Clinical Ethics in Anesthesiology. A Case-Based Textbook

1.Consent and refusal

6. Do not resuscitate decisions in pediatric patients

Kelly N. Michelson and Joel E. Frader

The Case

A 4-year-old boy with metastatic neuroblastoma undergoes stem cell transplantation following intensive chemotherapy and radiation. 17 days post-transplant, with the success of anti-cancer treatment and stem cell rescue unclear, he develops an acute bowel obstruction. His parents feel that, after many months of therapy, especially over the last few weeks, their son should not have to endure “heroic” treatment. They do not want him to have cardiopulmonary resuscitation (CPR) and the oncologists have agreed to a “do not resuscitate” (DNR) order, given the boy’s overall poor prognosis. The parents would like him to have palliative surgery to relieve the bowel obstruction. The anesthesiologists and surgeons request suspension of the DNR order for the surgery. The parents do not understand why it is acceptable to forgo resuscitation on the oncology unit, but not in the operating room.

Attempts to resuscitate a person from an apparently “lifeless” state date back to at least biblical times:

When Elisha came into the house, he saw the child lying dead on his bed. So he went in and closed the door on the two of them, and prayed to the Lord. Then he got up on the bed and lay upon the child, putting his mouth upon his mouth, his eyes upon his eyes, and his hands upon his hands; and while he lay bent over him, the flesh of the child became warm.(2 Kings 4:32–34)1

In 1878, Boehm described closed chest cardiac massage in cats, the basis for current CPR. The first successful resuscitation in humans using CPR, reported in 1960, involved five patients who experienced in-hospital cardiac arrests.2

The use of CPR spread and in 1966 the National Academy of Sciences’ National Research Council recommended all health care providers obtain CPR training. CPR became a nearly ubiquitous final procedure for all hospitalized patients experiencing cardiopulmonary arrest, regardless of circumstances. Later, healthcare providers questioned the indiscriminate use of CPR, particularly with terminally ill patients for whom resuscitations seemed to provide no benefit and who might experience suffering related to CPR. In some circumstances, professionals began to decide, albeit arbitrarily and without input from patients or their loved ones, who should or should not have CPR. Language such as “show code,” “Hollywood code,” or “slow code” emerged to describe sham resuscitations when healthcare providers chose not to make serious efforts to revive patients.

In 1974, the American Medical Association (AMA) described CPR as a procedure meant to prevent sudden, unexpected death that has no valid use in patients with terminal, irreversible illnesses. They stated, “Resuscitation in these circumstances may represent a positive violation of an individual’s right to die with dignity,” and proposed that actual orders be written in the patient’s chart when a physician determines that CPR is not indicated.3 In response, hospitals developed specific DNR order policies.

Arguments about the patients’ right to self-determination ensued. Some felt practitioners should assume that all patients prefer resuscitation unless they or their valid surrogates have clearly stated otherwise. Others argued for only offering CPR when “medically indicated.” A 1983 report of the President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral established a standard whereby consent for CPR was presumed unless specifically withdrawn following discussion between the patient or surrogate and the involved physicians.4


The original term, DNR, persists at many institutions. Some institutions use “DNAR” (“do not attempt resuscitation”), arguing it better recognizes that CPR often does not succeed. Recently, some have advocated using “AND” (allow natural death) noting that it focuses attention on improving the dying process, rather than on withdrawing therapies.

Simple DNR orders do not convey information about a patient’s goals for care and views about life-prolonging technology. Resuscitation goals do not always fit into a “yes” or “no” response. Patients may prefer to make specific decisions about multiple therapies, including intubation, vasoactive medication, defibrillation, chest compression, antibiotics, and laboratory testing. A patient might want intubation in the event of respiratory decompensation, for example, but prefer not to undergo chest compression or defibrillation for cardiac arrest. Which interventions make sense depends on patient/surrogate goals and the specifics of the clinical situation.

DNR orders in adults vs. children

In the US competent adults have the legal right to refuse unwanted medical therapies, provided they understand the consequences of refusal. A competent adult can refuse therapies based on: (1) religious or moral views, (2) views about what constitutes a good quality of life, or (3) a determination that a particular therapy is inappropriate for medical or other reasons. Often adults lack decision-making capacity at the time when a decision is required. Anticipating such a possibility, some adults have written or oral instructions, known as “advance directives,” that communicate their wishes should they lose decisional capacity. Alternatively, adults may identify a decision-maker and confer on them a “durable power of attorney for healthcare decisions” for situations when the patient cannot make or communicate their own decisions about care. The law generally expects alternative decision-makers to choose based on what the patient would have wanted. This is called substituted judgment.

