Reducing risk when patients leave against medical advice
By David L. Freedman, MD, JD, FAAEM, Attorney, Miller, Canfield, Paddock & Stone PLC, Ann Arbor, MI
Patients leaving against medical advice (AMA) create both medical and legal dangers and represent one of the more significant risk management problems for emergency departments (EDs). It is all too easy, when faced with the uncooperative and confrontational patient during a busy shift in the ED, to simply tell the nurse to "sign him or her out AMA." After all, why should you take time away from patients who need and want your attention to beg an uncooperative patient to stay, especially when you think the patient has a minor problem? Remember, though, that in the AMA situation, your conclusion that the patient’s problem is "minor" is likely based upon an incomplete evaluation.
The notion that a patient’s signature on a preprinted hospital AMA form automatically relieves the physician of responsibility in whole, or even in part, for a subsequent adverse outcome after a patient leaves AMA is dangerous at best. True enough, if a patient who is competent leaves AMA after a full disclosure of the possible risks of leaving, the physician should be protected from liability. However, the italicized words "competent" and "full disclosure" are key and often subject to dispute. Even though you thought the patient was competent and you had made a full disclosure, you might still find your fate in the hands of a jury who will decide whether the patient really was sufficiently "competent," and whether your discussion of the potential risks was complete enough.
A key strategy in addressing the AMA problem is to hone your human relations skills such that you are able to talk most patients who want to leave AMA out of leaving. In doing so, enlist any available help from other hospital personnel (nurses, social workers, clergy, etc.) and the patient’s family, and use reasonable coercion when necessary. No one will be 100% successful, but your success rate should approach that number. At times, your efforts will fail, and you will have no alternative but to allow the patient to leave AMA or restrain the patient. Restraints should, of course, only be used in compliance with applicable law and standards and guidelines (e.g., standards from the Joint Commission on Accreditation of Healthcare Organizations in Oakbrook Terrace, IL).
This article presents a number of reported cases involving patients leaving AMA, with a focus on what information must be given to a patient before allowing an AMA discharge, when a patient is sufficiently competent to make the decision to leave AMA, and what is proper documentation of AMA discharges. The cases are followed by a compilation of risk management suggestions to reduce your potential liability when patients want to leave AMA.
Documentation
The following cases illustrate the approaches of two jurisdictions with respect to required documentation in AMA discharges. Notice that there is a significant difference in these approaches.
Case No. 1: Joan B. Dick v. Springhill Hospitals Inc. and Dr. C. Lawrence Ennis.1
In this case, the Supreme Court of Alabama addressed what evidence a plaintiff must introduce in a medical malpractice case to counter a physician’s defense that the patient left AMA and, therefore, the physician should not be liable for the patient’s subsequent death. According to the court, the sole issue in the case was "whether . . . Dr. Ennis [the defendant emergency physician] advised Mr. Dick [the patient] that he should be hospitalized and that his refusal to be admitted to the hospital could have fatal consequences."2
On Feb. 26, 1984, William J. Dick Jr. was admitted to Springhill Memorial Hospital ("Springhill") in Mobile, AL, for injuries he received in an automobile accident. He was treated for various injuries that included a fractured right leg and a concussion. He remained in the hospital until March 9, 1984, when he returned home. On March 10, the day following Mr. Dick’s release from the hospital, he was returned to the Springhill ED by his wife and children with a complaint of pain in his left leg.
Mr. Dick was examined in the ED by Dr. C. Lawrence Ennis, who diagnosed "early phlebitis of the lower left leg."3 Dr. Ennis prescribed an analgesic and hot compresses to be applied to the leg. He also instructed Mr. Dick to contact his orthopedist within the next 2-3 days. Mr. Dick left the hospital that evening without signing the hospital’s AMA Form. Mr. Dick collapsed the following evening and was returned to the ED by ambulance. Unfortunately, Mr. Dick had suffered a cardiopulmonary arrest and expired. The cause of death was determined to have been due to a pulmonary embolus.
At this point, the physician’s and family’s versions of the facts diverge. Mr. Dick’s wife brought suit against the hospital and Dr. Ennis claiming that Dr. Ennis had "failed adequately to inform Mr. Dick of the serious nature of his medical condition and of his need to be hospitalized for immediate treatment."4 She claimed that, had Mr. Dick been properly informed, he would not have left the hospital against the advice of the physician.
Dr. Ennis’ version of events was as follows: When Mr. Dick first arrived at the hospital, he was "smiling, friendly, and cooperative, and was not in acute distress."5 He complained primarily of a "sore left leg," and he said that "he thought the rigorous physical therapy he had received at the hospital caused the pain in his leg."6 Dr. Ennis performed a complete physical examination and discovered no swelling, redness, or heat in the left leg. There were no masses in the left calf, although the calf was tender to deep palpation. Mr. Dick denied chest pain and shortness of breath.
