'Against Medical Advice' in the ED: Where We Are in 2009

By Edward Monico, MD, JD, Department of Surgery, Section of Emergency Medicine, Yale University School of Medicine.


In light of the existing health care crisis, patients who leave the emergency department (ED) against medical advice (AMA) will foreseeably represent an increasing population of emergency patients. Despite this, these patients have attracted little academic interest within the emergency medicine literature. Also, confusion persists among emergency medicine physicians regarding what safeguards to use to protect both patients and physicians from adverse medical and legal ramifications associated with this disposition. This article discusses the relevant issues surrounding AMA patients that exist to date, including terminology, the origin and source of physician obligation, a review of extant AMA literature, liability, outcome, and documentation.


No cogent discussion about AMA dispositions can proceed without first attempting to whittle down what is meant by the term "against medical advice." Health care providers and scholars invoke the phrase in practice and in the literature to describe an expanse of clinical encounters where a patient's decision causes a therapeutic or diagnostic goal to fall short of the provider's expectation. Both provider and scholar apply these three words like grammatical duct tape over an expanse of broken clinical scenarios in need of repair. Health care providers frequently use "AMA" even when the facts contradict the term's simplest conveyance. For instance, medical students, interns, and the occasional seasoned attending sees fit to describe patients who leave the ED waiting area without being seen as leaving "against medical advice" despite the fact that advice had never been rendered. On the other hand, academic scholars writing about AMA dispositions: limit the discussion to describe the timeliness of discharge,1 distinguish patients who leave AMA from patients who leave without being seen (LWBS) or complete their ED care,2 or simply never attempt to define the term at all.3

Legal and ethical obligations

The standard against which physicians are measured when confronted with a patient who wishes to leave AMA has developed apart from medicine's typical evolutionary course for standard of care. When it comes to AMA dispositions, the medical profession has departed from the typical self-regulation and scientific scrutiny characteristic of medical science. First, physicians have largely allowed other professions, political entities and legal standards to define physician obligations for AMA encounters. Second, to the extent that the medical community has defined any disclosure standard, it is based upon professional consensus and custom, not an objective analysis of patients' informational needs.4

The origin of the informed consent doctrine that underlies a patient's right to refuse treatment is found in the intentional tort of battery. Under this rubric, failure to obtain informed consent constituted an unlawful touching because the physician was limited by the consent given.5 In the 20th century, courts further defined and expanded informed consent doctrine to included negligence. The common thread running from tort to battery lies in the ancient Anglo-American prohibition of physical trespass and the fundamental right of every individual to determine what will be done with his or her body.6 The American Medical Association Code of Medical Ethics has implicitly espoused this view since at least 1981:

The patient's right of self-decision can be effectively exercised only if the patient possesses enough information to enable an intelligent choice. The patient should make his or her own determination on treatment. The physician's obligation is to present the medical facts accurately to the individual responsible for the patient's care for management in accordance with good medical practice...Social policy does not accept the paternalistic view that the physician may remain silent because divulgence might prompt the patient to forego needed therapy. Rational, informed patients should not be expected to act uniformly, even under similar circumstances, in agreeing to or refusing treatment.

The American College of Physicians and the American Society of Internal Medicine also endorse this view.7

Courts have recognized a fundamental common law right of a competent adult to refuse medical treatment.8 In some jurisdictions, the state's recognition of that right is particularly strong. For instance, in New York the common law right to refuse medical treatment has been codified in the legislature.9 Furthermore, the court in Fosmire declared the "fundamental common-law right is coextensive with the patient's liberty interest protected by the due process clause of our State Constitution."10 Generally, this right is subject only to the state's compelling and overriding interest to preserve life, prevent suicide, protect innocent third parties, and preserve the medical profession's ethical integrity.11

In addition to drawing from the common law and various state constitutions, courts have found patients' rights to refuse medical treatment to stem from the federal Constitution, as well. Although the U.S. Constitution contains no provisions expressly setting forth a constitutional right to refuse medial therapy, recent federal cases have found such a right implied within the more general constitutional protections embodied in the Fifth and 14th amendments. The derivation of a patient's constitutional right to refuse medical intervention arises from several fundamental guarantees which create a penumbral right to individual privacy and liberty that is no less important than the rights specifically articulated in the Constitution.12 For instance, the court in Satz v. Permlutter grounded the right of a terminally ill, elderly patient to refuse life-sustaining treatment in a constitutional right to privacy. The United States Supreme Court expressed similar sentiments, but chose to ground the right to refuse medical treatment in a 14th Amendment liberty interest rather than the right in privacy.13

