|| Checking for direct PDF access through Ovid
We summarise the evidence for an association between screening scores from the Alcohol Use Disorders Identification Test (AUDIT) and all-cause mortality.Using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, prospective cohort studies reporting all-cause mortality risk by AUDIT scores (complete AUDIT-10 or AUDIT-C) were identified through MEDLINE, Embase, PubMed and Web of Science up to September 2016. Risk estimates were pooled using random effects meta-analyses.Seven observational studies with 18 920 observed deaths among 309 991 participants were identified. At-risk drinking (ie, hazardous/harmful consumption, AUDIT-10 ≥8 and AUDIT-C ≥4) was associated with elevated mortality risk after 2–10 years of follow-up (pooled relative risk (RR)=1.24, 95% CI 1.12 to 1.37) compared with moderate drinking (AUDIT-10=1–7, AUDIT-C=1–3). Compared to past year abstainers (AUDIT=0), moderate drinkers had a lower mortality risk (RR=0.75, 95% CI 0.71 to 0.79) in US Veterans and a similar mortality risk (RR=0.99, 95% CI 0.72 to 1.38) in population-based studies. Most data came from studies among Veterans using the short AUDIT-C in men and showed a dose–response relationship (RR=1.04, 95% CI 1.04 to 1.05 for each AUDIT-C score among drinkers). Data for women and young adults were scarce.AUDIT screening scores were associated with mortality risk. The association was differential depending on the population examined, which may be related to prevalence of former drinkers among current abstainers. Due to heterogeneity between studies and the small number of populations examined, generalisability may be limited.