Sorry, you need to enable JavaScript to visit this website.
Skip to main content

Standard Deviations

James E. Sabin, MD | December 1, 2009
View more articles from the same authors.

Case Objectives

  • Understand the safety risks associated with patients being discharged against medical advice (AMA).
  • Recognize safety risks associated with being part of a stigmatized group.
  • Explain how the concept of "cultural competence" should extend to include the homeless intravenous drug user population.
  • Appreciate the role of ethics and psychiatric consultations in patients whose refusal to cooperate with treatment is perplexing.

Case & Commentary: Part 1


A 45-year-old man with an active history of intravenous (IV) drug use was admitted to the medical service with fever and low back pain. He was noted to have a leukocytosis and an elevated erythrocyte sedimentation rate (ESR), and a subsequent MRI showed an epidural abscess with surrounding osteomyelitis. After having his abscess successfully drained, the patient required a 6-week course of IV antibiotics for methicillin-resistant Staphylococcus aureus (MRSA). Given that he was homeless, uninsured, and actively abusing IV drugs, there was no appropriate care setting where he could complete his required therapy. Therefore, he remained hospitalized while awaiting Medicaid enrollment. After the first 2 weeks, he became increasingly abusive to the nursing staff, was found smoking on the unit, and twice threatened to leave against medical advice (AMA). In efforts to make the best of the situation, the physicians caring for the patient elected to discharge him on an oral antibiotic regimen with follow-up in a free health clinic.

This case highlights an unfortunately common problem with patients who are homeless and uninsured, and also have substance abuse problems, mental health conditions, or both: the diagnosis and desirable treatment may be clear, but actually carrying out the treatment can seem next to impossible.

The most important "error" in this case—a societal rather than a clinical one—occurred before the patient came to the hospital. The patient is uninsured. As a result, the hospital team had to care for him with the equivalent of one arm tied behind their backs. Had the patient been insured, it might have been possible to refer him to a setting other than the hospital that would have been more acceptable to him and would have allowed continuation of IV antibiotic therapy. A recent study suggested that as many as 45,000 deaths per year may be associated with lack of insurance.(1) This patient is at high risk for being part of that group.

Defining the clinical standard of care for a homeless IV drug user is complex. Anglo-American law requires us to comport ourselves with "due care" so as not to cause avoidable harm to others. Historically, physicians charged with negligence were judged by the standard of "customary practice"—how other physicians, locally or nationally, practiced. Over time, however, the legal standard shifted from actual practice patterns to "reasonable" practice patterns, to allow for the possibility that customary practice could be too slow in adopting evidence-based guidelines.(2)

Although the American Medical Association cautiously defines the clinical standard of care we owe to our patients as "competent care," presumably to protect physicians from being judged by overly high expectations, most health professionals (and patients) aspire to go beyond mere "competence" to an ideal of "excellence."(3) In this case, the hospital team provided excellent diagnosis and treatment planning. If physicians had failed to urge the patient to receive 6 weeks of IV antibiotics and explain the risks of not completing the course of treatment, they would not be meeting legal or clinical standards of care and would be vulnerable to a malpractice claim.

The first 2 weeks of treatment apparently went reasonably smoothly, but at that point a new clinical issue emerged—how best to deal with an abusive patient who threatens to leave against medical advice. This problem could have been anticipated and planned for at the point of admission. Discharges against medical advice (AMA) are not uncommon. A study of 3 million acute admissions showed a baseline AMA discharge rate of 1.44%.(4) But for patients with mental illness or substance abuse (or both), the rate is much higher, ranging up to an astounding 51%.(5) Among 1056 HIV-positive patients with a history of IV drug use, 25% signed out AMA.(6) Not surprisingly, patients discharged AMA do worse as a group. In one study, 21% were readmitted within 2 weeks compared to 3% in a matched control group.(7) Hospitals should develop strategies to prevent AMA discharges just as they develop strategies to prevent falls and hospital-acquired infections.

From the perspective of patient safety, this case emphasizes how important it is to focus on our competencies in dealing with culturally distinct patients and groups. A homeless IV drug user may be more "foreign" for hospital staff than an educated professional from Asia or Africa would be. Homeless IV drug users are a highly stigmatized group. It is common for clinicians to see a patient like the man in this case as an extreme "other." The staff wants to use IV medications to treat an illness. The patient uses IV "medications" to sustain an addiction. The staff wants the hospital to be a health-promoting environment. The patient smokes on the ward. The staff wants to treat patients with respect and empathy. The patient treats them with abuse. When differences like these create barriers to empathy and rapport, they also create risks to safety.

