Does moderate alcohol use affect health-care costs? A propensity analysis of female health-care workers.

To determine differences in health-care costs associated with moderate alcohol consumption among female health-care workers while controlling for other risk factors that may be correlated with alcohol use. Non-randomized, prospective, observational study of health-care costs by female health-care workers in a large managed care organization recruited between 1 January 1998 and 1 July 2000. Six hundred and eighty-five female employees, continuously and stably employed by the managed care organization, who received health-care through the affiliated managed care organization. All women completed a health risk appraisal as part of the company's Employee Wellness Plan; 218 women were categorized as moderate drinkers and 467 as abstainers/light drinkers. Total costs of in-plan and out-of-plan health-care utilization, by type of service, during the 6 month period after completing the health risk survey were calculated. Using 218 one-to-one matched pairs of moderate drinkers and abstainers/light drinkers, no significant differences in total, outpatient or inpatient costs were observed during the 6 month observation period. Pharmacy costs were significantly lower for moderate drinkers (–$43, 95% CI = –$88.82 to –$2.41), primarily due to differences in costs from anxiolytic (including barbiturates and benzodiazepines), hypnotic and sedative drug fills. Findings demonstrate the value of risk factor matching when studying the relationship between alcohol use and health-care utilization. The discovery of differential pharmacy utilization raises the possibility that alcohol consumption may reduce the use of prescribed central nervous system depressants. [ABSTRACT FROM AUTHOR]