A 21-year-old woman was transferred to Children's Memorial Hospital following an episode of supraventricular tachycardia (SVT) and subsequent cardiac arrest at an outside hospital. She was born with transposition of the great vessels, ventricular septal defect, tricuspid valve atresia, and coarctation of the aorta. She had had numerous palliative cardiac surgical procedures, including a "palliative arterial switch procedure" at 7 years. As a result, she had a long history of recurrent episodes of SVT as well as chronic noncompliance. She had not been seen here for 6 years prior to this admission.
On arrival to the referring hospital, she was awake and lucid and had a heart rate of 270. Our understanding, although this is poorly documented, is that she told the physicians that this was one of her typical episodes of SVT and that she had never responded in the past to either conventional vagal maneuvers or to the intravenous administration of adenosine. Rather, she stated that she should undergo cardioversion. Despite her requests, she was given intravenous adenosine. Within one minute, she developed a slow junctional rhythm, followed by hypotension requiring mechanical ventilation. She was transferred to the intensive care unit (ICU) where she received continuous intravenous infusions of dopamine and epinephrine as well as a dose of an intravenous fluoroquinolone antibiotic, levofloxacin. During her prolonged episode of hypotension, she had several episodes of cardiac arrest as well as three seizures.
Her past medical history was remarkable for an episode of pneumococcal meningitis at age 3 and a subsequent seizure disorder. Her medications included digoxin and diltiazem, although she had a long history of noncompliance with both medications and physician visits. She also had a 4-year-old son.
When she arrived at Children's Memorial Hospital, she was receiving mechanical ventilation and was alternately unresponsive and agitated. She was afebrile, her heart rate varied from 90 to 110, and her blood pressure was 160/80. There was jugular venous distention to the angle of the jaw. Her lungs were clear. Her precordium had a strong impulse along the left lower sternal border and her apex was displaced toward the left anterior axillary line. Sl was variable. S2 was single and loud. There were no murmurs. Her abdomen was soft and her liver was palpable 3 cm below the right costal margin. There was both clubbing and cyanosis of her fingertips. On neurological examination, she opened her eyes to both her name and noxious stimulation. She had normal corneal and doll's eyes reflexes bilaterally. She had no spontaneous motor movements, but she would withdraw both her lower extremities to noxious stimulation. Her deep tendon reflexes were 2+ on the right, 3+ on the left; she had a Babinski response on the left.
Her initial electrocardiogram revealed normal sinus atrial rhythm, with frequent atrial and ventricular premature contractions. There were short runs of unsustained ventricular tachycardia at a rate of 250 beats per minute. Her initial electroencephelogram (EEG) was normal and her magnetic resonance image (MRI) showed diffuse cerebral volume loss.
Robert Listernick, MD, moderator: Let's start off talking about her underlying heart disease.
Rae-Ellen Kavey, MD, pediatric cardiologist: Regardless of the complexity of her anatomy, she essentially had a single ventricle, irreversible pulmonary hypertension in her left lung, and normal right pulmonary artery pressures. Her multiple surgical procedures during childhood were attempts at palliation of what was ultimately going to be a fatal problem, the vascular disease in her left lung. The "palliative" arterial switch procedure was simply an attempt to provide more oxygenated blood to the systemic circulation.
Dr. Listemick: What would have been the surgical approach if she had been born in 2002?
Dr. Kavey: Patients who have a "palliative" single ventricle have passive flow of blood on the right side of the heart. A Fontan operation allows separation of the two circulations, creating a conduit from the right atrium to the pulmonary artery. While this operation may be lifeprolonging, it is not a cure, and these patients still have substantially shortened life expectancies. They have very limited exercise capacity as they cannot increase substantially their cardiac output. In addition, atrial arrhythmias are a common postoperative complication because the right atrial wall dilates.
Dr. Listernick: Was she a good candidate for heart transplantation?
Elfriede Pani, MD, pediatric cardiologist: She would have needed a combined heart-lung transplant to correct her anatomy and pulmonary hypertension. Her previous thoracotomies would have made this technically extremely difficult. Her risk of postoperative mortality would have been so high that this was not felt to be a viable option.
Dr. Listernick: Her MRI showed evidence of chronic volume loss. How does chronic cyanosis affect the brain?
Dr. Kavey: Multiple small thrombotic events can lead to intellectual deficits.
Dr. Listernick: One issue this case raises is just what is the responsibility of a treating physician when faced with an adult or parent who wishes to receive a particular therapeutic approach to a medical problem. The specifics of this case are not entirely clear, as there is much hearsay and the records from the referring hospital are incomplete. However, for our purposes, this leads to an interesting point for discussion.
Joel Frader, MD, pediatric ethicist: Obviously, if s extremely important to listen to patients and parents. I'd turn the question around and ask why shouldn't we pay attention and perform what is asked.