Dr. Ennis testified that he told Mr. Dick that, "although there were no definite physical symptoms, [he] might be suffering from phlebitis."7 Furthermore, Dr. Ennis testified that he informed Mr. Dick that phlebitis could result in a pulmonary embolus that could be fatal. Dr. Ennis also testified that he recommended hospitalization to Mr. Dick, although this recommendation was made "before Mrs. Dick entered the examining area."8 According to Dr. Ennis, Mr. Dick "appeared to understand this information, but indicated he did not want to be admitted to the hospital."9
Dr. Ennis further testified that he "briefly explained" the situation to Mrs. Dick following his earlier discussion with Mr. Dick.10 According to Dr. Ennis, Mrs. Dick expressed her concerns to Mr. Dick regarding his decision to go home, but he "remained adamant in his refusal to be admitted to the hospital."11 Dr. Ennis then telephoned Dr. Hall, Mr. Dick’s personal physician, who advised the prescription of an analgesic and hot compresses. Unfortunately, Dr. Ennis did not document on Mr. Dick’s medical record that he had recommended hospitalization or that he had requested Mr. Dick to sign the hospital’s AMA form stating that he was leaving the hospital against Dr. Ennis’ advice.
Mrs. Dick testified at her deposition that "after examining her husband, Dr. Ennis told them that Mr. Dick had a problem, perhaps phlebitis, but they were lucky because it was below the calf."12 This statement, if true, would suggest that Dr. Ennis would not have recommended hospitalization. Mrs. Dick testified that "she asked Dr. Ennis to keep her husband at the hospital, and Dr. Ennis responded that Mr. Dick did not want to stay at the hospital."13 It was her opinion that, had her husband been told that phlebitis was serious, as Dr. Ennis claimed to have done, her husband would not have left the hospital. Mrs. Dick denied being informed that her husband’s condition was serious or life-threatening and that, had she been so informed, she would not have taken her husband from the hospital. Mrs. Dick conceded that she was not present during Dr. Ennis’ examination of her husband.
Mr. Dick’s daughter, who remained with her mother in the waiting room, testified that Dr. Ennis did not tell them that Mr. Dick could be at risk for developing a pulmonary embolus that could be fatal. She also testified that she did not believe that Dr. Ennis could have told her father that he should be hospitalized because her father would have told this to the family and did not.
The parties were in sharp disagreement as to whether Dr. Ennis had advised hospitalization for Mr. Dick and, if he had, whether he had fully informed Mr. Dick of the risks of refusing hospitalization before Mr. Dick left the hospital. Remember that Mr. Dick did not sign the AMA form. The court noted that Mr. Dick and his wife did sign another form that was located on the same page as the AMA form. This could reasonably be interpreted as evidence that they were probably not asked to sign the AMA form and, therefore, Mr. Dick had not left the hospital AMA.
The lack of a signature on the AMA form proved not to be of significance. Dr. Ennis testified that he only asked patients to sign the AMA form in "obvious cases," and this was not an obvious case of thrombophlebitis.14 In addition, the plaintiff’s expert testified that "he would find no fault if Dr. Ennis had in fact informed Mr. Dick of the risks of refusing hospitalization, but simply had failed to note it in Mr. Dick’s record."15
The court upheld the trial court’s summary judgment for the defendants, holding that Mrs. Dick had failed to carry her burden "to put forth any facts to show that Dr. Ennis did not inform Mr. Dick fully."16 This was largely because "Mrs. Dick and her daughter conceded that Dr. Ennis was alone with Mr. Dick during the time Dr. Ennis stated he conveyed this information to Mr. Dick," and "Mrs. Dick and her daughter also admitted that they have no knowledge of the substance of the conversations Dr. Ennis had with Mr. Dick while the two were alone."17
Commentary: This very "physician-friendly" ruling placed an insurmountable burden on the plaintiff in this case. While I would agree that a signature on an AMA form is not the most important part of the AMA process, I would encourage all physicians to obtain a patient’s signature on something that clearly indicates that the patient is leaving AMA. That signature could be on the AMA form or the written discharge instructions, or both. I like the use of the discharge instructions because, rather than relying on generic AMA form language, the physician can clearly state the specific risks of leaving AMA, including the risk of death. Despite the outcome of that case, I think it is absolutely crucial to involve the family when a patient wants to leave AMA. First, the family may be able to convince the patient to stay. Second, the family will have a much more difficult time testifying that the physician’s explanation of the risks was insufficient.
Remember also that this case represents the opinion of one state’s supreme court (Alabama), and we do not all practice in that state. Before getting too comfortable about AMA discharges and becoming lackadaisical in your AMA procedures, consider the following Arkansas case.
