Background: Discharge from the inpatient hospital setting against medical advice (AMA) poses significant challenges to continuity of care and portends increased risk of 30-day readmission and all-cause mortality. However, there are no practice guidelines for management of a patient who wishes to leave AMA and limited data is available on the clinical characteristics of such patients, communication between providers and patients at discharge, and post-discharge clinical outcomes.
Methods: A retrospective chart review of patients discharged AMA between March 31, 2019, and March 26, 2024, was conducted at an academic medical center under IRB approval. Patients were included if they were 18 years or older and discharged within the study timeframe with discharge disposition marked “Leaving against medical advice”, and excluded if they did not meet these criteria. Demographic information, medical comorbidities, and admission diagnosis were initially extracted, after which charts were manually reviewed for pertinent information surrounding the discharge encounter (e.g. documentation of patient decision-making capacity), discharge planning such as medications and appointments, and 30-day readmission outcomes.
Results: In total, 690 patients were identified for inclusion. Of these, 60% were male, with comorbidities most frequently observed including substance use disorder (43%), hypertension (36%), hyperlipidemia (31%), and anxiety (29%). Most patients (69%) were managed and discharged under Hospital Internal Medicine (HIM) service, with average length of stay 2.66 days. Patient’s decision-making capacity was documented in 47% of cases, while discussion of risks and benefits appeared in 67% of documentation. Prescriptions were ordered at discharge in 44% of cases. The median LACE+ index score was 59 (IQR 52-73), consistent with an observed 34% 30-day readmission rate, with median time to readmission being 5.0 days (IQR 1-13 days).
Conclusions: In this retrospective chart review, key characteristics associated with patients discharged AMA included male sex, substance use disorder, and admission to non-surgical services. AMA discharge management varied significantly amongst providers, with key elements–including assessment of decision-making capacity and discharge medication orders–frequently not performed. Consistent guidelines for shared decision-making and discharge planning during these encounters may improve clinical outcomes such as 30-day readmission (34%). Study strengths include a robust sample size and incorporation of validated scoring systems such as the LACE+ index. Limitations include the absence of a control group. Further studies can identify variations in AMA discharge management and potential effects of post-discharge outcomes.