

Conclusions
Male-dominated occupations, such as police officers and military personnel, showed higher levels of hazardous and harmful alcohol use, indicating that interventions tailored specifically for these occupational groups may be needed.