Case No. 2: Raymond Thomas v. Dr. Leslie Sessions, et al.18
In this Arkansas case, the issue was whether the physician, Dr. Sessions, had advised admission and, if the patient left AMA, whether he was properly informed by Dr. Sessions prior to making his decision to leave. The facts of the case were summarized by the court as follows:
In the late afternoon of Oct. 24, 1984, Misty Sturgis called on her neighbor, Mr. William E. Thomas. She found him in acute distress: sweating profusely, weak and trembling, extremely flushed, experiencing nausea, chest pain, and numbness in his left arm. At her insistence, Thomas agreed to go to the Jefferson Regional Medical Center [in Pine Bluff], and Ms. Sturgis drove him there, arriving around 5 o’ clock. Ms. Sturgis asked the emergency room personnel to do an EKG. Another friend, Ms. Bernadette Allen, arrived, and the two stayed with Thomas waiting for the doctor to see him. Ms. Allen testified to Thomas’ chest pain, as well as generalized pain. After some two hours, Thomas left the emergency room, though whether he left of his own volition or was refused admission is a disputed question. Thomas collapsed later that evening, was returned to the emergency room at 9:43 p.m., and expired shortly thereafter. Death was attributed to myocardial infarction [MI].19
In the subsequent malpractice suit, it was claimed that "Drs. [Carl] Bell [an admitting physician] and Sessions had conferred by telephone and concluded, without taking a proper history or performing proper tests and examinations, that Thomas was dehydrated due to alcoholism, resulting in a misdiagnosis and abandonment in that Thomas was refused admittance to the hospital."20 The doctors’ version of the facts was that Dr. Sessions, the emergency physician, and Dr. Bell, the prospective admitting physician who had treated Mr. Thomas four years earlier, conferred by telephone and concurred in recommending that Mr. Thomas be admitted to Jefferson Regional Medical Center ("Jefferson"), or sent to Riverview in Little Rock (apparently a substance abuse treatment facility) for a 28-day program of detoxification and rehabilitation. The physicians contended that Mr. Thomas refused both and left the hospital AMA.
The plaintiff disagreed and alleged that Mr. Thomas was refused admittance to Jefferson and was, therefore, abandoned. In the trial court, Jefferson and Dr. Bell were dismissed by summary judgment. The case went to trial against Dr. Sessions, and a jury returned a verdict in her favor. The plaintiff appealed and contended there were material issues of fact as to Dr. Bell that should have precluded summary judgment and trial errors with respect to Dr. Sessions that merited reversal of the verdict.
The Supreme Court of Arkansas noted that, while "[t]he two doctors were unequivocal in their contention that Mr. Thomas [had] refused admittance to Jefferson," that was not an undisputed fact.21 That is, there was credible evidence that Mr. Thomas did not leave Jefferson AMA and had been refused admittance. There were two key pieces of evidence to support the plaintiff’s claim. First, Rebecca Amos, a registered nurse on duty in the ED at the time Mr. Thomas was initially seen, was unable to support the defendants’ contention that Mr. Thomas had refused admission to Jefferson. She testified that she had no recollection of Dr. Bell saying anything about admitting Mr. Thomas to Jefferson. Rather, Dr. Bell had only discussed a transfer to Riverview for admission. This supported the plaintiff’s claim that Dr. Bell had never offered admission to Jefferson. Second, and more importantly, the court noted that "the records of the emergency room do not reflect that Thomas refused hospitalization at Jefferson Regional, only that he refuses transfer’ [a reference to Riverview]."22
The court noted that whether Mr. Thomas had refused transfer to Riverview (the alcohol rehabilitation facility) was irrelevant, as Mr. Thomas was suffering from MI, not dehydration from alcohol abuse. The court then observed that the question was not simply whether Mr. Thomas had refused admission to Jefferson; rather, assuming Mr. Thomas had refused admission to Jefferson, whether "his decision was based on an informed understanding of his condition."23 Just as a person must be informed to consent to treatment, a person must be reasonably informed to refuse treatment.
The plaintiff was prepared to prove that, contrary to the version of facts presented by the physicians, Mr. Thomas had been refused admission to Jefferson. This was based upon the testimony of his two friends. Ms. Sturgis testified that "she was told by emergency room personnel to take Mr. Thomas home, there was nothing the hospital could do for him.’"24 Ms. Allen testified that she was told by a doctor at Jefferson that "Mr. Thomas had been drinking or was drunk, that he had an alcohol problem, and needed to be taken to Little Rock because there was nothing they could do for him.’"25 Remember, the testimony of the nurse tended to support the plaintiff’s version, not the physicians.
In this case, a diagnosis of alcohol-related dehydration was made without the benefit of a blood alcohol level. In retrospect, there was little, if any, support for this diagnosis since the patient returned some two hours after discharge at which time a blood alcohol test was done and showed a level of zero. Also, he died of an MI, symptoms of which were, in retrospect, present at the time of his earlier visit.
Commentary: The court did not say, but it is possible (perhaps probable) that the emergency physician had seen Mr. Thomas before and knew that he had an alcohol abuse problem. We also can guess that Mr. Thomas might have presented in the past (perhaps with some regularity) with symptoms somewhat similar to those he had on his March 13 visit and, on those previous occasions, he was never found to have any significant problem, other than alcohol abuse. We have made the point repeatedly in ED Legal Letter that each time a patient presents to the ED, whether it is a repeat visit or the patient is a "frequent flier," it is imperative that the emergency physician perform a full and complete evaluation of the patient, as determined by his or her present complaint. Reliance on a previous assessment by you or another physician, or allowing yourself to be influenced by your previous contacts with or knowledge of the patient, invites disaster. The bottom line is that the fact that a patient may have been seen 20 times in a row for chest pain related to alcoholic gastritis and reflux, does not guarantee that the next visit for chest pain will be because of gastritis, not acute MI.
