Journal Logo



The Vagaries of Pediatric Patients Leaving Against Medical Advice

Roberts, James R. MD

Author Information
doi: 10.1097/01.EEM.0000490514.99455.fa
    At 3 a.m., this hard-working father, who has to go to work in five hours, is just another distraught and exhausted parent of a cranky febrile 3-month-old. Statistically, the infant probably only has a benign URI, but prolonged waiting times are the norm, albeit the bane of all, and quality evaluations take even more time. Parents have the autonomy and right to refuse testing, treatment, and even admission of their child as long as they are properly informed of the relevant issues and consequences, and are competent. A quick evaluation should allow the prescient clinician to make conclusions that will be beneficial to all; often special attention to the about-to-depart goes a long way in allowing cooler heads to prevail. Occasionally, the clinician must dig in and simply take control, and proceed with testing and even admission against the wishes of the parents. One cannot simply invoke child abuse statutes for no good reason, and overestimating the issues and progressing to a legal entity can get you into trouble with the courts. The physician is clearly between a rock and a hard place with this one. Pay very close attention to the paperwork — your charting has to be squeaky clean — if this child leaves AMA.

    Patient autonomy generally supersedes physician recommendations when an adult patient wants to leave the ED against medical advice (AMA), and the only requirement for a patient to leave the ED abruptly is that he is mentally competent and has been appropriately apprised of the risks of leaving. A properly executed and detailed AMA form may offer some protection to the physician and hospital, but it is not blanket immunity if an adverse event occurs. Patients who leave AMA are at higher risk for untoward events, recurrent admissions, and an overall increased cost to all involved.

    But what about children who are taken out of the ED by their parents? The general public might be somewhat sympathetic to a physician who does his level best to keep a difficult patient in the ED, particularly one that has a known rather than only a potential medical issue, but emotions run high when a child is involved. You don't want to be on the wrong end of a disaster when a child was plucked from the ED by his parents, with subsequent allegations that they were not fully informed of the risks. Three years later, in court, the milieu, tone, and contentions of each party are vastly different from what happened during that unfortunate ED encounter.

    Predictors of Pediatric Emergency Patients Discharged Against Medical Advice

    Reinke DA, Walker M, et al.

    Clinical Pediatr


    This is one of a few articles in the medical literature that specifically deals with pediatric patients removed from the ED AMA. It focuses mainly on the characteristics of parents and patients who choose to leave AMA, and the effect of such actions on return visits and hospital admissions. This four-year retrospective study from St. Louis Children's Hospital ED looks at children who were evaluated by a physician or a nurse practitioner and whose parents took them from the ED before medical evaluation or therapy was completed. Subjects were excluded if they were not evaluated or they left without notifying the staff (elopement).

    A total of 94 of the 250,000 children were identified as leaving AMA. The most frequent chief complaints were abdominal pain, URI, otitis media, asthma, trauma, and fever. No difference was seen in the sex of the child, race or ethnicity, or insurance status between those who left and those who stayed. The time of arrival, time to be seen by medical personnel, and day of arrival also were not significantly different. Half of the patients were seen within 30 minutes of arrival, but eight percent waited longer than three hours. No difference was seen on patient age, mode of arrival, mother or father in attendance, or the use of consultants. More than half of the children were discharged AMA before completion of diagnostic testing. As one might suspect, chart documentation was poor.

    About a quarter of the patients returned to the ED within 15 days, three times the rate of those routinely discharged. The complaint was similar most of the time, and the admission rate at the time of the second visit was about 25 percent, not significantly different from control groups.

    The characteristics of those who left AMA appeared no different from adults leaving AMA. Adolescents, however, were at increased risk, as were patients with more severe illness or complex medical problems. Abdominal pain was a predictor of AMA, possibly because the workup is time-consuming, complex, and may require consultation. Adolescents who self-registered were also at increased risk for AMA. The presence of sexually-related illness, with concerns about parental knowledge, may have been a factor.

    Providers consistently failed to document alternative therapy following AMA discharge, and the authors revised their AMA form to address understanding of the diagnosis, proposed management, alternative therapies, risk of no treatment, and specific follow-up instructions. Focusing on adolescence and patients with more complex medical problems may decrease instances of AMA.

