• Asher J Schranz, Casey Tak, Li-Tzy Wu, Vivian H Chu, David A Wohl, and David L Rosen.
• Department of Medicine, University of North Carolina, Chapel Hill, NC, USA. aschranz@med.unc.edu.
• J Gen Intern Med. 2023 May 1; 38 (7): 161516221615-1622.

BackgroundHospitalizations for infective endocarditis (IE) associated with opioid use disorder (O-IE) have increased in the USA and have been linked to high rates of discharge against medical advice (DAMA). DAMA represents a truncation of care for a severe infection, yet patient outcomes after DAMA are unknown.ObjectiveThis study aimed to assess readmissions following O-IE and quantify the impact of DAMA on outcomes.DesignA retrospective study of a nationally representative dataset of persons' inpatient discharges in the USA in 2016 PARTICIPANTS: A total of 6018 weighted persons were discharged for O-IE, stratified by DAMA vs. other discharge statuses. Of these, 1331 (22%) were DAMA.Main MeasuresThe primary outcome of interest was 30-day readmission rates, stratified by discharge type. We also examined the total number of hospitalizations during the year and estimated the effect of DAMA on readmission.Key ResultsCompared with non-DAMA, those experiencing DAMA were more commonly female, resided in metropolitan areas, lower income, and uninsured. Crude 30-day readmission following DAMA was 50%, compared with 21% for other discharge types. DAMA was strongly associated with readmission in an adjusted logistic regression model (OR 3.72, CI 3.02-4.60). Persons experiencing DAMA more commonly had ≥2 more hospitalizations during the period (31% vs. 18%, p<0.01), and were less frequently readmitted at the same hospital (49% vs 64%, p<0.01).ConclusionsDAMA occurs in nearly a quarter of patients hospitalized for O-IE and is strongly associated with short-term readmission. Interventions to address the root causes of premature discharges will enhance O-IE care, reduce hospitalizations and improve outcomes.© 2022. The Author(s), under exclusive licence to Society of General Internal Medicine.

20 keywords...

hide…