Special Report

Patients refusing care: Weigh risk vs. autonomy

by Jay C. Weaver, JD, EMT-P, Attorney, Private Practice; Paramedic, Boston Public Health Commission, Boston, MA

For emergency department (ED) personnel, the most troublesome patients often are those who don't want treatment at all.

Two percent of all people who seek care from an ED leave before optimal care can be rendered.1 A small number refuse care from the beginning, having been brought to the hospital against their will by law enforcement personnel or prehospital care providers. Regardless of the circumstances under which the patient arrives, a decision to leave the ED against medical advice (AMA) places ED personnel in an awkward predicament, forcing nurses and physicians to choose one of two options: They can allow the patient to leave, or they can restrain him and force him to accept treatment. Like most decisions made in the ED, this one must be made instantly.

Choosing unwisely can have profound legal implications. A physician or nurse who allows a patient to leave prematurely invites a claim of medical malpractice. Conversely, ED personnel who force unwanted treatment on a reluctant patient expose themselves to claims of assault, battery, and even wrongful imprisonment.2 Because an employer can be held liable for the on-duty actions of its employees, the hospital may be named as a defendant also.

Patient autonomy

Americans have long enjoyed a right to accept or reject medical care as they see fit. This right flows primarily from the common law, but the Supreme Court has found a constitutional basis for it as well. A handful of jurisdictions, such as Delaware, have granted a statutory right of refusal to their residents. Physicians and nurses cannot force treatment on a patient simply because they believe the patient will benefit from such care. This holds true even when the patient's life hangs in the balance. As the Massachusetts Supreme Judicial Court said in Harnish v. Children's Hospital: "Every competent adult has a right to forego treatment, or even cure, if it entails what for him are intolerable consequences or risks, however unwise his sense of values may be in the eyes of the medical profession."3

Not all patients have the right to refuse treatment, of course. Minors generally cannot refuse lifesaving care. Nor can individuals who lack the capacity to make rational decisions, or who pose a threat to themselves or others because of mental illness. The duty of ED personnel to provide treatment in the presence of a court order or judicial determination of incompetence is clear. Things become considerably more complex when a normally healthy patient simply decides to leave.

Patient consent and the 'informed refusal'

In general, treatment cannot begin until the patient has consented to it. To be valid, the consent must be informed … that is, the patient must possess sufficient knowledge about the nature of his condition, the risks and benefits of the proposed treatment, and the risks and benefits of available alternatives to permit a reasoned decision.4

Patients cannot expect their physicians to disclose everything they know about the treatment they intend to provide. Not only would most patients have difficulty comprehending such information, but the disclosure process itself would be too cumbersome to permit timely care. Today, some jurisdictions require practitioners to make "those disclosures which a reasonable medical practitioner would make under the same or similar circumstances."5 Others apply a patient-based standard, requiring the disclosure of facts that a reasonable patient would deem significant in deciding whether to accept the proposed treatment.6

Because children are thought to be incapable of making reasoned decisions, patients younger than 18 generally cannot give valid consent. Exceptions are made for "emancipated" minors … individuals younger than 18 who are married, pregnant, living independently, or serving in the military. Increasingly, courts also have recognized the right of "mature minors" to make autonomous health care decisions. Children as young as 8 have been permitted to give consent or refuse care after demonstrating an appreciation of the consequences of their decisions. In the setting of the ED, it is always better to seek consent from the parents or guardians of a minor than to rely on this doctrine, however.

In a true emergency, it is not always necessary to obtain consent. The law has long recognized that legal formalities must take a backseat to the preservation of life and function. In Jackovach v. Yocom, the Iowa Supreme Court refused to hold a surgeon liable for acting without parental consent in saving the life of a 17-year-old who had fallen beneath a train.7 Underlying this decision was concern that physicians would place lives at risk out of concern over liability. Consent to lifesaving treatment may therefore be presumed in situations where prolonged discussion would seriously jeopardize the patient's welfare, or where the patient cannot give effective consent because he is unconscious, disoriented, or sedated.

A medical emergency does not open the door to whatever treatment a practitioner deems advisable, however. To the contrary, in the absence of consent, a physician or nurse may provide only that care which is necessary to maintain the status quo. Similarly, consent to treatment may not be presumed if the patient previously objected to it. In Shine v. Vega, an ED physician incurred liability by sedating and performing endotracheal intubation on an asthmatic who had repeatedly objected to it, and who wanted to leave the ED.8 The defendant physician learned the hard way that a practitioner may not override a patient's decision simply because the patient's condition deteriorates.