    Comment: This article has minimal specific information. Other than focusing one's efforts on better documentation and perhaps realizing that adolescents leave more often than children, the authors offer no concrete, constructive suggestions. One can readily understand the reticence of a teenager — or an adult — with symptoms of an STD to stay in the ED for fear of publicizing his visit. Presumably, adolescents who arrived alone were considered competent adults, so parents/guardians were not contacted, an unfortunate fallout of the restrictive HIPAA. Many patients left prior to diagnostic testing, so it would be difficult to clearly identify risks of leaving AMA other than to say the diagnosis was not yet certain.

    Removing obviously ill children from the ED is a difficult issue. It is clear that no standard of care exists. A rather fascinating questionnaire survey by Salem, et al., asked board certified emergency physicians to comment on a mother who removed a febrile infant from the ED. (Cal J Emerg Med [abstract] 2002;3[3]:45.) The mother in this fictitious case did not appear to be intoxicated or confused, but was upset with nursing issues and became angry. She subsequently left the ED. Thirty-two percent of almost 700 board certified EPs said they would allow the mother to leave and take no further action, 32 percent said that they would request an AMA form prior to leaving, 19 percent stated that they would allow the mother to leave but report the case to child protective services, and 17 percent said they would call security to prevent the mother from leaving. Clearly most physicians believed in parental autonomy to make decisions about their seemingly ill children. It is also clear that obvious reflexive action is not to restrain the child or parent, particularly when there was no evidence of parental negligence or an obvious suspicion for a serious illness in an unstable patient.

    It is universally suggested that physicians document specific medical risks of leaving on the AMA form. Many patients left prior to diagnostic testing, so it's difficult to clearly identify specific risks other than to disclose that the diagnosis was uncertain. “You could die” is a common warning, but it's a bit tricky to chart that a child could die if he left and then justify why you allowed him to be taken away. Things are more clear-cut when the child is obviously gravely ill and reasonable parents can often sense impending doom. The issue of parents removing only potentially ill children or those with vague, nonspecific, improving symptoms from the ED can be a dilemma for all, especially for the child who has no say in the matter. Any abdominal pain can presumably be due to appendicitis, and that disease can kill.

    This unexplained fracture conjures up concern for child abuse, but was actually due to Osteogenesis Imperfecta.

    No standards seem to exist other than recognizing parental autonomy and informed consent. Hopefully, common sense by all parties will prevail, and the professionals (physicians, nurses) will eschew the good-riddance approach, but an outraged, sleep-deprived, hostile, intolerant, and occasionally intoxicated parent is tough. At 4 a.m., the desire to leave is great, especially when parents conclude that the ED visit was not really necessary, and it is certainly taking too long for a seemingly minor illness. Giving acetaminophen in triage can make a lot of fussy kids look a whole lot better, and nonverbal body language and sympathetic nurses can give assurances that prompt an exit.

    Patients who Leave the Pediatric Emergency Department without Treatment

    Dershewitz RA, Paichel W

    Ann Emerg Med


    The main reasons that 223 of 27,000 pediatric patients (1.1%) left AMA were they waited too long (60%); the medical process spontaneously resolved (10%); and a plethora of transportation, anger, and misunderstood instruction issues. About five percent of patients had their concerns allayed by a nurse before physician evaluation. I don't think waiting times have decreased since this study was done, but the authors don't offer any specific recommendation except to work faster and try to communicate better with patients. Wait times are the bane of EDs, with little hope on the horizon. It's no secret that efficiency leads to patient satisfaction and fewer AMAs.

    A study by Bourgeois found that about 2.5 percent of visits left AMA from Children's Hospital of Boston. (Ann Emerg Med 2008;52[6]:599.) They identified an urban location, self-pay insurance, less acute medical problems, race and ethnicity, and arrival times as factors associated with leaving AMA. Again, no suggestion on how to address the issues. AMA issues are generally similar in adults and children.