Pediatric patients often lack the capacity to make medical decisions, either because of their neurodevelopmental status or their medical conditions. Principles of law and ethics rely on parents to make medical decisions on behalf of their children, based on the best interest of the child.5 The best interest standard requires that the decision-maker determine the net benefit for the patient of each option, and the appropriate course of action is the one with the greatest overall patient benefit. Older children or adolescents have some capacity to make, or assist in making medical decisions. The American Academy of Pediatrics recommends that the pediatric patients contribute to decisions to the extent of their ability.6 (More in-depth discussions of pediatric informed consent and the rights of minors can be found inChapter 5)

DNR in the OR

The Federal Patient Self-Determination Act of 1990 requires hospitals and healthcare organizations to advise patients during every admission of their right to create advance directives and to inquire if the patient has completed one. The law refocused attention on limits of resuscitation and intensive care. Professionals noticed that many advance directives provided inadequate clarity and specificity for clinical decision-making. Ambiguities included whether limits on resuscitation applied in the operating room. Even at end-of-life, surgical interventions may be undertaken to ease continuing care (e.g. placement of tracheostomy or gastrostomy tubes or ports for IV medications), or to surgically palliate acute problems (e.g., relief of pain from a pathologic fracture or nerve root interruption to prevent pain transmission).

Many surgeons and anesthesiologists argue that providing general anesthesia is similar to providing continuous resuscitation. During surgery patients are under continuous close observation that differs fundamentally from care outside the OR. Most patients dying on hospital units die from causes directly related to their underlying conditions, such as cancer, congestive heart failure, and degenerative neurological disease. By contrast, in the OR cardiopulmonary collapse can occur as a direct result of anesthesia or surgical manipulation (e.g. hemorrhage). Moreover, resuscitative efforts under the well-monitored and expectant conditions of the OR have higher success in returning patients to their baseline functioning than do resuscitations outside the OR. Finally, some argue that the OR provides a poor environment for end-of-life care. OR personnel typically do not deal with terminal situations and the OR has no readily available space to accommodate grieving families and clergy.7

In the early 1990s, institutional policies honoring patients’ or surrogates’ DNR wishes, as reflected in physicians’ orders, typically permitted automatic suspension of DNR orders for surgical procedures and during the immediate post-operative period. Such policies did not sufficiently reflect respect for patients’ autonomy. Over time, there has been more general acknowledgment of the rights of patients and their valid surrogates to authorize or reject available treatments, particularly at emotionally and religiously sensitive periods, such as the end of life.8 A shift occurred from paternalistic physician-centered decisions, to shared decision-making between patients or surrogates, and healthcare professionals. Mirroring guidelines from the American Society of Anesthesiologists, the American College of Surgeons, and the Association of OR Nurses, some hospitals have adopted “required reconsideration,” whereby a review of existing directives limiting treatments occurs before a procedure is undertaken. Discussion of perioperative DNR orders should involve the patient or surrogates, the anesthesiologist, the surgeon, and other professionals involved in the patient’s care.9 It should include the following points:

(1) Careful review of the goals of the procedure(s);

(2) Discussion of the meaning of the treatment limits and how proscriptions on interventions might compromise or complicate the anesthetic and/or the operation;

(3) Discussion of the likelihood of successful reversal of any anesthesia- or surgery-related complication;

(4) Agreement on what, if any, limits on resuscitation will remain during the procedure;

(5) Establishment of time boundaries for reinstituting the DNR orders if they are suspended;

(6) Assurance that suspension of treatment limitations will not inhibit sound decision-making – including a decision to withdraw life support – should the patient’s evolving condition alter the expected clinical course.

The anesthesiologist and/or surgeon should provide clear and comprehensive documentation of the agreement in the medical record and ensure that all those involved with the proposed surgery understand and accept the agreement.

Is the cause of cardiac arrest relevant?

Reluctance of OR personnel to accept DNR orders rests to a great extent on the notion that actions by surgeons or anesthesiologists which cause or appear to cause life-threatening events (e.g. hypotension, hypoxemia, hemorrhage, arrhythmia, etc.) differ fundamentally from events that may happen elsewhere. The assumption is that iatrogenic death, or near death, deserves a different response from other events. This claim has some merit in that instability in the OR is often easily reversed with a small likelihood of major adverse effects. But it does not explain why one would not have a similar responsibility to attempt to reverse all potentially fatal events that could be “blamed” on medical interventions. For example, a patient may experience cardiac arrest as a result of severe sepsis, a known complication in immunocompromised patients who have received anti-neoplastic agents. Few would insist on reversing DNR orders for every event whose etiology could possibly be traced to a medical intervention. A patient, parent, or other surrogate might well ask why acceptance of death in one circumstance (on the nursing unit) should not apply in the OR, especially in a setting where the patient is mercifully unconscious and will likely not suffer. The temporal and causal relationship between anesthetic delivery or surgery and death may well affect how physicians and surgeons feel about what happens, but does not necessarily make an important moral or psychological difference to the patient or loved-one who accepts the inevitability of the patient’s death.

