Clinical Ethics in Anesthesiology. A Case-Based Textbook

1.Consent and refusal

2.Informed refusal - DNR orders in the patient undergoing anesthesia and surgery and at the end-of-life

David M. Rothenberg

The Case

A 67-year-old male with oxygen-dependent COPD requires a series of electroconvulsive therapies (ECT) for severe depression refractory to medical therapy. The patient has consented to general anesthesia with mask ventilation for these procedures, but refuses endotracheal intubation irrespective if it might be life-sustaining in the event of respiratory or circulatory arrest. As such, he requests that a do-not-resuscitate (DNR) order be enacted and remain in place during the perioperative period and throughout the entirety of his hospitalization.

In 1960, cardiopulmonary resuscitation (CPR) was introduced to clinical practice and was intended to apply to those patients in whom their hearts were considered “too good to die.” Unfortunately, the more liberal use of CPR in patients with hearts essentially “too poor to live” prompted ethical consideration of when CPR should be withheld, thus averting the situation whereby death, rather than life, was delayed.

The DNR order is founded on the ethical principle of respect for individual autonomy and the legal doctrine of informed consent. Autonomy to make one’s own decisions regarding medical therapy is not a modern day concept. Indeed, John Stuart Mill proposed this idea in his 1869 treatise On Liberty in which he stated, “Considerations to aid his judgment, exhortations to strengthen his will, may be offered to him, even obtruded on him, by others; but he himself is the final judge. All errors which he is likely to commit against advice and warning, are far outweighed by the evil of allowing others to constrain him to what they deem his good.”1

The doctrine of informed consent was defined by Justice Benjamin Cardozo in the case of Schloendorff v Society of N.Y. Hospital.2 In this case a young woman, who consented only to a gynecological examination under ether anesthesia, was instead furtively subjected to uterine surgery. Justice Cardozo of the New York Court of Appeals described the wrong committed as a “trespass,” emphatically stating “Every human being of adult years and sound mind has a right to determine what shall be done with his own body.” A competent patient’s refusal of treatment, including the request to refuse potentially life-sustaining therapy, must be honored. However, individual autonomy is not a legal or moral absolute and can be superceded by the state’s rights to protect innocent third parties (e.g., minors), prevent suicide, and maintain the integrity of the medical profession. Concerns about maintaining the integrity of the medical profession and the individual integrity of physicians are often invoked in discussions of futility as it applies to a patient’s DNR order.

The theory of medical futility originated in the ancient philosophical teachings of Plato and was later espoused by Hippocrates, who stated that, in general, the role of medicine “is to do away with the sufferings of the sick, to lessen the violence of the diseases, and to refuse to treat those who are over-mastered by their diseases, realizing that in such cases medicine is powerless.”3

The Council on Ethical and Judicial Affairs of the American Medical Association simply affirms Hippocratic dogma in stating: “When efforts to resuscitate a patient are judged by the treating physician to be futile, even if previously requested by the patient, CPR may be withheld.”4Although the predominance of medical opinion continues to be that physicians are not and should not be required to provide futile treatment, defining futility may be a more difficult and critical task. It has been implied that medical futility should apply when there is an absence of at least a modicum of medical benefit or, in the setting of cardiac arrest, when the patient predictably has near 0% survival after CPR. For patients with diseases such as metastatic cancer, acute stroke, sepsis, or multilobar pneumonia, CPR should probably not be offered in the event of a cardiac arrest. Schneiderman et al. suggested that, in terms of probability, futile therapy be defined as a chance of success being less than 1 in 100.5 Although isolated cases of long-term survival in these instances may exist, if these occurrences are so rare as to be unexplainable and therefore have little predictive value, then this definition of futility would still apply.

In discussing futility, it is imperative to differentiate “physiologic” or “quantitative” definitions of futility – based on scientific data that suggest that CPR will not be effective – from more value-laden or “qualitative” definitions of futility, in which physicians may impart their own beliefs on what they perceive as the patient’s preference, and suggest in this context that CPR is not worth attempting.

