Screening for Unhealthy Alcohol Use: Associations with Health Care Utilization, Costs, and Health-Related Quality of Life

Abstract

Background. Hospital readmissions and emergency department (ED) visits within 30 days of discharge are common among older adults. Unhealthy alcohol use could be a risk factor due to greater risk for injuries, poorer medication adherence and self-care, and medication interactions. As alcohol screening becomes more widely implemented, research is needed to determine whether screening results could identify patients at risk for post-discharge acute care and expenditures, and who may benefit from post-discharge interventions. Cost utility analyses (CUA) could help health care systems determine how to invest scarce resources. However, CUAs for alcohol interventions are limited due to the lack of preference weights for the full spectrum of alcohol use. Objectives. Research objectives are to 1) examine the association between unhealthy alcohol use and 30-day readmissions and ED visits, 2) examine the association between unhealthy alcohol use and 30-day acute care expenditures, and 3) estimate preference weights for the spectrum of alcohol use in the U.S. population. Methods. Objectives 1 and 2 were conducted among 579,330 older VA patients hospitalized for a medical/surgical condition who were screened for unhealthy alcohol use in the year prior to hospitalization. Objective 3 was conducted among 17,440 respondents to the National Health Interview and Medical Expenditure Panel Surveys. Adjusted preference-weights for two health-related quality of life (HRQOL) measures were estimated for different categories of alcohol use. Results. High-risk drinking was associated with increased risk for readmissions relative to low-risk drinking, but only among medical inpatients and not after adjusting for differences in SES and social support. High-risk drinking in medical inpatients was associated with increased acute care utilization but not increased acute care expenditures in the 30 days post-discharge. Nondrinking was associated with increased risk for readmissions, ED visits, and acute care expenditures. Preference weights did not differ with respect to low-risk drinking, with the exception of nondrinking and moderate-risk drinking. Conclusions. Patients with high-risk drinking based on alcohol screening results may not need targeting for post-discharge interventions. Self-reported unhealthy alcohol use may not be associated with poorer HRQOL, but lack of differences could also reflect the insensitivity of generic HRQOL measures to alcohol-specific domains.