Discharge against medical advice (AMA) leads to worse outcomes in burn patients including higher readmission rates, wound complications, and elevated costs, although protective and risk factors remain incompletely characterized. Burn encounters were identified in two national databases, the Nationwide Inpatient Sample (NIS), 2016-2021, and the Burn Care Quality Platform (BCQP), 2016-2022. Encounters were stratified by AMA discharge. The primary outcomes were the overall incidence of AMA discharge and the effects of inpatient psychosocial or behavioral interventions and burn center status on AMA discharge, evaluated with 1:1 nearest neighbor propensity-score matching of AMA and non-AMA encounters on demographics, burn characteristics, and established risk factors for AMA discharge, without replacement and with maximum caliper distance of 0.1. In the NIS and BCQP, respectively, 5,950 (2.8%) of 214,390 and 2,923 (1.4%) of 205,993 burn encounters were discharged AMA. Significant predictors of AMA discharge included male sex, Medicaid or self-pay, smaller burns, and more recent year (p<0.01). The incidence of AMA discharges increased over the study period in both the NIS (IRR 1.2; 95% CI: 1.1-1.2; p<0.01) and the BCQP (IRR 1.2; 95% CI: 1.2-1.2; p<0.01). Within a propensity-score matched model, burn center care was associated with a 1.2% decrease in AMA discharge (ATE -0.01; 95% CI: -0.02- -0.00; p=0.01); psychosocial behavioral interventions did not impact the likelihood of AMA discharge (p=0.85). Existing interventions to alleviate AMA discharge are insufficient to significantly alter patterns of AMA discharge. Treatment at burn centers may result in lower likelihood of AMA discharge in burn patients.
Shah et al. (Thu,) studied this question.