Within the framework of “physiologic” futility, a DNR order should be written and CPR should be withheld when an arrest occurs. Patients and/or their families should be informed of this decision, although to suggest that a choice exists could imply that there may be a potential benefit from CPR, inadvertently sending a mixed message. Respect for patient autonomy does not mean that patients have a right to demand nonbeneficial and potentially harmful treatment. Physicians who agree to such demands act purely on behalf of the patient’s psychological welfare rather than on the patient’s rights. Brody reiterates “that there are some questions of professional ethics that physicians are entitled to decide among themselves unilaterally. Regarding those matters, the society is entitled to decree that the profession of medicine will not be practiced in that way … But the society is not entitled to dictate to the profession the contents of professional integrity.”6 In protecting the integrity of the medical profession, however, public trust must not be compromised. DNR policies founded on the premise of cost containment tend to be value laden and, if implemented, are likely to foster a paternalistic approach to patient care, as well as to erode patient confidence that physicians are acting with their best interests in mind.

Guidelines for establishing standards of medical futility have been suggested in an effort to create an open dialogue with society, in general, and the medical community. These guidelines include (1) acknowledging that the word futility is widely used and (2) defining futility as a treatment that fails to achieve the goals of medicine. Social and medical acceptance of futility as it pertains to CPR can then hopefully be attained by using this definition in formal outcome studies.

DNR orders may be written based on physician assessment of futility, or based on patient request. The majority of patients who request DNR orders do so not because they wish to die, but rather as a reflection of how they perceive their quality of life before or after the resuscitation. Diminished quality of life has been defined as diminished capacity to resume work and/or impaired cognitive function specifically as it applies to awareness of one’s environment. It has been shown that, in patients who survive critical illnesses, quality of life is worse if the patient requires CPR. Physical or mental impairment, chronic disability, or persistent vegetative states are additional reasons patients or their surrogates may request a DNR order. Patients also express desires for DNR orders when they believe that CPR will merely prolong the dying process. Irrespective of the reasons behind the decision, the individual patient’s values are more relevant than those of the physician in determining resuscitation status when quality of life is the issue in question.

Physician insecurity in applying a DNR order has sometimes led to the ethically and medically inappropriate practice of the “slow-code” or “show code.” Additionally, a DNR order should not be misinterpreted as an order to withhold or withdraw other vital therapies. Indeed, the misconception of this order to mean “do-not-round” has led to the suggestion that the order be renamed no-CPR, or do-not-attempt resuscitation, to convey the idea that even if CPR were to be initiated, the likelihood of success would be extremely low. A DNR order does not preclude the use of other life support therapies including, in selected instances, surgery or intensive care. Most recently a call for renaming these orders to “allow natural death” (AND) has been suggested in order to eschew the negative connotation of do-not-resuscitate.7 The problem with this term is that it may inappropriately lead to the withholding of all other forms of therapy exclusive of CPR.

Our patient is justified in requesting a DNR order in advance of ECT as he perceives his life to be insufferable should he require prolonged mechanical ventilatory support. An argument for a physician-initiated DNR or AND order based on medical futility does not apply because of a lack of scientific evidence citing survivability if this type of patient suffers a respiratory or cardiac arrest.

Combinations of both medical and social factors have led to the more frequent application of the DNR order both in a pre-hospital and in-hospital setting. An aging population with end-stage malignancies, stroke syndromes, multiple organ failure, and Alzheimer’s dementia contributes to the growing use of the DNR order. They are also major social factors related to the enactment of patient’s Self-Determination Act of 1990,8 which mandated that all hospitals receiving Medicare reimbursements provide patients with information that would aid in their decisions regarding end-of-life issues such as a DNR order. As DNR orders continue to be written, it is inevitable that new ethical dilemmas will arise. Such dilemmas are increasingly common in the pre-hospital and pre-surgical settings.

Pre-hospital DNR orders

Pre-hospital DNR policies have emerged recently and serve three primary purposes: (1) to provide continued respect for patient autonomy following hospital discharge, (2) to prevent futile resuscitation efforts in the field, and (3) to protect the well-being of emergency medical service (EMS) personnel. Patients who maintain a DNR order following discharge from the hospital should have this order honored in a pre-hospital setting should an arrest occur at home or at a chronic care facility.

Attempts to clarify whether a patient has a pre-existing DNR order based on quality of life in an out-of-hospital setting may be difficult and time consuming. Therefore recommendations have been made to identify such patients preemptively, and include a DNR identification bracelet or “no-code” tattoo, or a central, computerized file system allowing EMS personnel to access and verify resuscitation status at the scene. In the event that the patient’s wishes are unclear, full resuscitative measures often are instituted by medical rescue technicians irrespective of the probability of successful resuscitation. Emergency department physicians and ethicists have thus considered the need for a pre-hospital DNR policy based on medical futility. As of 1991 when the US Federal Patient Self-Determination Act was mandated, Montana was the only US state to have a policy that provided civil and criminal immunity for EMS technicians who terminated resuscitative efforts in good faith in the field. At that time, other states restricted the use of pre-hospital DNR orders to only hospice programs, patients being transferred between institutions or terminally ill patients. Difficulty was felt to arise in establishing the validity of an advance directive or the competency of a conscious patient who refused such therapies as endotracheal intubation or intravenous vasopressors.