The medical record also did not support the physicians’ version. According to the court, the "[r]ecords of the emergency room make no mention of Thomas refusing admission at Jefferson Regional, only that he refused transfer to Riverview."26 Finally, the plaintiff’s expert, Dr. Wayne Smith, testified that "the records generated on the evening of the 24th and at the hospital and Dr. Bell’s office the following morning belie the contention that Mr. Thomas was offered admission at Jefferson Regional Medical Center.’"27 In short, except for the testimony of the defendant physicians, who obviously were interested parties, the evidence all tended to support the plaintiff’s contention that Mr. Thomas had been refused admission at Jefferson.
The court reversed the judgments for the physicians and remanded the case to the trial court for a new trial.
Case No. 3: M.L. Drummond and Sue Drummond v. Richard E. Buckley, MD.28
Whether the plaintiff, M.L. Drummond, left the ED AMA was the primary issue in this case. In addition, the case also had an interesting side issue regarding the requirement that a plaintiff must support his or her claim that the standard of care was breached through the use of expert testimony.
On March 4, 1985, Mr. Drummond had surgery performed on his lumbar spine (presumably a laminectomy) because of a herniated disc. The surgery was performed at Memorial Hospital in Gulfport, MS, by Dr. Harry Danielson. Immediately following the surgery, Dr. Danielson left on a trip to California, and Dr. Richard Buckley attended to Mr. Drummond in Dr. Danielson’s absence. Mr. Drummond was discharged from the hospital on March 10, 1985.
On the morning of March 13, 1985, Mr. Drummond called Dr. Buckley complaining of "pain and swelling in the area of the wound, drainage from the wound, and a fever the night before."29 Mr. Drummond had an appointment that day to see the hospital administrator (for reasons unknown), and Dr. Buckley said he would meet him in the ED when he arrived for his meeting. At approximately 11:40 that morning, Dr. Buckley examined Mr. Drummond and found the surgical wound to be "red, tender, and swollen, and concluded that the wound was infected."30 While those facts were uncontested, what occurred next was a matter of dispute between the parties and was the primary issue in the case: Did Mr. Drummond leave the hospital AMA?
Dr. Buckley testified that "he told Drummond that he should enter the hospital so that the infection could be properly treated."31 According to Dr. Buckley’s version of the facts, Mr. Drummond refused and stated that "he had an appointment to see the administrator, that he did not have time to come into the hospital at this time, that he would see about it when he got his business with the administrator taken care of. He commented that he might come back that afternoon after he had seen the administrator, and he told me that he surely would see Dr. Danielson the next day."32
Mr. Drummond’s recollection of the discussion was different. He testified that, while "he does not remember all of the details of that conversation with Dr. Buckley, because he was scared at the time . . . had he been told to enter the hospital, he probably would have done it, because he was in a lot of pain at the time."33 The court interpreted Mr. Drummond’s statement that he "probably would" have stayed as he "would" have stayed. As the court put it, the primary issue in the case was "whether the patient or the physician is to be believed concerning [this] conversation between the two."34
Given those essentially polar opposite versions of the circumstances of Mr. Drummond’s discharge from the hospital, a signed AMA form or discharge instructions documenting the fact that the patient had been advised to stay in the hospital and was leaving AMA would have been crucial supportive evidence of Dr. Buckley’s version. Either of those documents would have been key evidence for the defense; unfortunately, they did not exist. In fact, there was absolutely no documentation that the patient had left AMA.
Mr. Drummond’s ED record included Dr. Buckley’s notations that "Drummond had had recent back surgery and came to the emergency room complaining of drainage of the surgical wound, swelling in the area of the surgery, and increased pain."35 There was, unfortunately, no documentation that admission was advised or that Mr. Drummond left against Dr. Buckley’s advice. The only recorded recommendations made by Dr. Buckley were "that the patient should engage in Phisohex soaks, take the antibiotic ampicillin, and see Dr. Danielson the next day."36
Dr. Buckley explained the lack of documentation as follows: "My normal practice is to write down what is done, which is what the patient allowed me to do. Had I written in there, recommended admission and he refuses, it would certainly have tainted his medical record and that is not my normal practice to write AMA or things of that nature on the chart . . .. As for when he went to see another physician in the emergency room again and they saw he had an AMA, it would certainly have an adverse effect on his medical care."37 While the trial court found this explanation reasonable, the Mississippi Supreme Court was skeptical.
