THE EFFECT OF PATIENT GENDER ON THE PREVALENCE AND RECOGNITION OF ALCOHOLISM ON A GENERAL MEDICINE INPATIENT SERVICE

Objectives: 1) to determine the rate of alcoholism among general internal medicine inpatients, 2) to assess the recognition and referral rates of these patients by their physicians, 3) to determine the effect of patient gender on physician recognition of alcoholism, and 4) to compare the observed alcoholism rates with rates reported in frequently cited studies, controlling for gender distribution. Design: Cross-sectional study, face-to-face interviews. Setting: A large, county-owned metropolitan teaching hospital. Patients/participants: Adult patients admitted to an inpatient general medical firm. From among 95 consecutive admissions, 78 patients (81%) entered the study. Intervention: The Michigan Alcoholism Screening Test (MAST) was administered to all study subjects. Chart reviews provided evidence of physician recognition and referral of patients with alcoholism. The observed rate of alcoholism was compared with rates reported in frequently cited studies after stratifying by type of service sampled and alcoholism assessment method used. Rates were then standardized for gender using the direct method. Measurements and main results: Twenty-two patients (28%) were found to be alcoholic by MAST criteria (scores of 5 or higher). Scores in the range indicative of alcoholism were observed more frequently among the 36 men than among the 42 women (p = 0.002) and varied by age group. Only the interaction between gender and age group was significant (p = 0.023). Sixteen of the 22 patients (73%) with alcoholism by MAST criteria were identified as alcoholic by physician evaluation. Physicians were significantly more likely to identify as alcoholic those patients with MAST scores higher than 29 and tended to more readily identify men who had alcoholism than women. Among physician-identified patients, only about one in five was referred for rehabilitation. The standardized alcoholism rate found (291/1,000) ranked about halfway between the highest and the lowest standardized rates from nine other studies of medicine inpatient services (465/1,000 and 112/1,000). Conclusions: Patient gender affected the prevalence of alcoholism and influenced its recognition by physicians. Alcoholism by MAST criteria was found in one in eight female and nearly one in two male inpatients. Physician recognition was higher for men and for more severely affected patients. An understanding of gender effects is essential to the appropriate interpretation of the results of screening tests for alcoholism and to understanding differences in reported crude rates of alcoholism among studies. Supplementing clinical impressions with the routine use of standardized methods for detecting alcoholism is recommended.