Factors associated with healthcare utilization and mortality in alcohol use disorder hospitalization

Abstract Objective To examine the outcomes of alcohol use disorder (AUD)-hospitalizations. Methods We used the U.S. National Inpatient Sample (NIS) data from 1998-2016 to examine predictors of Healthcare utilization (total hospital charges, discharge destination, length of hospital stay) and in-hospital mortality for AUD-hospitalization outcomes. We used logistic regression to assess the association of patient demographics (age, sex, race/ethnicity, household income), comorbidity (Deyo-Charlson index), insurance payer (Medicare, Medicaid, private, self-pay or other), and hospital characteristics (location/teaching status; hospital bed size; and region) with outcomes Results There were 5,590,952 primary AUD-hospitalizations from 1998-2016; of these 106,419 (1.9%) died in-hospital. The mean age was 48 years, 73% were male, 59% white, and 57% had a Deyo-Charlson comorbidity score of zero. In multivariable-adjusted analyses of AUD-hospitalizations, older age, female sex, higher Deyo-Charlson index, rural location or hospitals with medium or large bed sizes were associated with significantly higher in-hospital mortality and healthcare utilization. Similarly, South or Western U.S. hospital location, White race, or highest income quartile with higher healthcare utilization. Compared to a private insurance payer, Medicare or Medicaid insurance payers were associated with higher healthcare utilization, but lower in-hospital mortality. Conclusions We identified several independent associations of modifiable and non-modifiable factors with healthcare utilization and mortality outcomes for AUD-hospitalizations. These findings provide an opportunity for prognosis, resource allocation and the development of interventions to improve outcomes of AUD-hospitalizations in the future.