During the early morning hours of March 14, 1984, Mr. Drummond’s condition worsened. According to Mrs. Drummond, her husband’s "breathing had become really shallow and . . . he was in a great deal of pain."38 Mr. Drummond was then taken by ambulance to the Memorial Hospital ED. Mr. Drummond was admitted to the hospital that morning and, over the next several months, he received extensive inpatient and outpatient treatment for his infection. He spent a significant portion of his inpatient stay in the intensive care unit and, while he apparently eventually recovered, he incurred "very large medical bills for the treatment necessitated by his infection."39 It should be noted that a second issue in the case, which the Supreme Court ruled was a question of fact for the jury (i.e., the case could not be dismissed by the court), was whether Mr. Drummond was harmed by Dr. Buckley’s alleged malpractice. That is, would the months of treatment have been avoided if Mr. Drummond had been admitted on March 13, rather than March 14?
It was established that the standard of care required that Mr. Drummond be admitted to the hospital for treatment on March 13. (How this standard was established is the interesting side issue mentioned above and discussed below.) As a result, the case turned on whether there was a question of fact for the jury as to whether Mr. Drummond had left the ED on March 13 AMA (i.e., the content of the disputed conversation). The court found support for Mr. Drummond’s version in that "the record entry made by Dr. Buckley for [the March 13 ED visit] did not mention any recommendation that Drummond be hospitalized."40 That lack of documentation was sufficient to raise a question for a jury to decide as to whether Dr. Buckley had ever recommended hospitalization.
There was an interesting additional twist in the Drummond case. Shortly before trial, the one expert identified by the Drummonds who would testify at trial was disqualified by the circuit court: "Investigation by defendants into the background of the proffered expert witness had shown that the witness’ experience and credentials were highly suspect, and the witness refused to cooperate with discovery to clear up the numerous questions that had been raised regarding his credentials."41 In the mind of the trial judge, this created an insurmountable obstacle for the plaintiff’s case since, in Mississippi, as in other states, expert testimony is required to establish that the defendant physician breached the standard of care. The court refused the plaintiff’s motion for a continuance, and granted summary judgment for the defendant, Dr. Buckley, because the plaintiff was without an expert to testify as to the standard of care.
The Supreme Court of Mississippi, however, saw it differently. While the plaintiff may not have had its own expert witness, the Supreme Court agreed with the plaintiff that "the depositions of Buckley and Danielson established that the standard of care demanded that [Drummond] be told that he needed to be hospitalized or that his wound be cultured in the emergency room."42 That is, the plaintiff was able to use the deposition testimony of the defendant and Dr. Danielson, who was a defense witness, to establish the standard of care in its case. The relevant testimony was as follows:
[PLAINTIFF’S] COUNSEL: What did the standard of care require with regard to recommendations to Mr. Drummond about hospitalization on March 13th, 1985?
[DEFENDANT’S] COUNSEL: Note my objection.
DR. BUCKLEY: Exactly what I did, recommended him to be hospitalized.43
According to the court: "There is no magical form to which a plaintiff’s supporting expert opinion must conform, so long as its import is apparent."44
Once a qualified expert has given a delineation of the specific acts that need to be done to adhere to the standard of care, the court ruled that "the question of whether a breach of that standard occurred becomes a factual inquiry focusing on whether the physician did the delineated acts."45 In Mississippi, it is not necessary for an expert to opine that the defendant’s acts did not conform to the standard of care. It was, therefore, up to the jury to decide whether Dr. Buckley had recommended admission, i.e., whether Dr. Buckley had satisfied the standard of care he had set forth in his earlier deposition.
It is impossible to conclusively determine from the record what exactly was told to Mr. Drummond by Dr. Buckley. Therefore, it is impossible to determine without trial whether Dr. Buckley conformed to the conduct that he himself admitted was minimally required.46
No further expert testimony was necessary because, according to the court, the case came down to a simple yes or no question: Did Dr. Buckley recommend admission to Mr. Drummond when he saw him in the ED on March 13, 1985?
The Mississippi Supreme Court reversed the trial court’s judgment for Dr. Buckley and remanded the case for further proceedings.
Competence to Leave AMA
Case No. 4: Miller v. Rhode Island Hospital, et al.47
In this case decided by the Supreme Court of Rhode Island, the court dealt with the issue of competency to consent to, or refuse, medical treatment. The case involved an apparently intoxicated patient who had been involved in a motor vehicle accident and refused a recommended peritoneal lavage. He was restrained by hospital personnel, and the peritoneal lavage was performed against his expressed wishes. He brought suit and won a judgment for battery against the hospital, which the hospital appealed.
At trial, Mr. Miller testified that on Feb. 12, 1987, he had "approximately two alcoholic beverages with his lunch."48 He left the restaurant with a friend at about 3 p.m. and then, between approximately 5 p.m. and 8:45 p.m., he consumed "between five and 10 additional drinks" at another restaurant.49 Mr. Miller left the second restaurant as an unrestrained passenger in his friend’s car, which was subsequently hit by an oncoming vehicle. According to the court, Mr. Miller’s injuries included "lacerations over his right eye, on the bridge of his nose, and on the right side of his forehead," as well as "a bruise to his ribs."50
Mr. Miller was transported by ambulance to the hospital, where his condition was treated as a surgical emergency. He was taken to the trauma room, where he was evaluated by three trauma team physicians, and various diagnostic tests were performed. It was undisputed at trial that Mr. Miller’s blood alcohol level was 0.233.