This points to potential conflicts in the goals of care for patients and loved-ones, versus those of the professionals. All parties need to clarify these goals and develop practical guides that all parties can accept. In many cases, everyone will agree on preserving life if temporary instability develops in the OR, so that remaining expected weeks or months pass with greater comfort or even function. In others, patients or surrogates may feel that some measures, such as chest compressions, electroshock, or even mechanical ventilation beyond the surgical suite or recovery room imply a level of personal invasion or indignity which is unacceptable, given the patient’s overall condition and prognosis. In those cases, there are few ethical arguments that support the idea that the preference of anesthesiologists or surgeons for life extension should simply override the feelings and beliefs of patients and family members.

Case resolution

The fact that the bowel obstruction occurs so soon after transplantation makes this case especially problematic, because there is likely little reliable information regarding prognosis. Decision making with such uncertainty makes physicians uncomfortable. The oncologists may well favor “aggressive” treatment, hoping the patient will remain free of neuroblastoma and have marrow recovery. The surgeons and anesthesiologists, also hoping for a good medical outcome, may view relief of bowel obstruction as relatively simple and limits on trying to sustain the boy’s life as wrongheaded.

The child’s parents see this picture quite differently. Their son has already endured the rigors of initial cancer surgery and chemotherapy that failed to eliminate the disease. He has also experienced considerable discomfort and distress related to “conditioning” prior to the stem cell transplant, felt the difficulties of post-transplant, pre-engraftment treatment, and faces a very uncertain outcome. Bowel obstruction represents “one more blow” and while they accept an attempt at surgical correction, they feel “heroic” attempts to revive him from a cardiac arrest in the OR would only impose additional burdens on their son and on them.

Extensive conversations between professionals and parents may or may not overcome their concerns about cardio-pulmonary arrest and CPR. They may accept a time-limited reversal of a standing DNR order. Or they may persist in their view that CPR and mechanical ventilation beyond the OR represent unacceptable invasiveness, given the boy’s overall prognosis. Should they refuse to accept a temporary suspension of the DNR order, few ethical theories could justify ignoring their wishes.

A somewhat thornier ethical issue concerns whether professionals have an obligation to proceed with surgery if they voice “conscientious objection” to providing care without authorization to resuscitate. In the absence of finding substitute physicians who could willingly respect the parents’ wishes, simply refusing to give the anesthetic may be regarded as “abandoning” the patient and/or exerting coercive power over the patient/surrogates. Such willful expressions of power over patients raise questions about what it means to accept the mantle of “professional” as one who puts the interests of patients ahead of self-regarding considerations. (For more on physician conscientious objection see Chapter 43.)

Key points

• DNR orders developed in response to the realization that CPR is not appropriate for all patients, particularly those with terminal illness and otherwise dismal prognosis.

• The patient’s/surrogates’ goals should determine the appropriateness of resuscitative interventions in and outside of the OR.

• Pediatric patients may or may not have the capacity to participate in medical decision-making, depending on age and medical condition. They should be involved in decision-making to the degree they are capable. In general, parents function as surrogate decision-makers, acting in the overall “best interest” of the child.

• Automatic suspension of DNR orders in the setting of anesthesia and surgery does not sufficiently recognize patients’ rights to self-determination.

• DNR orders in pediatric patients undergoing anesthesia and surgery, as in adults, should undergo “required reconsideration” in which the involved parties review treatment limitations in light of their benefits and risks, and agree upon and document changes in orders, if any.

• When patients or their surrogate decision-makers, such as parents, do not wish to suspend DNR orders in the setting of surgery, few ethical arguments support ignoring their wishes.


1 The Holy Bible. The New Revised Standard Version. (1989). New York: Oxford University Press.

2 Kouenhoven, W.B.Jude, J.R, and Knickerbocker, G.G. (1960). Closed-chest cardiac massage. JAMA173, 1064–7.

3* Standards for cardiopulmonary resuscitation (CPR) and emergency cardiac care (ECC). (1974). JAMA227(7), Suppl:837–68.

4* Burns, J.P.Edwards, J.Johnson, al. (2003). Do-not-resuscitate order after 25 years. Crit Care Med31(5), 1543–50.

5* American Academy of Pediatrics Committee on Bioethics. (1994). Guidelines on foregoing life-sustaining medical treatment. Pediatrics93(3), 532–6.

6* Committee on Bioethics, American Academy of Pediatrics. (1995). Informed consent, parental permission, and assent in pediatric practice. Pediatrics95(2), 314–17.

7* Ewanchuk, M. and Brindley, P.G. (2006). Perioperative do-not-resuscitate orders – doing ‘nothing’ when ‘something’ can be done. Crit Care10(4), 219.

8* Waisel, D.B.Burns, J.P.Johnson, al. (2002). Guidelines for perioperative do-not-resuscitate policies. J Clin Anesth14(6), 467–73.

9* Fallat, M.E. and Deshpande, J.K. (2004). Do-not-resuscitate orders for pediatric patients who require anesthesia and surgery. Pediatrics114(6),1686–92.