Background: Self-directed discharges, also referred to as discharge “against medical advice” (AMA), comprise 1–2% of hospital discharges. Practice guidelines suggest avoiding the designation of “AMA” discharge as this is associated with adverse outcomes. The term “AMA discharge” persists, however, and characteristics of clinician documentation of self-directed discharges remain underexplored. We used retrospective chart review at our tertiary academic medical center to review hospitalists’ and nurses’ documentation around the time of self-directed discharges to identify how often the term “AMA” was used and other documented details of the discharge.
Methods: Using our electronic health record, we identified 1126 discharges designated as “AMA” for hospital medicine patients between 2017 and 2023. Every sixth case was reviewed, yielding a sample of 165 patients. Two physician reviewers independently assessed the charts to review demographic information and hospitalists’ and nurses’ discharge documentation, with a third physician resolving discrepancies and verifying inter-rater reliability.
Results: Demographic data regarding the selected patients is summarized in Table 1. Many of the patients had a psychiatric disorder (n=34, 21%) and/or an unhoused or marginally housed status (n=32, 19%), and most of the patients (n=96, 58%) had a history of substance use disorder. Hospitalists used the term “AMA” in 81% of notes (n=134) and neutral alternatives like “self-directed discharge” in 17% of notes (n=22). Nurses used the term “AMA” in 70% of notes (n=116) and neutral alternatives like “self-directed discharge” in 11% of notes (n=18). Documented barriers to medically recommended discharge included a combination of therapeutic needs (n=126, 76%), pending diagnostic studies (n=34, 21%), and/or care coordination barriers (n=25, 15%). Documented patient-stated reasons for self-directing their discharge included disagreement with or mistrust of the treatment plan (n=30, 18%) and dissatisfaction with the pain control regimen (n=19, 12%), though 31% of charts (n=51) lacked documented rationale. Capacity assessments were absent in 23% (n=38), documentation of risk-benefit discussions was missing in 13% (n=21), and both were absent in 18% (n=29). Concerns about staff safety around the time of the self-discharge were noted in 28% of cases (n=46), with “verbal abuse” being the most frequently cited issue (13%, n=24).
Conclusions: This review of documentation practices regarding self-directed discharges found that most hospitalists and nurses still use the term “AMA” and fail to consistently document capacity assessments and risk-benefit discussions. Our study highlights the need for further education and standardization around documentation practices of self-directed discharges to ensure comprehensive, accurate, and patient-centered documentation and avoid perpetuating bias and stigma in this largely marginalized patient population.