    Comment: No one has successfully addressed the issue of unreasonable parents snatching their child from the ED prior to complete evaluation. While there are appropriate reasons for this to occur, the EP is on the wrong end of the deal on this one. Not only do they have to schmoose the irate parents, they have to make snap judgments regarding the potential illness of a child with minimal or no data. Infants with fever rarely have life-threatening illnesses, and one is often left with a gestalt or a gut feeling about whether a child has a serious bacterial illness requiring further treatment or evaluation. No one should allow a parent to take a child from the ED when an LP demonstrated bacterial meningitis, even invoking the required force, restraint, and security guards. The more vexing problem is intervening when the child does not appear on death's door, yet you are just not sure about the seriousness of the condition. Sometimes all the child needs is Tylenol, a bit of time, and some fluid to perk up.

    Sometimes, it requires blood tests, spinal taps, and CT scans. As with adults, ACEP guidelines allows for parents to have great autonomy over the medical care of their children. If you want to get really steamed, read an EMN article detailing a case where an EP was sued for performing an LP on a child with suspected meningitis against the wishes of the mother. (2007;29[7]:1; Amazingly, this legal action was supported by the Center for Individual Rights in Washington, D.C. Check out their website for a real eye-opener on dubious personal rights organizations. ( I find this litigation totally appalling as well as just plain stupid, and yet another situation where the EP can't win.

    I have never personally called a security guard to take a child away from the parents unless there was obvious child abuse or a known life-threatening medical condition. A physician should rarely be the surrogate for an ill child when the parents are in the room. A retrospect analysis, however, is a powerful tool, and a bad outcome is often taken as evidence of physician misconduct. Parents need to take responsibility for their child's welfare in the long run, and cannot blame physicians for their own hasty mistakes or poor judgment. The best one can do is to try to diffuse any hostile situations, let common sense prevail, and do all the other negotiations and warm and fuzzy techniques that we have learned in dealing with adults who want to make foolish decisions about their health. Proper documentation should at least convey your gargantuan efforts to revise medical care for the defenseless and helpless child who is legal custody of his parents.

    Most state laws allow parents to control health issues affecting their children, and physicians are empowered to intercede in cases of obvious child abuse or neglect. Parents' rights, including religious beliefs, do not include the right to deny life-sustaining medical intervention for their children. Laws differ by state, and some require a second physician to come to the same conclusion as the first one attempting to take control of a child's health issues. Most courts allow parents to refuse medical treatment if there is no threat to life or potential for serious impairment.

    Herein lies the conundrum. Without a period of observation or certain testing, it's merely an educated guess in some cases that a child has a serious or a life-threatening illness. Under these circumstances, there is no agreed-upon standard of care other than to provide documentation of the potential risks and the competency of the parents. Unfortunately, the last thing one attends to in a stressful case is the paperwork, and clearly it is rarely pristine. Even when it is airtight, you can spend three weeks in court trying to convince a jury that your story is more believable than the parent's recollection of the events.

    This is not just an American problem. A review article from Iran concluded that a more satisfactory facility and more effective communication between medical staff and parents could have prevented the majority of the 5.3 percent of pediatric cases from being removed from a hospital AMA. (World J Pediatr 2010;6[4]:353.) Sound familiar?

    Between a Rock and a Hard Place: When Parents Refuse Treatment for Their Children in the ED

    McDonnell WM

    ED Legal Letter


    This erudite and informative article gives a more legal slant to the conundrum of children being removed from ED medical care by their parents. It's peppered with case law references, and it's an easy read, although it's discouraging. The entire article should be studied in detail in journal club.

    The author, a pediatric emergency physician and lawyer, reiterates the fact that courts recognize a strong public interest in allowing parents to make reasonable medical decisions on behalf of their children (citing nothing less than the 14th Amendment to the U.S. Constitution). The operative word here is “reasonable,” and that's in the eye of the reviewer, often a retrospective eye at that. The rock and hard place is the potentially conflicting issue of child welfare and parental autonomy. Physician obligations and parental rights can conflict, especially in a stressful ED where little rapport has been established in an already-charged scenario of long waiting times, hostile counterparts, and ill children.