David Steinhorn, MD, pediatric intensivist: One reason to listen, but not necessarily to regard with absolute credibility what patients say, is that they often misunderstand sophisticated diagnoses or the subtleties of conflicting therapies. The treating physician has to decide which patient-dictated treatments are reasonable and which shouldn't be performed.
Dr. Frader: I agree, but we shouldn't assume that the patient or the parent is wrong. Why not assume that they're right and put the burden on the physician to provide evidence to the contrary?
Dr. Kavey: Ideally, the patient should have an information sheet with all the details of her illness written down. However, she had a long history of noncompliance and had not been seen for several years. Regardless, apparently she quite clearly explained that she had had multiple episodes of SVT in the past and that adenosine had never helped. Certainly, this information should have been taken into account before a therapeutic decision was made.
Dr. Frader: If the patient appeared to have adequate decision-making capacity, her information should have been taken seriously. If there are sound medical reasons supporting alternate therapies, those should be discussed with the patient or parent.
Dr. Listernick: Ideally, how should she have been treated?
Dr. Kavey: Her underlying rhythm was almost certainly atrial flutter, as it had been in the past. Adenosine will not convert a patient in atrial flutter to sinus rhythm; rather, it will cause a blockade at the atrioventricular node leading to a slower ventricular rhythm from which many patients will become symptomatic. Once recognized, atrial flutter should be treated by cardioversion.
Denise Goodman, MD, pediatric intensivist: At the outside hospital, she underwent cardioversion into a very slow junctional rhythm. Following this, she developed a second unrecognized rhythm disturbance before being transferred. On arrival here, she was noted to be in Torsades de pointes. This is a polymorphic ventricular tachyarrhythmia that has the appearance of slow ventricular flutter without discernible QRS complexes or T waves. The ventricular activity has constantly changing amplitudes and undulations that appear alternately above and below the baseline. Torsades de pointes almost always occurs in the presence of prolongation of the QTc interval. It is usually unsustained, but occasionally may persist for more than a minute. It may be caused by electrolyte abnormalities such as hypokalemia or hypomagnesemia, or it may be associated with the administration of a number of drugs including quinidine, procainamide, phenothiazines, or, as in her case, the fluoroquinolones. Use of this group of antibiotics should be avoided in patients who have atrial arrhythmias for just this reason. She underwent cardioversion again to restore a normal sinus rhythm.
Dr. Listernick: Moving on, she was transferred out of the ICU. Her mental status was extremely abnormal. Although she was noncommunicative, she would respond to verbal commands with purposeful movements. After several days, she developed progressively increasing agitation to the point where she became extensively bruised from banging against the bed rails. She no longer responded to external stimuli. Sedatives were used extremely cautiously because of concern regarding potential exacerbation of her previous arrhythmias. Her EEG became progressively disorganized. One week later she developed marked tachypnea and hypoxia necessitating mechanical ventilation.
Obviously this case raises multiple issues. First, did we ever talk to her or her family about advance directives for end-of-life care? Who gets to make these decisions?
Dr. Pahl: This family had been noncompliant for years and had not been seen for a long time. Ideally, I try to discuss this issue, including others such as birth control, long before an emergent situation. However, we all clearly don't do this enough.
Dr. Fraden The presumption should be that the adolescent be included in these discussions. An adolescent who has had experience with chronic illness, particularly repeated hospitalizations and exposure to high-risk interventions such as intensive care and mechanical ventilation, has at least some understanding of what she is facing. Depending on the particular adolescent, including factors such as intelligence, maturity, family structure, and others that may be hard to measure, she should have the authority to either accept or reject recommended care, including recommendations about limiting life support, should that become an issue.
Dr. Steinhorn; Waiting until the last minute to deal with these issues is a perennial problem both in academic health centers and in the general community. Both general pediatricians and academic subspecialists need to be proactive in obtaining these advance directives long before these patients reach the ICU.
Dr. Pahl: How does a physician implement a do not resuscitate (DNR) order?
Dr. Frader: Although the details are not yet clearly worked out by the Illinois Department of Health, there is a universal DNR form in use in the state, printed on bright orange paper. It applies to both adults and children. It should be immediately available to health care workers, including emergency medical services providers, in the home and on arrival at the hospital. Although the law is still somewhat murky, this form should be sufficient for use without revision once the patient is admitted to the hospital.
Dr. Listernick: How can we assess her mental status and make a prognosis as to longterm neurologic function?