When asked, Mr. Miller responded that "his head, eyes, back, and ribs caused him pain."51 Mr. Miller overheard the physicians discussing a procedure involving "an abdominal incision" (the peritoneal lavage), and he began questioning a doctor as to what was proposed.52 He testified that the procedure was explained to him in a brief fashion. Mr. Miller responded, "No, I don’t want you to do that,’ and he attempted to sit up and engage the doctor in a dialogue."53 It was Mr. Miller’s testimony that he was told by the doctor that: "Since you have been drinking, you’re not in a position to know the extent of any injuries, and this is our standard procedure for a situation of this kind."54 According to Mr. Miller, he was not asked to consent to the procedure and the physician did not inquire as to the availability of immediate family members.
Mr. Miller then attempted to get up, and the doctor tried to restrain him. Mr. Miller began yelling and, according to Mr. Miller, "the doctor pushed him, [he] struggled, and the doctor called a security guard."55 Mr. Miller was restrained by being strapped to the gurney and was "administered anesthesia through a syringe [apparently sedation]."56 When Mr. Miller regained consciousness, he found a "3-inch long incision on his stomach."57 Mr. Miller left the hospital the following morning AMA. While Mr. Miller eventually left the hospital AMA, the focus of the case was on the events that occurred the day before when the peritoneal lavage was performed. Mr. Miller refused the procedure, and the physician had him restrained. Whether he was attempting to leave AMA at that time is not known and is not necessary for this case to be relevant to our discussion. The issue was: When is a patient competent to refuse treatment? Patients who leave AMA are really a subset of those who refuse treatment, which makes the court’s analysis relevant to the AMA discussion.
In deciding the issue of Mr. Miller’s competence to refuse treatment, the trial court distinguished medical competence from legal competence. The trial court then ruled that medical competence was immaterial, and that the only relevant issue was Mr. Miller’s legal competence. Next, the court ruled that, if a patient is legally competent (and all adult patients are presumed to be legally competent unless previously determined to be legally incompetent), a physician may not perform a surgical procedure on the patient without the patient’s consent. The court then made a huge leap and ruled that intoxication never rebuts the presumption of legal competency and that, since Mr. Miller presumably awoke the morning of the accident legally competent, he remained legally competent despite his intoxication, which the court assumed was a "temporary condition." The judge so ruled despite the proffered testimony of the surgery resident who read notes from the medical records that stated that Mr. Miller "was intoxicated, combative, and uncooperative."58 The trial judge summarized his view of the law as follows:
I know of no decision anywhere where it is said that a person who is intoxicated, is not of sound mind, to wit, is not legally competent. *** A person whose judgment is impaired because of intoxication may make the wrong choice. But that choice is made not because of incompetency, but because if the person was sober, the person would not have made that choice, and here is the essential distinction, as this court sees it, this plaintiff, when he got up that morning, presumed he was legally competent, and he was legally competent when he sobered up. This court does not believe that the fact of intoxication creates a state of temporary incompetency, which permits the hospital to perform a procedure against a patient’s will.59
As a result, the jury was instructed that essentially no level of intoxication could render the patient legally incompetent and, as a result, protect the physicians from the plaintiff’s claim of battery.
In addressing that issue, the Rhode Island Supreme Court began by acknowledging that there is an ongoing tension between society’s competing interests of individualism — reflected in informed consent doctrine — on the one hand, and protecting people through effective emergency medical treatment. Given those competing interests, the court was compelled to find the middle ground: "To chart a course between the perils of insufficient emergency medical care and violation of a patient’s individual liberty."60
The court then provided a brief review of informed consent law. It observed that it is a well-established law that a patient generally has a right to control his or her body and to make an informed decision regarding medical treatment. That is reflected in the doctrine of informed consent, a right that is applicable only to competent adults (minors in some cases) and is limited in emergency situations. However, to keep the exception from devouring the right, the emergency exception must be "carefully circumscribed."61 For example, there is no emergency exception when a competent patient refuses treatment. "Competency" was, of course, the primary issue in the case.
According to the court, the trial court’s overly narrow focus on legal competency led to an incorrect ruling: "Legal competence and medical competence are two different standards. That a person is legally competent does not, however, mean that the person is capable of making medical decisions."62 According to the court, the fact that Mr. Miller was legally competent at the time of the peritoneal lavage, did not mean he was medically competent, i.e., capable of making medical decisions. That is, "a finding of legal incompetence . . . is not a prerequisite to determine that a patient lacks the ability to make decisions regarding treatment."63 Therefore, a patient may be legally competent, but medically incompetent.
What, then, is the standard for determining medical competency? The court joined a consensus emerging among other states that the standard of mental competency is that which is applicable to one’s capacity to contract. This is based upon individual circumstances and not derived from a general presumption (i.e., the presumption that adults are legally competent). This standard had been elucidated by the court in a previous case where it had held that the test for mental capacity to consent to, or refuse, medical treatment was whether the patient has "sufficient mind to reasonably understand the condition, the nature and effect of the proposed treatment, [and the] attendant risks in pursuing the treatment, and not pursuing the treatment . . . ."64 As to the trial court’s ruling that intoxication could never affect a patient’s competence to consent to treatment, the court’s comment was brief: "We disagree."65
Commentary: This case clearly outlines the information that should be disclosed to a patient, and documented in the medical record, before the patient leaves AMA.
