Physician Detection of Unhealthy Substance Use on Inpatient Teaching and Hospitalist Medical Services

Background: Screening, brief intervention, and referral to treatment for substance use depends on reliable identification. The goal of this study was to determine the rate of detection of unhealthy substance use by physicians on teaching and nonteaching medical services at a community teaching hospital. Methods: This cross-sectional study was conducted from February to June 2009. All new medicine admissions to the Teaching Service or the nonteaching Hospitalist Service were assessed for unhealthy substance use using the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C) and Drug Abuse Screening Test (DAST). All patients identified with substance use completed the Alcohol, Smoking, and Substance Involvement Screening Test (ASSIST). Medical record review was also performed to assess physician documentation. Results: Of 442 eligible patients, 414 consented to participate. Patients on the Teaching Service were more likely to be male, younger, unmarried, non-white, uninsured or receiving publicly funded insurance, and current smokers (p < .01 for all comparisons). Overall, the detection rate for unhealthy substance use was 64.3% (63 of 98 confirmed cases), with service-specific rates of 73.4% for the Teaching Service, compared with 47.1% for the Hospitalist Service (p = .011). ICD-9 coding accounted for 53.1% of identified cases on the Teaching Service and 14.7% of identified cases on the Hospitalist Service (p < .001). Assignment to the Hospitalist Service, being married, and isolated unhealthy drug use were independently associated with decreased physician detection rates. Conclusions: Our study suggests that unhealthy substance use is more likely to be detected on a Teaching Service than on a Hospitalist Service.