Gretchen Wieck, MD, pediatric neurology fellow: There are multiple factors that need to be taken into account before one can even begin to try to prognosticate. First, one must make sure that she was not receiving any drugs that might alter her mental status, thus confusing the picture. Next, patients who have been in the ICU for prolonged periods may develop a delirium or a reversible psychosis. She was receiving no such drugs and had been in the ICU for only 1 day prior to this assessment. If the initial EEG is relatively normal, organized with a good background, there is certainly a chance for clinical recovery, although this is by no means assured. If the initial EEG is markedly abnormal and disorganized, the prognosis for neurologic recovery is poor.
Although her EEG on admission was normal, it became increasingly disorganized over the next several days. Her encephalopathy never improved; she was unresponsive to verbal stimuli and continued to have marked agitation. A repeat MRI was unchanged; there was no radiographic evidence of ischemic damage. By the end of one week, this accumulated information led us to conclude that she had sustained profound irreversible neurologic damage.
Dr. Steinhorn: There's a well-described entity following hypoxic-ischemic brain injury called paroxysmal thalamic storm. Patients, who are generally in a vegetative state, will have episodic hypertension, hyperthermia, tachypnea, tachycardia, and posturing. Although the etiology is poorly understood, the current view is that it represents disruption of brainstem autonomic function.
Karen Emerick, MD, pediatric gastroenterologist: We have found that the Glasgow Coma Scale has been extremely helpful in grading the level of encephalopathy in patients who have liver failure. Using a standardized scoring system minimizes the vagaries of interobserver differences as well as those of serial examinations by the same observer.
Dr. Listernick: Once her physicians agreed that she had an irreversible encephalopathy, how did you proceed?
Dr. Pahl: The first issue we had to decide was who should be making decisions regarding her care. She had been living independently for at least 4 years, and there was an initial sense that she had been estranged from her family.
Dr. Frader: Although the word "estranged" was used repeatedly to describe her relationship with her family, on closer examination her parents would baby-sit occasionally for the patient's child and were responsible for some of their daughter's financial support.
Dr. Listernick: What are the legal aspects of establishing who is responsible for decision-making for incapacitated patients?
Dr. Frader: According to Illinois law, there is a clearly defined hierarchy of individuals responsible for making such decisions. The law specifies an order of priority and includes court-appointed surrogates, parents, spouse, adult children, or any adult sibling of the patient. Most, but not all, states have a similar list. If there are no individuals who meet these requirements, the law is much less clear. The absolute safest route is to obtain a court-appointed guardian. However, this is often a slow, cumbersome process, and many hospitals have policies in place that allow decisions to be made without going to court.
In this case, the question was raised as to whether the parents should be disqualified from holding this responsibility; after some discussion, there was clear agreement that there was no good reason to bar the parents from assuming this role.
Dr. Pahl: What are acceptable reasons to bar the parents from being the decision makers?
Dr. Fraden Certainly any obvious conflicts of interest or if members of the health care team feel that the parents would not act in the best interest of the patient. A court would have to make the definitive determination.
Dr. Listernick: Once it was agreed that the parents should be the decision makers, how was the information presented?
Dr. Frader: As a team, we explained the irreversible nature of her encephalopathy. What was most striking was a comment made by the mother. She said, in effect, that she had known for a long time that her daughter was going to the and wondered why it took so long for the physicians to tell her! They agreed that we should provide only comfort care from that point on.
Dr. Steinhorn: Obviously patients and their families are generally a lot more insightful than we believe. If s often helpful to ask families about their perception of their child's illness as a way of furthering the discussion.
Dr. Listernick: As I understand it, she was successfully extubated. Since the goal was to keep her comfortable, how might one handle her fluids, nutrition, and medications? For instance, would it be acceptable to withhold nasogastric feedings?
Dr. Frader: With the agreement of the family, that is a perfectly acceptable option. Most importantly, there is no legal obstacle to withholding feedings in this context. The family was going to consider this option, but she died before any decision could be made.
Dr. Listernick: Don't patients who aren't fed suffer?
Dr. Frader: There are data both from people who voluntarily stop eating or drinking for political reasons and from tortured prisoners. They say that after roughly 48 hours they stop feeling hungry and after approximately 72 hours they stop feeling thirsty. More important, the vast majority of patients in whom this would be considered have severely impaired consciousness. However, we would typically provide sedation and analgesia for such patients in order to ensure that we had minimized suffering.
Dr. Steinhorn: The ethical definition of suffering requires the individual to be conscious or aware. Were these medications necessary?
Dr. Frader: First, we really can't absolutely answer the question, metaphysically or physiologically, as to her level of awareness. Second, if s important to realize that these medications also allow her caregivers and observers to feel confident that any suffering has been alleviated.
Dr. Steinhorn: That's exactly my point. The intention to ease suffering should always be present; that's the metaphysical aspect that nobody can quantify or answer objectively. It's a well-established concept in the medical ethics literature that the alleviation of perceived suffering is beneficial.
Dr. Listernick: Thank you everybody.