Alpha Omega Alpha Honor Medical Society

2011 Research Abstract

Validation of the ASSIST for detecting at-risk alcohol use in urgent care clinics

Investigator: Justin Peterson, Mercer University School of Medicine

Co-investigators: J. Aaron Johnson, PhD, J. Paul Seale, MD, Nichole McCollum, Audrey Green, Anna Lee, Leslie Rumph, and Alex Woychek

Mentor: J. Aaron Johnson, PhD, Mercer University School of Medicine and Medical Center of Central Georgia, Department of Family Medicine

Screening, Brief Intervention, and Referral to Treatment (SBIRT) is effective in reducing alcohol use. While the SBIRT identifies patients with alcohol use disorders, the primary emphasis is on detecting and intervening with patients reporting at-risk alcohol use (exceeding NIAAA recommended daily or weekly limits). The World Health Organization developed the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST), an inexpensive international screening test. Much of the research surrounding the development of the ASSIST was focused on the detection of alcohol use disorders rather than at-risk drinking. This study uses a sample of 442 patients visiting medical center-affiliated urgent care clinics to identify ASSIST scores with the best specificity and sensitivity for detecting at-risk alcohol use. Patients completed a brief screen in the waiting room. Those identified as current drinkers were asked to complete a survey while in the exam room. The survey contained the ASSIST and 90 day timeline follow-back (TLFB), the “gold standard” for detecting at-risk alcohol use, as well as other measures. When compared to the TLFB, none of the ASSIST scores offer strong sensitivity and specificity. A score of 4+ has strong sensitivity (Se – 83.5) and modest specificity (Sp – 42.4) while a score of 6+ has the best combination of sensitivity and specificity (Se – 64.2, Sp – 63.6). Many SAMHSA funded SBIRT projects use an alcohol ASSIST score of 11+ to identify the need for a brief intervention. While offering a high specificity (Sp-90.3) a cutoff score of 11+ will not detect 60 percent (Se – 39.6) of at-risk alcohol users who might benefit from a brief intervention. While more research in varied healthcare settings would help to clarify the results of this study, the use of a lower ASSIST cutoff score is recommended to provide services to the largest possible percentage of at-risk drinkers.

Updated on May 14, 2012.


© 2017 Alpha Omega Alpha Honor Medical Society