Risk Management Strategies
• Make all reasonable efforts to avoid having a patient leave AMA.
Any time a patient leaves the ED AMA, prior to your completion of the patient’s evaluation and treatment, bad things can happen, and you might be required to defend your actions. At best, you will have to respond to the complaint letter the patient may write to the hospital administration; at worst, you will be defending your actions in court.
• Use all available resources to prevent a patient from leaving AMA.
It is crucial to involve the patient’s family in such situations for two reasons. First, they may be able to assist you in talking the patient out of leaving. Attempt to identify the one family member that you seem to have the best rapport with and focus on building an alliance with that individual such that he or she might assist you in de-escalating the situation. Second, always remember that, if the patient does leave AMA and dies, he or she is no longer available to testify as to how you repeatedly warned of the risk of leaving and implored him or her to stay. If you did not involve the family, they might not understand this and will be more likely to sue.
Yes, there are potential confidentiality issues involved when you enlist the aid of the family. However, while you will have to use your best judgment as to what must remain confidential, confidentiality should take a back seat to your efforts to get the patient to stay in the hospital.
• ALWAYS, ALWAYS carefully document when a patient leaves AMA.
When you are ultimately unable to convince the patient to stay and the patient is leaving AMA, carefully document exactly what you have told the patient regarding the risk of leaving.
• Do not simply have the patient sign a generic hospital AMA form.
There is nothing wrong with having the patient sign such a form. However, to protect yourself from the patient claiming that, while he or she may have signed the form, they did not read the form and were not aware of its import, write the specific risks down on the discharge instructions in plain English. You should list any potential serious complication you can think of and, in most cases, death. For example, if your chest pain patient is leaving AMA, include in your discharge instructions the following: "You are leaving against my medical advice. You may be having a heart attack, and I have recommended that you stay in the hospital. By leaving, you risk having a heart attack and suffering permanent damage to your heart, brain damage, or death." Read that to the patient, and have him or her sign it. Get the spouse or other family member to sign it as well, if possible. Sometimes, the presentation of that frank presentation of the potential risks to the patient will get the patient to reconsider and stay.
Carefully document in the medical record the conversation you had with the patient regarding why staying in the hospital was necessary and the risks of leaving. Document which family members you discussed that with and what personnel you enlisted to assist you in trying to get the patient to stay, e.g., a social worker. Have a nurse cosign your note or document separately the efforts that were made to get the patient to stay.
• Never let incompetent patients sign out AMA.
If you think the patient is competent enough to sign out AMA, carefully document the basis for this conclusion. Do not simply enter a conclusive statement in the medical record (e.g., "patient is competent"). Rather, you should include documentation of an adequate mental status examination.
• Correlate your level of coercion with the risk to the patient.
If the patient refuses to have a wound sutured, and you are comfortable that, with proper wound care (absent sutures), the wound will heal with only a larger scar and a small increased risk of infection, your efforts to get the patient to stay for sutures will be less than your efforts to get a chest pain patient to stay.
• Negotiate and compromise with patients.
The patient may be willing to stay if you agree not to perform some test. For example, the patient may be refusing a CT or MRI, perhaps because of the cost. Perhaps the patient clearly needs sutures but will accept the use of tissue adhesive, an inferior alternate in your opinion under the circumstances. In such situations, you should, of course, document that treatment is proceeding contrary to your best advice. If you can get the patient to stay, you might, in time, get the patient to cooperate with your treatment plan.
In the case of a child, negotiate reasonably with the parents. For example, you might give in on insisting on the lumbar puncture that they are convinced will paralyze their child if they will agree to admission and parenteral antibiotics that will provide meningitis coverage. If you have any concerns about any significant risk to the child, discuss with the parents your obligations under your state’s child protection law and involve children’s protective services and the police if necessary. Often, merely explaining the child protection system to the parents will get them to back off and cooperate.
• Never abandon a patient.
I sometimes hear physicians say that once a patient leaves AMA, they will not (some even say "legally may not") provide treatment to the patient. This is absolutely wrong. For example, if you advise admission for parenteral antibiotics and the patient refuses, offer available alternatives such as discharging the patient to return for twice-daily intravenous infusions of antibiotics or going home with oral antibiotics. Explain that they are not your first choice and might not be as effective.
• Always tell the patient that he or she may return at any time.
I sometimes hear of physicians instructing patients who are leaving AMA that, since they refuse to follow medical advice, they should seek subsequent treatment elsewhere. Never do this.
• Involve the police if necessary.
If there is concern that the patient might harm himself or herself, or others, and you are unable otherwise to prevent the patient from leaving, call the police for assistance.
• Always worry more about malpractice and abandonment than battery or false imprisonment.