Lastly, Congress created the Emergency Medical Treatment and Actives Labor Act (EMTALA) to provide adequate emergency medical services to the indigent and uninsured who seek emergency care. EMTALA sets out an obligation that physicians must adhere to when patients refuse to consent to treatment:

A hospital is deemed to meet the requirement of paragraph (1)(A) with respect to an individual if the hospital offers the individual the further medical examination and treatment described in that paragraph and informs the individual (or a person acting on the individual's behalf) of the risks and benefits to the individual of such examination and treatment, but the individual...refuses to consent to the examination and treatment.14

Patient and hospital characteristics

Most of the academic literature focusing on AMA dispositions isolates one frame of the overall AMA picture. For instance, some authors focus on patient characteristics, subliminally inferring there must be something about patients that want to make them leave AMA. Generally, these studies demonstrate that refusing medical care is a commonality among the socio-economically impoverished typically defined as those without insurance or covered by Medicaid. One study demonstrated how certain presenting symptoms such as nausea and vomiting (9.7%), abdominal pain (7.9%), and nonspecific chest pain (7.6%) might foreshadow an AMA disposition, but stopped short of suggesting that leaving AMA is part of the symptom complex.2 Studies evaluating AMA discharges among hospitalized patients seem to support the lower income category prevalence, and add African American race, men, and younger age to the mix.

Other studies focus on external factors that might contribute to the decision to leave AMA. Studies that focus on hospital factors generally conclude that location in large urban areas, hospitals with a greater proportion of minorities, and hospitals with the lowest Herfindahl indices (an economic indicator of market power) somehow lend themselves to AMA discharges.15 While overcrowding contributed to AMA discharges,16 teaching hospitals had fewer AMA discharges when compared against non-teaching hospitals.17 One study that remarked on the patient– physician dynamic revealed that one of the most common reasons for leaving AMA was dissatisfaction with care18 and lack of trust in the health care provider.19

Taken together, these studies seem to suggest that multiple factors have their hands on the wheel steering patients toward leaving AMA. Which factor or factors predominate on any given occasion may be the product of how environmental factors influence individual circumstance. Identifying how these factors play on each other and their role in the dynamic milieu of the ED may be the only way to truly understand AMA encounters.

Physician liability in AMA encounters

Physicians' angst regarding their liability arising out of AMA encounters may be well founded. These patients sue the physician and hospital nearly 10 times as often as the typical ED patient, yielding a rate of about one in 300 AMA cases versus the usual rate of one in 20,000–30,000 ED visits.20 Physician misconceptions about AMA encounters may be one factor that contributes to AMA litigation risk. For example, belief that providing patients leaving AMA with aftercare resources in the form of discharge instructions exposes physicians or institutions to greater liability is one such misconception. In fact, aftercare instructions may decrease AMA liability.21 The ability of a patient's signature on an AMA form to confer legal immunity to the physician is another common misconception. Physicians who discharge patients AMA enjoy no definitive legal protection from the consequences of their patients' choices.22 This is due, in part, to failure to adequately adhere to and document AMA protocols. In a 2008 presentation to the American College of Emergency Physicians (ACEP),20 Dr. Robert Bitterman stated the most common failures included:

• Failure to adequately document that the patient couldn't be found or left of his or her own free will prior to the medical screening examination;

• Failure to insist that the patient sign the AMA form, and instead discharged the patient without first documenting the patient's refusal to follow the recommended treatment plan;

• Failure to adequately inform the patient of the risks of leaving which are specific to the patient's chief complaint.

In general, a physician's liability may hinge on whether there was a duty to commit the patient to the hospital involuntarily or release him AMA. A second wave of rulings may turn on the sufficiency of the information exchange and how it is reflected in the medical record. Demonstrating in the medical record that the patient's ability to make an informed decision was done in a competent manner with sufficient information to make such a decision puts the physician in the best possible position to defend a claim of medical malpractice. This is done under the caveat that the patient is not a clear danger to himself or others, nor is unable to care for himself.