    Case law confirms that a patient or parent must agree to evaluation prior to any physical contact. Failure to do so exposes the physician to potential claims of assault and battery. Of course, federal law (EMTALA) also requires that physicians provide an appropriate examination and stabilizing treatment to all. It's easy when the patient is incompetent or unconscious but less clear when dealing with potentially ill children. Even critically ill patients have the right to refuse medical care if they so choose. And that's not only end-of-life decisions, but also extends to the ability to refuse admission or further treatment for severe asthma if the patient is capable of making such decisions. How's that for an example of the quintessential gray area?

    To make things even more complicated, the principle of parens patriae (parent of the state) gives the state an interest in the welfare of all adults and children. This concept is not a mechanism for the physician but one for the state to override parental rights. It might help with feeding tubes and chemotherapy, but it is of minimal practical guidance at midnight when the parents want to take their febrile child home before the EP has a chance to evaluate him fully.

    One must be careful when trying to protect the child through child abuse laws. Child abuse does not have to be proven by the physician, merely suspected. Physicians are liable if they don't appropriately suspect or report child abuse, but are also liable if they overstate or exaggerate medical risk to institute care against parental wishes. It appears that not only the child, but also the physician, can get squashed between that rock and that hard place.

    Child Abuse or a Rare Medical Condition?

    The literal poster child for a potential minefield in suspected child abuse is Osteogenesis Imperfecta (OI). Unexplained or multiple fractures in a child usually raise a red flag on anyone's list of child abuse characteristics, but OI can be clandestine and clinically mild. Medical personnel have been prosecuted when erroneously filing child abuse complaints about children with OI. Watch out for this one, and look for the blue sclera.

    From the website of the Osteogenesis Imperfecta (OI) Foundation ( “A child is brought into the emergency room with a fractured leg. The parents are unable to explain how the leg fractured. X-rays reveal several other fractures in various stages of healing. The parents say they did not know about these fractures, and cannot explain what might have caused them. Hospital personnel call child welfare services to report a suspected case of child abuse. The child is taken away from the parents and placed in foster care.”

    Also see Wardinsky TD, et al., “The Mistaken Diagnosis of Child Abuse: A Three-Year USAF Medical Center Analysis and Literature Review,” Mil Med 1995;160[1]:15.

    Reader Feedback: Readers are invited to ask specific questions and offer personal experiences, comments, or observations on InFocus topics. Literature references are appreciated. Pertinent responses will be published in a future issue. Please send comments to [email protected].

    AMA Forms are a Necessary Risk Management Tool

    Dr. Roberts: Thank you for your article about AMA forms. (“Standard AMA Forms Offer No Real Protection,” EMN 2016;38[5]:22; You performed a great service to emergency physicians through your outstanding articles in EMN down through the years. Therefore, I reluctantly disagree with the headline of this article and with some of the statements minimizing the importance of the AMA form.

    The article cited from the family practice literature did not prove that AMA forms “should be considered legally worthless.” In my opinion, the headline, “Standard AMA Forms Offer No Real Protection” is inaccurate and misleading. The text of the article somewhat negates the headline by describing how to properly complete the AMA form and the importance of properly assessing the patient's capacity to understand.

    Law is perhaps the ultimate anti-science. In the practice of law, anecdote does matter. In fact, when attorneys present their arguments in court, they primarily cite case law, which is nothing more than a series of anecdotes. In my 10 years of defending emergency physicians in medical malpractice lawsuits, we had three cases where patients left the emergency department against medical advice, but the emergency physicians did not attempt to have the patients sign an AMA form. In all three cases, the patients stated that the physician did not want them to stay in the emergency department and that the emergency physician decided to discharge them to home. We could not get these cases dismissed because of the lack of an AMA form.

    In these three cases, the physicians decided not to complete AMA forms because they did not want to confront the patients. However, this process does not ever require confrontation. Most patients refuse medical advice for reasons unrelated to the quality or nature of medical care. They do so for personal reasons or responsibilities. (J Emerg Med 2011;41[4]:412.) The entire process of interacting with such patients can be performed in an empathetic and supportive manner, always encouraging the patient to return when possible.