Never allow your medical judgment to be unduly influenced by concerns of a patient’s subsequent claim of battery or false imprisonment. If a patient demands to leave AMA and you detain him or her, those will be the patient’s potential claims. If you wrongfully allow the patient to leave and there is an adverse outcome, the claim will be malpractice and/or abandonment. Consider the intoxicated or otherwise arguably incompetent patient with chest pain as an example. If he leaves AMA and suffers a cardiac arrest, you potentially will be involved in a malpractice lawsuit with the damages being death. If, on the other hand, you restrain the patient, the most you will have to defend is a claim of battery or false imprisonment with damages consisting of hours or days of lost "freedom." I know which claim I would prefer to defend.
• Remember that AMA is a process, not a form.
If you are going to allow a patient to leave AMA, you must carry on a complete informed consent discussion, if possible, with the patient. A signature on a form is not the end of matter. You must provide the patient will all information that a reasonable patient would require in order to make a decision to forego further care. As mentioned above, when possible, you should involve the family in this discussion. This is not always possible — for example, some patients slip out of the department unnoticed.
• Have systems in place to identify patients leaving prior to completion of their treatment.
EDs should be designed such that patients attempting to leave can be identified and efforts may be made to assess their competency and get them to stay. If they are competent and refuse to stay, they then can be informed of the risks of leaving. This is particularly important for psychiatric patients who may be suicidal or homicidal.
• Have policies and procedures in place to handle difficult AMA situations.
All EDs should prospectively address difficult AMA situations (e.g., the intoxicated or potentially suicidal patient) and develop policies and procedures for handling them (e.g., involvement of the local police agency).
• Attempt to minimize patient waiting.
Long patient waits tend to increase the incidence of patients leaving AMA. Patients who are triaged to lower-priority status understandably tend to wait the longest during busy times and might, as a result, leave AMA. Understand patients’ frustration with waiting when they might be terrified that, despite what the triage nurse might have said, they believe they might have a serious condition.
While it is simply impossible to eliminate long patient waits on occasion, efforts to minimize the frequency should be made (adequate staffing, effective on-call arrangements, sufficient treatment rooms, dedicated X-ray and laboratory, etc.).
• Apologize early and often.
In the ED, the rule should be to "apologize early and apologize often." There is merit in starting nearly every patient encounter in the ED with an apology for the wait. Make sure someone is apologizing at regular intervals. When the wait is prolonged, volunteers can be very helpful in this regard. If the patient has a complaint (e.g., the triage nurse or registration clerk was rude), apologize. While it might well have been the patient who was rude, not the nurse or clerk, just apologize. Explaining to the patient that the department is very busy, there was a huge number of "really sick patients" (i.e., "you’re not that sick"), or the clerk had a bad day, is almost universally a waste of time and will get you nowhere. Just apologize. It is always quicker and usually more effective.
Conclusion
It is a crucial risk management principle that the number of patients leaving the ED AMA should be as few as possible. It is, of course, impossible to totally eliminate having some patients leave AMA, but direct all efforts at reducing that number to the absolute minimum. In those rare cases in which a patient leaves the department AMA, he or she should be allowed to do so only after the physician has fully disclosed the risks of leaving in terms that are understandable to the patient. When a patient leaves AMA, always make sure that the medical record documents:
1) the patient’s competence to make the AMA decision (vital signs, mental status, etc.);
2) your discussion of all potential serious risks of leaving (be liberal in your inclusion of the risk of death);
3) your efforts to get the patient to stay;
4) your ongoing concern for the patient (e.g., instruction to the patient that he or she may return at any time).
Endnotes
1. 551 So.2d 1034 (Ala. 1989).
2. Id. at 1037.
3. Id. at 1035.
4. Id.
5. Id. at 1036.
6. Id.
7. Id.
8. Id.
9. Id.
10. Id.
11. Id.
12. Id.
13. Id.
14. Id.
15. Id. at 1037.
16. Id.
17. Id. at 1038.
18. 818 S.W.2d 940 (Ark. 1991).
19. Id. at 941.
20. Id.
21. Id. at 942.
22. Id.
23. Id.
24. Id. at 943.
25. Id.
26. Id.
27. Id.
28. 624 So.2d 264 (Miss. 1993).
29. Id. at 265.
30. Id.
31. Id.
32. Id.
33. Id. (emphasis added).
34. Id.
35. Id. 266.
36. Id.
37. Id. (footnote omitted).
38. Id.
39. Id. at 267.
40. Id. at 269.
41. Id. at 267.
42. Id. at 268.
43. Id.
44. Id.
45. Id. at 269.
46. Id. at 270.
47. 625 A.2d 778 (R.I. 1993).
48. Id. at 779.
49. Id.
50. Id.
51. Id.
52. Id. at 780.
53. Id.
54. Id.
55. Id.
56. Id.
57. Id.
58. Id. at 781.
59. Id. at 782.
60. Id. at 783.
61. Id. at 784.
62. Id. at 785.
63. Id.
64. Id. at 785-786, citing In the Matter of Schiller, 372 A.2d 360, 367 (N.J.Super. 1977).
65. Id. at 786.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content