Routine Alcohol Screening Scores Linked to Mortality Risk in Primary Care Patients

A retrospective cohort study examining more than 500 000 primary care patients found that alcohol consumption levels reported on routine healthcare screenings were associated with all-cause mortality risk, with the strongest associations observed in younger adults.¹

The study, published in Alcohol: Clinical and Experimental Research, analyzed data from 531 851 adult patients at Kaiser Permanente Washington who completed the three-item Alcohol Use Disorder Identification Test, Consumption (AUDIT-C) screening between March 1, 2015, and December 31, 2021. Patients were followed for up to 8 years, during which 21 548 deaths (4.8%) occurred.¹

Using Cox proportional hazards models adjusted for demographic and clinical covariates, researchers found a convex (J- or U-shaped) association between AUDIT-C scores and mortality. Compared with patients reporting low-risk alcohol use (the reference group), those reporting very high-risk use had a 58% higher hazard of death (HR = 1.58 [95% CI: 1.35-1.84], p < 0.001), while those reporting no alcohol use had a 40% higher hazard (HR = 1.40 [95% CI: 1.36-1.44], p < 0.001).¹

Patients with moderate-risk alcohol use demonstrated a 14% lower hazard of death compared with the low-risk group (HR = 0.86 [95% CI: 0.83-0.90], p < 0.001). High-risk alcohol use showed no significant difference in mortality compared with low-risk use (HR = 0.97 [95% CI: 0.83-1.13], p = 0.72).¹

The association between alcohol consumption and mortality varied significantly by age (likelihood ratio test p < 0.001). Among patients aged 18-29 years, those with very high-risk alcohol use had nearly a threefold higher hazard of death compared with low-risk drinkers (HR = 2.98 [95% CI: 1.45-6.14], p = 0.003). This pronounced pattern diminished with increasing age and was not observed in patients aged 75 years and older.¹

No significant effect modification by sex was identified (likelihood ratio test p = 0.20).¹

“These findings confirm earlier research linking alcohol use and mortality, but this is the first study to do so using routinely collected alcohol screening scores in a large, generalizable primary care population,” lead author Tessa Matson, PhD, MPH, a collaborative scientist at Kaiser Permanente Washington Health Research Institute (KPWHRI), said in a press release.² “This reinforces the value of AUDIT-C as a practical tool that enables primary care clinicians to recognize high-risk alcohol use and engage patients in timely, supportive conversations about reducing their risk.”

The study population was predominantly middle-aged (52% between ages 30 and 59), female (58%), White (70%), and non-Hispanic (94%). Most patients had commercial insurance (63%), with smaller proportions having Medicare (19%), Medicaid (3%), or state-subsidized insurance (6%). Current tobacco use was reported by 10% of patients, and 3% had a substance use disorder diagnosis documented in the previous 2 years.¹

"This study adds to the growing body of research demonstrating the clinical utility of AUDIT-C scores completed in routine care," the authors wrote. The findings suggest that brief alcohol screening measures administered during routine primary care can identify patients at elevated risk for major adverse health outcomes.¹

The AUDIT-C is a validated three-item measure that assesses past-year alcohol consumption through questions on drinking frequency, quantity, and frequency of heavy episodic drinking. Scores range from 0 to 12, with higher scores indicating greater alcohol consumption and risk.¹

The authors noted several limitations, including the inability to account for changes in drinking patterns over time, potential underreporting of alcohol consumption on routine care screenings, and limited generalizability due to the predominantly commercially insured, non-Hispanic White population from a single state. The AUDIT-C's past-year timeframe does not differentiate between individuals who have never consumed alcohol and those who recently stopped.¹

References: