Can 2 screening questions accurately detect alcohol and other drug abuse in the primary care setting?

Journal of Family Practice, June, 2001 by Mandy Gittler, Frederick Chen

Brown RL, Leonard T, Saunders LA, Papasouliotis O. A two-item conjoint screen for alcohol and other drug problems. J Am Board Fam Pract 2001; 14:95-106.

* BACKGROUND Currently, most screening questions in the primary care setting address only alcohol abuse, but it is also important to screen for other substance abuse. The authors of this study examined the validity of a 2-question screening test for alcohol and other substances.

* POPULATION STUDIED The participants were 1252 patients randomly selected from 3 University of Wisconsin primary care clinics. They ranged in age from 18 to 59 years; 68% were women; and the population was racially diverse. Sixty-nine percent had private insurance. The prevalence of substance disorders in this study population was high (23%).

* STUDY DESIGN AND VALIDITY This prospective cohort study sought to identify 2 questions that could accurately screen for substance abuse. Screening questions were developed through a literature search and a series of focus groups. The interviewers asked patients the screening questions and then administered the Composite International Diagnostic Interview-Substance Abuse Module (CIDI-SAM), a series of questions based on the Diagnostic and Statistical Manual of Mental Disorders, third edition, revised used for identifying substance use disorders. The sensitivity of the screening questions was defined as the proportion of patients with a current substance abuse problem, as detected by CIDI-SAM, who had positive screening results. The high prevalence of substance abuse in this study will improve the positive predictive value of the screening test.

This study was well designed and had a high response rate. Although the interviewers were not blinded while administering the screening questions and the CIDI-SAM, they were monitored closely, and a random subset of patients underwent urine drug testing. The authors also tested the validity of the screening questions with questions about patient comfort and truthfulness.

* OUTCOMES MEASURED The primary outcome measured was identification of either alcohol or drug abuse.

* RESULTS Among 1136 participants (91%), a positive response to either of 2 questions was 80% sensitive and specific for current alcohol or drug abuse. With the high prevalence of substance use in this sample, the positive predictive value of the 2 questions was 52% and the negative predictive value was 93%. Comfort with the interviewer was reported by 84% of the participants, and withholding any information by only 11%. Random urine substance screening of 231 patients revealed 14 positive test results in patients with a negative CIDI-SAM, but these results did not change the performance of the screening test.

RECOMMENDATIONS FOR CLINICAL PRACTICE

A positive response to either of 2 questions ("Have you felt you wanted or needed to cut down on your drinking or drug use in the last year?" and "In the last year, have you ever drunk or used drugs more than you meant to?") detected current substance abuse with 80% sensitivity and specificity. As with all screening tests, the performance of the questions will vary with the prevalence of substance abuse in your patient population.

Mandy Gittler, MD Frederick Chen, MD, MPH University of Washington Seattle E-mail: mgittler@u.washington.edu