In its ethical meaning, "standard of care" refers to the treatment that should be provided. Over the past 40 years, the United States has come to a widely shared consensus that informed consent determines what should be provided. In this case, a 6-week course of IV antibiotics is clinically indicated, but if the patient is competent to decide whether to accept the recommendation and his physicians explained the risks of not following it in ways he understood, the ethical standard of care would require respect for his decision.

Decision-making capacity is defined by four criteria: (i) ability to communicate a choice, (ii) ability to understand the relevant information, (iii) ability to appreciate the situation and its consequences, and (iv) ability to reason about treatment options.(8) Although the case report does not suggest lack of competence, the significant potential harm from nonadherence to the recommended treatment, the frequency of psychotic disorders among the homeless (in one systematic review, the pooled prevalence of psychotic disorders was 12.7% [9]), and the potential for subtle cognitive deficits (10) call for an especially careful assessment of competence by a skilled clinician.

Research and clinical experience suggest that when homeless people reject services they do so "on the basis of a desire to be independent, a lack of active participation in services, poor therapeutic relationships, lack of provider cultural competence, and side effects from medication."(11) For a man accustomed to independent life on the streets, a hospital can be highly stressful. He is not accustomed to following orders and obeying rules in a socially appropriate manner. Outreach workers spend months trying to develop trusting relationships with homeless clients and even with that effort may not succeed. Once this patient begins to feel better, the personality factors that led to his homelessness and drug addiction are likely to reassert themselves.

In a case like this, it is as important to bring in a collaborator skilled in dealing with homeless IV drug users as it is to bring in an interpreter for patients who do not speak English. The collaborator could be a clinician, someone the patient knows and trusts (such as a shelter worker or priest), or even a homeless person he looks up to. In a case from Massachusetts involving a homeless man with alcoholism and advanced cancer, a priest became the crucial go-between for the hospital staff.(12) Sometimes adjustments in the hospital regimen can help, such as having a volunteer take him outside to smoke on a regular basis. The case report does not tell us how the patient's substance abuse condition was handled, but a Canadian hospital reported a 50% reduction in AMA departures when HIV-positive patients with a history of IV drug use were treated with methadone.(6)

The case report tells us that "the physicians caring for the patient elected to discharge him on an oral antibiotic regimen with follow-up in a free health clinic." An ethics consultation should be considered before making a decision like this, since the clinical consequences of less-than-optimal therapy might be high. A third party might be able to help the staff identify impediments to creating a better outcome. Abusive patients can elicit negative staff attitudes that could make the prospect of discharging the patient appealing. And financial concerns on the part of the hospital about a long hospital stay for an uninsured patient could also skew clinical judgment. Ethics consultants are typically skilled at dealing with thorny situations like this one and offer expertise and guidance to providers in managing such situations.

Case & Commentary: Part 2

The patient was readmitted to the same hospital 3 weeks later with a recurrent epidural abscess and worsened osteomyelitis. He reported taking "most of his antibiotics" but never made it to a clinic for scheduled follow-up care. He required a more extensive surgical debridement, and he also developed MRSA endocarditis. The patient became quite deconditioned postoperatively, but, after nearly 3 weeks of inpatient care, he threatened to leave AMA once again. Although he was ambulating at that point, he still required additional IV antibiotics and lacked any viable option for institutional nonhospital care. The providers "negotiated" with the patient to remain hospitalized, but he ultimately left AMA after an additional week. He was subsequently lost to follow-up.

Sadly, but not surprisingly, the patient's condition worsened. Ideally, the hospital would have learned from the first AMA departure as a sentinel event and would be better prepared to deal with the patient (and others like him) when he returned. It would be important to try to learn—from the patient's perspective—what makes it difficult to stay in the hospital and which of his values are satisfied by getting back on the street. This kind of conversation can be challenging for staff, for whom goals like "doing what I want and not what I'm told to do" or "getting a fix" may seem unworthy and hard to understand.(13) And hospitals that serve substantial numbers of homeless patients should build relationships with community agencies that serve the homeless population to provide partners for creating flexible responses to patients like this one. Finally, a system that flags patients' charts after they have left the hospital AMA may alert providers on future admissions about opportunities for early intervention and the risk for a repeated occurrence.

It would be important to use ethics consultation, psychiatric consultation, or both, to ensure that any remediable impediments to adherence have been identified and that culturally appropriate approaches are identified. Are there previously unrecognized grounds for considering involuntary commitment? Have we understood the patient's real agenda? Are there people in the patient's life who can be brought into the treatment process? Are there ways in which the hospital and staff are unwittingly contributing to the patient's reluctance to stay?

