Validating the ability of a single-item assessing drunkenness to detect hazardous drinking

Abstract

Background: To simplify the screening process to identify problem drinking, researchers have developed single-item, quantity-based instruments. Quantity-based items, however, suffer from several limitations. Objective: Determine whether a non-quantity-based single item assessing drunkenness would be able to identify those who exhibited signs of hazardous alcohol use, as determined by Alcohol Use Disorders Identification Test (AUDIT)-C scores. Methods: Between September and November of 2011, the alcohol-related behaviors of 1062 self-identified current drinkers were assessed with self-report measures (i.e. AUDIT-C and one item assessing frequency of drunkenness) in addition to corresponding biologic samples (i.e. breath alcohol concentration (BrAC) samples). We assessed the concurrent validity of the single item to identify hazardous alcohol use established via gender-based AUDIT-C thresholds. Convergent validity of the single-item was assessed by determining its relationship to BrAC levels. Results: The single item accounted for 0.791 of the area under the received operating characteristics curve for hazardous alcohol use (p < 0.001). When employing a cut-off of 1, the single-item question was 95.9% sensitive in identifying hazardous alcohol use. Conclusion: The results reported herein highlight the validity (both concurrent and convergent) and potential utility of a non-quantity-based single-item alcohol screen for assessing drunkenness. Additional research is warranted to confirm the utility of this one-item drunkenness measure to capture risk of injury and hazardous drinking.

• Alcohol
• AUDIT-C
• drunkenness
• hazardous drinking
• screening-instrument
• single-item

Notes

¹“X” is replaced with “4” for female and “5” for male respondents. A response ≥ 1 is considered positive.

²“X” is replaced with “4” for female and “5” for male respondents. Responses are coded as never, more than 12 months ago, 3–12 months ago and within the past 3 months.