    Important decisions from state supreme courts held for the patients because the physicians did not have the patients sign an AMA form. (J Emerg Med 2011;41[4]:412.) These cases were decided by important courts and have great precedential value in their states and persuasive value in other states. Physicians will have significant difficulties convincing courts that a patient left against medical advice if they fail to document such refusal. The AMA form is standard practice and provides excellent documentation.

    In addition to your outstanding advice to focus on the patient's capacity to understand, you correctly stated that patients discharged against medical advice have a higher rate of readmission and death. This fact emphasizes the importance of properly documenting such refusals. Certainly, patients who sign an AMA form can still file a lawsuit, but patients can file a lawsuit for no good reason at all. We live in a society where 83 percent of lawsuits against physicians are groundless. (New Engl J Med 1996;335[26]:1963.)

    When properly completed, the AMA form provides strong evidence that the patient left the emergency department against medical advice. This creates an assumption of risk on behalf of the patient. This means that a patient who has the capacity to understand will assume the risk of the improper discharge, an important legal principle that will assist in the physician's defense.

    Physicians who work in the United States practice in the world's most dangerous legal environment. We can ill afford to assume additional risks. I do not believe the text of your article advises physicians to dispense with the use of AMA forms, but I believe the headline of the article, the paper cited by the article, and some of the sweeping conclusions in the article will encourage some physicians to stop using AMA forms. In many cases, these physicians will assume the risks of their patients' dangerous decisions. — Larry D. Weiss, MD, JD, Baltimore, MD

    Dr. Roberts responds: Thanks for your insightful comments, Dr. Weiss. I see that you are a lawyer and have significant knowledge of the topic. I think you slightly misunderstood the article, but maybe I did not convey my thoughts clearly enough. The key word in the headline is “Standard.” If you look at the standard AMA form example I included, it is obvious that it was written by a lawyer, and the physician did nothing except sign it.

    I did not suggest that an AMA form is of no protection, but thought I strongly promulgated that an expanded and detailed form always be used, even if the patient refuses to sign it. I emphasized that the one-size-fits-all, 20-year-old standard form is woefully inadequate to document any required interactions or concerns of the clinician. That hastily signed form, in my opinion, is essentially worthless. Hence, I included a copy of an AMA form that I believe is far superior. It actually conveys massive clinician involvement to inform the patient and documents that the interaction occurred in an all-knowing environment for both clinician and patient.

    I have been in two malpractice cases myself where a standard AMA form was used, and a trial nonetheless ensued. The physicians technically won the cases, but lost many worrisome hours fretting about it and had to spend a week or more in court. I think we both agree that the AMA form is standard practice and the form I suggested but not the standard ones provides excellent documentation. I see no documentation of anything other than the patient “signed his health away” on the standard uninformative AMA form that most hospitals use.

    Maybe a detailed physician note on the chart negates the need for this expanded form, but when this one is signed by the patient, there is no argument of what transpired. I disagree with your conclusion that this article will prompt physicians to avoid an AMA form. Quite the contrary, I think it strongly advises them to use one, but to use one that has enough information on it to document the situation correctly.

    Read InFocus and Earn CME!

    Earn CME by completing a quiz about this article. You may read the article here, on our website, or in our iPad app, and then complete the quiz, answering at least 70 percent of the questions correctly to earn CME credit. The cost of the CME exam is $10. The payment covers processing and certificate fees.

    Visit for more information about this educational offering and to complete the CME activity. This enduring material is available to physicians in all specialties, nurses, and other allied health professionals. Lippincott Continuing Medical Education Institute, Inc., is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. Lippincott Continuing Medical Education Institute, Inc., designates this enduring material for a maximum of 1 AMA PRA Category 1 Credit™. Physicians should only claim credit commensurate with the extent of their participation in the activity. This activity expires Aug. 31, 2016.

    Learning Objectives for This Month's CME Activity: After participating in this CME activity, readers should be better able to differentiate potential medical outcomes for children whose parents remove them from the ED AMA and prepare a plan to discourage parents from prematurely removing their children from the ED.

    Share this article on Twitter and Facebook.

    Access the links in EMN by reading this on our website or in our free iPad app, both available at

    Comments? Write to us at [email protected].

    Wolters Kluwer Health, Inc. All rights reserved.