Refusals of care, like consents to treatment, must be informed. ED personnel have an obligation to describe for a departing patient the nature and severity of the patient's condition; the nature, risks, and benefits of the proposed treatment; and the consequences of refusing that treatment. In Truman v. Thomas, the defendant physician, Dr. Claude Thomas, saw the plaintiff as a patient for more than five years. At each examination, he suggested a Pap smear, but the plaintiff always refused to have one performed. When the plaintiff developed cervical cancer, she sued Dr. Thomas, claiming that he had been negligent in failing to describe the significance of the test she had repeatedly declined. The California Supreme Court agreed, and upheld the trial court's finding of liability on the part of Dr. Thomas.9

When a patient indicates a desire to leave without treatment, ED personnel will be tempted to describe the potential consequences in the broadest possible terms. This may be the easiest approach, but it isn't necessarily the best. Warning a patient that "you might die without treatment" covers every eventuality, but it lacks the specificity needed to be effective. More than one court has held that such a warning is not likely to be taken seriously by patients, and therefore does not operate as a disclosure at all.10 The same holds true of boilerplate release forms, which talk of risks in general terms and do not put the patient on notice about the risks unique to his or her condition.11

A refusal cannot be informed if the patient's condition does not permit him to comprehend the information being provided. Before a patient is allowed to leave, his decision-making capacity must be assessed. Various tests have been devised to guide this process.12 The more serious a patient's condition is, the more carefully his decision-making capacity should be scrutinized.

Handling an AMA refusal of care properly

A risk of liability arises with every AMA refusal of care. This risk can be minimized by an ED and its staff, however, through the adoption of certain principles and procedures.

A physician should talk with the patient about the risks involved. In many EDs, nurses obtain the patient's signature on a release-of-liability form as part of the discharge process. There is nothing wrong with this practice. At some point prior to the patient's departure, however, a physician should personally describe the risks involved in the refusal of treatment. Ultimately, it is the physician who must decide whether to allow the patient to leave, and he cannot avoid responsibility by delegating to a resident or nurse the task of disclosing necessary information.

Obtain the patient's signature on a form specific to the patient's situation. By customizing the refusal-of-care form with details about the patient's condition, proposed treatment, and potential consequences of refusal, and by obtaining the patient's signature on this form afterward, the hospital goes a long way toward proving that treatment was not withheld, and that the patient was appropriately informed when he left AMA. If the patient refuses to sign a refusal-of-care form, document the circumstances thoroughly. Quote the patient liberally.

Do not allow patients to waive disclosure. A patient cannot make a knowing waiver of rights unless he knows what rights he is giving up. Patients who not only refuse treatment, but also refuse to participate in a discussion about the consequences, do not make an informed refusal.13 ED personnel must do everything possible to hold such a discussion, and if the patient absolutely refuses, the interaction should be documented in detail.

Communicate and document. When it becomes clear that a patient intends to leave AMA, the patient's primary care physician should be contacted. The patient should be informed of specific scenarios that should prompt immediate return to the ED, and some kind of followup, whether by telephone or return visit, should be planned. Every aspect of the interaction should be documented. In appropriate circumstances, the refusal-of-care form must expressly state that the patient declined a "medical screening examination" and "stabilizing treatment" otherwise, the hospital may face severe penalties under the Emergency Medical Treatment and Active Labor Act (EMTALA).

Do not appear vindictive. Behaving angrily in response to the patient's noncompliance has been shown to add to patient resentment, which in turn increases the likelihood of litigation. Reassure the patient that he is free to return to the ED.


1. Agency for Healthcare Research and Quality. 2006 National Healthcare Disparities Report. Rockville, MD: U.S. Department of Health and Human Services, Agency for Healthcare Research and Quality; December 2006. AHRQ Pub. No. 07-0012.

2. Schloendorff v. Society of New York Hosp., 105 N.E. 92, 93 (N.Y. 1914).

3. Harnish v. Children's Hospital, 439 N.E.2d 240 (Mass. 1982).

4. Salgo v. Stanford Univ. Bd. of Trustees, 317 P.2d 170 (Cal. App. Ct. 1957).

5. Natanson v. Kline, 350 P.2d 1093 (Kan. 1960).

6. Canterbury v. Spence, 464 F.2d 772 (D.C. Cir. 1972); Cobbs v. Grant, 502 P.2d 1 (Cal. 1972).

7. 237 N.W. 444 (Iowa 1931).

8. 709 N.E.2d 58 (Mass. 1999).

9. 611 P.2d 902 (Calif. 1980).

10. See Wells v. Van Nort, 125 N.E. 910 (Ohio 1919).

11. See Bourgeois v. McDonald, 622 So.2d 684 (La. Ct. App. 1993).

12. Jones RC, Holden T. A guide to assessing decision-making capacity. Cleve Clin J Med 2004;71:971-975.

13. See, e.g., Mark Strasser. The right to remain uninformed, 11. J Med. & Philosophy 265 (1986).