The Determinants of Against Medical Advice Hospital Discharges

Discharges against medical advice (AMA) are associated with adverse patient outcomes, higher readmission rates, and increased hospital expenses. Understanding the determinant of AMA is expected to guide the development of preventive measures.

A retrospective cohort study comparing adult inpatient AMA and non-AMA discharges at SUNY Upstate Medical University Hospital over a one-year period. Data was analyzed at-the-person as well as at-the-hospitalization level. The independent variables studied were age, race, sex, mental health, substance use, alcohol use, Social Deprivation Index (SDI), length of stay (LOS), season, and history of prior AMA. Fisher's exact test, Chi-square tests, and Wilcoxon rank sum tests were used for descriptive and bivariate analyses, logistic regression, and generalized estimating equation (GEE) models for multivariate analyses.

At-the-person level, increased risk for AMA (p < 0.05) was associated with being male, African-American, young adult, having a history of substance or alcohol use, and higher SDI. At-the-hospitalization level, AMA discharges featured shorter LOS, prior AMA, substance use, or alcohol use. Adjusted logistic regression identified male sex, African-American race, substance use, alcohol use, younger age, and SDI ≥ 80 as independent predictors of AMA discharge at-the-person level. GEE models yielded similar results at-the-hospitalization level, with prior AMA history being the strongest risk factor.

AMA discharges present a significant challenge for hospitals and compromise patient care. This study identified several key risk factors predictive of AMA. Knowledge of the determinants of AMA would allow for early identification of at-risk patients and for potential early interventions to prevent AMA discharges.