Physicians and nurses may be confused and deeply troubled when clinical, ethical, and legal standards point in different directions, as they do in this case. Allowing the patient to interrupt his treatment and leave the hospital may feel like a violation of Hippocratic precepts. The hospital team is required to provide safe care, but the patient is not obliged to treat himself in a safe manner. The ethical standard of care requires that a competent patient's informed refusal should be honored even if the patient's choice does not meet clinical standards. If all that can reasonably be done to create a therapeutic alliance has been done, and if consultation confirms that the patient is competent to make the decision to leave the hospital AMA, the appropriate responses for the clinical team are sadness and determination to learn as much as possible from the case, not shame or guilt.

Take-Home Points

  • Patients from stigmatized groups, like homeless IV drug users, require culturally competent care as much as patients from foreign cultures do.
  • Homeless IV drug users are predictably at high risk for leaving the hospital AMA. Hospitals should plan in advance how to deal with that risk.
  • Patients whom staff find difficult to like, or care for, are vulnerable to safety risks created by reduced empathy and rapport.
  • Ethics consultation, psychiatric consultation, or both can be useful for patients with whom it is difficult to establish a therapeutic alliance.
  • Patients who sign out AMA are at risk for worse outcomes.


James E. Sabin, MD Director, Ethics Program, Harvard Pilgrim Health Care Clinical Professor, Departments of Population Medicine and Psychiatry, Harvard Medical School

Faculty Disclosure: Dr. Sabin has declared that neither he, nor any immediate member of his family, has a financial arrangement or other relationship with the manufacturers of any commercial products discussed in this continuing medical education activity. In addition, his commentary does not include information regarding investigational or off-label use of pharmaceutical products or medical devices.


1. Wilper AP, Woolhandler S, Lasser KE, McCormick D, Bor DH, Himmelstein DU. Health insurance and mortality in US adults. Am J Public Health. 2009;99:2289-2295. [go to PubMed]

2. Rich BA. Medical custom and medical ethics: rethinking the standard of care. Camb Q Healthc Ethics. 2005;14:27-39. [go to PubMed]

3. Sulmasy DM. What is conscience and why is respect for it so important? Theor Med Bioeth. 2008;29:135-149. [Available at]

4. Ibrahim SA, Kwoh CK, Krishnan E. Factors associated with patients who leave acute-care hospitals against medical advice. Am J Public Health. 2007;97:2204-2208. [go to PubMed]

5. Brook M, Hilty DM, Liu W, Hu R, Frye MA. Discharge against medical advice from inpatient psychiatric treatment: a literature review. Psychiatr Serv. 2006;57:1192-1198. [go to PubMed]

6. Chan ACH, Palepu A, Guh DP, Sun H, Schechter MT, O'Shaughnessy MV, Anis AH. HIV-positive injection drug users who leave the hospital against medical advice: the mitigating role of methadone and social support. J Acquir Immune Defic Syndr. 2004;35:56-59. [go to PubMed]

7. Hwang SW, Li J, Gupta R, Chien V, Martin RE. What happens to patients who leave hospital against medical advice? CMAJ. 2003;168:417-420. [go to PubMed]

8. Appelbaum PS. Assessment of patients' competence to consent to treatment. N Engl J Med. 2007;357:1834-1840. [go to PubMed]

9. Fazel S, Khosla V, Doll H, Geddes J. The prevalence of mental disorders among the homeless in western countries: systematic review and meta-regression analysis. PLoS Med. 2008;5:e225. [go to PubMed]

10. Burra TA, Stergiopoulos V, Rourke SB. A systematic review of cognitive deficits in homeless adults: implications for service delivery. Can J Psychiatry. 2009;54:123-133.[go to PubMed]

11. Stanhope V, Henwood BF, Padgett DK. Understanding service disengagement from the perspective of case managers. Psychiatr Serv. 2009;60:459-464.[Available at]

12. Penson RT, Fergus LA, Haston RJ, et al. The Kenneth B. Schwartz Center at Massachusetts General Hospital hematology-oncology department: hope for the homeless. Oncologist. 2003;8:488-495. [go to PubMed]

13. Berger JT. Discharge against medical advice: ethical considerations and professional obligations. J Hosp Med. 2008;3:403-408. [go to PubMed]

This project was funded under contract number 75Q80119C00004 from the Agency for Healthcare Research and Quality (AHRQ), U.S. Department of Health and Human Services. The authors are solely responsible for this report’s contents, findings, and conclusions, which do not necessarily represent the views of AHRQ. Readers should not interpret any statement in this report as an official position of AHRQ or of the U.S. Department of Health and Human Services. None of the authors has any affiliation or financial involvement that conflicts with the material presented in this report. View AHRQ Disclaimers
Related Resources From the Same Author(s)
Related Resources