AMA outcomes

There is a paucity of outcome data dedicated to patients who leave the ED AMA. One reason for this is that outcome data require a return visit and re-evaluation — two events that do not seem to occur with AMA patients in a statistically meaningful way. The scarcity of return visits maybe due to patient belief that resolution of symptoms negates the need for further follow-up and that the decision to leave AMA will invoke derision from staff members.23 Also, attempts to track these patients have proved difficult.24 Other impediments to learning the fate of AMA patients include the fact that most of the AMA literature is comprised of studies that are either small,25 include single-site settings,26 or focus on specific medical subpopulations such as patients suffering from asthma.27

One study concluded that AMA patients have a higher number of rehospitalization rates and more severe symptoms at the time of discharge.28 A similar study demonstrated that during a 30-day follow-up period, patients who left AMA had significantly more emergent hospitalization and ED visit rates compared with control groups.29 What these studies suggest is that the prognosis of patients who leave emergency services AMA falls statistically between patients who consent to admission and those for whom admission is not recommended.30 What this means for non-statisticians is that patients with asthma, for instance, who left AMA were more likely to have an asthma relapse within 30 days resulting in higher revisits and readmission to the hospital when compared to other discharged patients.27

AMA documentation

Curiously, the liability risks AMA disposition represent to emergency physicians have not translated into a universal understanding of how to document the interaction between the physician and the patient leaving AMA. An amazing amount of ambiguity persists regarding what to tell patients and what to memorialize in the medical record. Inadequate documentation may actually increase the likelihood of proving liability against the provider when an invalid AMA is obtained and no care is provided. The sobering reality remains that even in the presence of a valid AMA, the question of liability may be more dependent on the expertise of legal counsel than on the facts. However, to put themselves in the best possible position to defend a claim of medical malpractice, physicians should have sufficient understanding of what contents constitute the medicolegal floor in AMA documentation. Generally, it should be said that most hospital AMA forms contemplate some or all of the following (or more) would be documented in the narrative space provided on the document. Simply signing the blank form, in my opinion, amounts to signing a blank check if a claim of malpractice follows.

A threshold question may be if the patient possesses the competence and capacity to make medical decisions concerning health care. Simply stating the patient "understood" may be difficult to substantiate in court. Instead, physicians should document the patient's ability to provide a lucid history, reasoning, and exam findings such as Glasgow coma scores that contribute to assertions of competence and capacity. The lack of mitigating factors such as alcohol and distracting injury should also be acknowledged.

Second, physicians must document that the patient was informed of the extent and limitation of the evaluation conducted up to the point the patient expressed the desire to leave AMA. For instance, documentation that a patient was informed that a negative electrocardiogram and cardiac enzymes does not conclusively eliminate the possibility of an acute coronary syndrome would be of paramount importance when chest pain patients refuse further evaluation for that symptom.

Documentation should also reflect that the patient and physician were on the same page during discussions of the presenting signs and symptoms and that the patient was made aware of the specific concerns the physician had regarding the presentation.

Demonstrating that the patient was made aware of the risks of not receiving treatment is the next factor. This would include, for example, death from chest pain that the physician thought might represent unstable angina. Other required documentation includes enumerating reasonably foreseeable complications, such as worsening asthma resulting in the need for prolonged hospitalization.

Alternatives, if they exist, should be discussed and documented. Prolonged observation in the ED or an observation center would be typical examples of alternatives to leaving AMA.

It should be explicitly stated that the patient left AMA as well as what specific care and treatment were refused. Simply writing the words "patient refused care" may be legally insufficient and construed as a conclusion without the contextual reference of what care was refused.31

Lastly, the AMA note should contain that patients were provided the opportunity to ask questions, timely follow-up was provided, and discharge instructions containing this information were given to the patient.


Patients who leave the ED AMA may represent an increasing liability risk to emergency medicine physicians in light of the current health care crisis and the burden that crisis places on emergency care. Acknowledging this population through increased academic interest, self preservation, or both, is needed to enable the medical community to better understand what these patients mean in terms of health care risk and liability potential.


