What Are Best Practices for Patients Discharged against Medical Advice?

Of note, one study found that race did not act as an independent predictor of AMA discharge when adjusted for age, gender, and socioeconomic factors.¹¹

The AMA population is clinically heterogeneous. Among patients with pneumonia, for example, Saitz et al showed that a patient’s documented clinical severity did not independently predict AMA discharge, suggesting that there is great clinical heterogeneity even among AMA patients with similar admission diagnoses.¹²

These studies highlight the clinical and demographic heterogeneity within this population, suggesting that patients discharged AMA require individualized attention from hospitalists and other healthcare providers.

Patients describe numerous motivations for leaving the hospital prematurely, including needing to pick up public-assistance checks, personal financial issues, and familial obligations.¹³ Interestingly, in the cohort of HIV patients referenced above, discharge on the day welfare checks were distributed was an independent predictor of AMA discharge.⁶ In focus groups composed of patients discharged AMA and their treating nurses and physicians, several themes were identified as potential contributors to AMA discharge, including drug addiction, pain management issues, external obligations, wait time, the physician’s bedside manner, being in a teaching hospital, and communication issues.¹⁴

Clearly, patients have a diversity of reasons for requesting to be discharged AMA, and further research is necessary to define clear and potentially modifiable risk factors.

Discussion

The clinical scenarios outlined above present two patients with very different clinical presentations and outpatient support systems as well as demonstrate the great variability in clinical risk at the time of discharge AMA. These examples emphasize the importance of an individualized approach to care for each patient.

In Case No. 1, the patient is admitted with a mild asthma exacerbation with persistent bronchospasm, though she clinically appears well and has reliable follow-up. In contrast, in Case No. 2, the patient has life-threatening disease and no established primary care physician or mechanism for outpatient care. These examples demonstrate extremes on the clinical and psychosocial spectrum of patients requesting an “early” discharge and suggest that no two patients at risk of AMA discharge are the same. Patient 1 could likely be safely managed at home with close outpatient follow-up, while Patient 2 presents a high-risk scenario with very few safe outpatient treatment options.

We suggest that an individualized approach be taken for each patient, with attention to both clinical and psychosocial risk. In clinically low-risk cases (e.g., Case No. 1), an approach that prioritizes shared decision making and coordination with the outpatient care team may be preferable to an AMA discharge, particularly given the often adversarial nature of the later.² In such cases, a collaborative approach may provide greater opportunity for harm reduction, provision of appropriate prescriptions, and follow-up appointments. In clinically high-risk patients such as Case No. 2, however, premature discharge is clearly inappropriate. Even in such clinically high-risk cases, however, we argue that a collaborative strategy aimed at identifying and addressing the patient’s psychosocial concerns is appropriate, as such an approach promotes shared decision making, builds trust between the patient and the care team, and therefore may facilitate improved follow-up at the time of discharge. Research is needed to formally assess the optimal approach for this patient population, including impact on rates of AMA discharge and the quality of post-discharge follow-up.

At present, the decision to classify a discharge as AMA falls solely on the treating provider, and we suspect that there is great variability in practice patterns, particularly as there are few established professional society practice guidelines regarding this difficult issue. As with all discharges from the hospital, the burden falls on the provider to engage the patient in shared decision making and ensure that the patient has the capacity to understand the risks and benefits of the proposed treatment plan. It is in this spirit that simply “filling out an AMA form” does not provide legal protection to a physician who does not adequately explain the full risks and benefits of refusal of inpatient treatment.^2,15