1. Alfandew DJ. "I'm going home": Discharge against medical advice. Mayo Clin Proc 2009;84:255-260.

2. Ding R, Jung JJ, Kirsch TD, et al. Uncompleted emergency department care: Patients who leave against medical advice. Acad Emerg Med 2007;14:870-876.

3. Solomon R. Ethical issues in emergency medicine. Emerg Med Clin North Am 2006;24:733-737.

4. Canterbury, 464 F.2d at 786

5. Shugrue RE, Linstromberg K.The Practitioner's Guide to Informed Consent 24 Creighton L. Rev 882 (1991).

6. Schloendorff v. New York Hosp., 464 F.2d 772, 780 (N.Y.1914).

7. American College of Physicians and American Society of Internal Medicine, Ethics Manual. Ann Int Med (1998).

8. Fosmire v. Nicoleau, 551 N.E.2d 77, 80 (N.Y.1990).

9. 551 N.E.2d 77 (N.Y. 1990).

10. See id. (quoting Rivers v. Katz, 495 N.E.2d 337, 341 [N.Y. 1986]).

11. See id.

12. Griswold v. Connecticut, 381 U.S 489 (1965).

13. Cruzan v. Dir., Mo. Dep't of Health, 497 U.S. A261, 277-278 (1990).

14. EMTALA Statute: 42 USC 1395dd (b)(2)

15. Franks P, Meldrum MS, Fiscella K. Discharges against medical advice: Are race/ethnicity predictors? J Gen Intern Med 2006;21: 955-960.

16. Bernstein S, Aronsky D, Duseja R, et al. The effect of emergency department crowding on clinically oriented outcomes. Acad Emerg Med 2009;16:1-10.

17. Ibrahim S, Kwoh K, Krishnan E. Factors associated with patients who leave acute-care hospitals against medical advice. AJPH 2007;97:2204-2208.

18. Dubow D, Propp D, Narasimham K. Emergency department discharges against medical advice. J Emerg Med 1992;10:512-516.

19. Chappel JN, Schnoll SH. Physician attitudes: Effect on the treatment of chemically dependent patients JAMA 1977;237:2318-2319.

20. Bitterman RA. "Against medical advice–When should you take 'no' for an answer?" ACEP Scientific Assembly–Chicago, Ill October 30, 2008.

21. Strinko JM, Howard CA, Dchaeffer SL, et al. Reducing risk with telephone follow-up of patients who leave against medical advice or fail to complete an ED visit. J Emer Nurs 2000;26:223-232.

22. Devitt PJ, Devitt AC, Dewen M. An examination of whether discharging patients against medical advice protects physicians from malpractice charges. Psychiatr Ser 2000;52:899-902.

23. Jerrard DA, Chasm R. Patients leaving against medical advice from the emergency department: How many of them actually return to continue their evaluation and what are the reasons the others do not? Ann Emerg Med 2004;44:4 S130.

24. Pennycook AG, McNaughton G, Hogg F. Irregular discharge against medical advice from the accident and emergency department–A cause for concern. Arch Emerg Med 1998;9:230-238.

25. Hwang SW, Li J, Gupta R, et al. What happens to patients who leave hospital against medical advice. CMAJ 2003;68:417-420.

26. Weingart SN, Davis RB, Phillios RS. Patients discharged against medical advice from a general medicine service. J Gen Intern Med 1998;13:568-571.

27. Baptist AP, Warrier I, Arora R, et al. Hospitalized patients with asthma who leave against medical advice: characteristics, reasons, and outcomes. J Allergy Clin Immunol 2007;19:924-929.

28. Smith DB, Telles JL. Discharge against medical advice at regional acute care hospitals. Am J Public Health 1991;81:212-215.

29. Ding R, Jung JJ, Kirsch TD, et al. Uncomplicated emergency department care: Ppatients who leave against medical advice. Acad Emerg Med 2007;14:870-876.

30. Lee TH, Short LW, Brand DA, et al. Patients with acute chest pain who leave emergency departments against medical advice: Prevalence, clinical characteristics, and natural history. J Gen Intern Med 1988;3:21-24.

31. Miller S. Obtaining a valid AMA. JEMS 1996;21:54-55.