As of 2002, however, 42 US states had enacted out-of-hospital DNR protocols based on scientifically validated studies assessing the futility of CPR in the field. Futility in a pre-hospital setting needs to be determined completely on a physiologic basis (as was previously discussed), and without value-laden quality of life judgments. Outcome data in these multiple, large studies indicate a near 0% survival if there is no return of spontaneous circulation in the field, especially in the setting of blunt trauma.9 Although studies of these types seem to validate the definition of physiologic futility as it relates to CPR in the field, application of “termination of resuscitation rules” based solely on futility are not absolute. Significant variability regarding the timing of termination of resuscitative efforts and the pronouncement of death has been noted despite having defined rules, especially in large cities where many different base hospitals provide on-line medical control of EMS personnel during out-of-hospital cardiac arrest.

Finally, concern regarding the health of both EMS personnel and innocent bystanders during high-speed ambulance transports, the increase of risks of HIV and hepatitis B exposure during resuscitation, economic constraints in terms of use of limited resources (especially in small communities), and costs of futile resuscitation have all been offered as additional reasons to establish pre-hospital DNR policies. Although fear of medical liability, in concert with the public’s expectations that EMS personnel will initiate and continue CPR, may have limited the use of pre-hospital DNR orders in the past, federal and/or widespread state adoption of such preset rules of out-of hospital termination of basic and advanced cardiac life support appear imminent.

Pre-anesthesia DNR orders

Patients with preexisting DNR orders often require anesthesia for surgical procedures necessitated by the need to improve quality of life. Common surgical procedures that are often performed as a means of palliation or comfort of care in patients with DNR orders include the following:

• Tracheostomy

• Indwelling central line access

• Feeding gastrostomy/jejunostomy

• Emergency surgery (hemorrhage, bowel obstruction, peritonitis.

In addition, it is not uncommon for patients with DNR orders to require an anesthetic for nonsurgical procedures such as electroconvulsive therapy as described for our patient. Given that the practice of anesthesiology is inseparable from the process of intraoperative resuscitation, an ethical dilemma may occur when a patient with a DNR order presents for surgery. In this setting automatic resuscitation may violate patient autonomy and give rise to a situation directly prohibited by the DNR order, whereas automatic withholding of life-support may mean ignoring a reversible complication, thus violating the spirit of patient’s wishes.

Although arguments have been made in favor of maintaining the DNR order at the time of surgery, the majority of medical and ethical opinion favors a policy that has been best termed “required reconsideration.”

In keeping with the philosophy of Mills, the anesthesiologist should stress to the patient that to have the greatest likelihood of surgical success, simple intraoperative interventions and/or resuscitation may be necessary. These may include pharmacologic reversal of residual anesthetic agents (e.g., naloxone for the respiratory depressant effects of opiates, or flumazenil for similar effects of benzodiazepines), antiarrhythmic therapy for iatrogenically induced arrhythmias (e.g., ventricular tachycardia secondary to either an accidental intravascular injection of a local anesthetic or direct myocardial irritation from a central venous catheter placement), or correction of hypoxemia due to accidental esophageal or unilateral bronchial intubation. The contention that these mishaps are more likely to result in an intraoperative arrest in a patient with a preexisting DNR order has not been substantiated. Indeed patients should understand that survival following resuscitation, specifically CPR, has been shown to be considerably improved if the arrest occurs in an operating room versus the general hospital ward. With these “exhortations,” it is hoped that most patients or their surrogates would opt to have the surgery performed with the DNR order temporarily suspended. Patients should be reassured that resuscitation measures would be employed only if necessary, and any form of life support would be discontinued if no longer required or no longer beneficial in reversing the patient’s underlying condition. This approach also allows distinguishing between, and treatment of those arrests that are readily reversible and those in which resuscitation is physiologically futile.

Within the context of this “informed consent” discussion, the patient or surrogate is entitled to certain options as they relate to the pre-anesthesia DNR order. The patient may: agree to undergo the surgery with the DNR order temporarily suspended; undergo surgery but may stipulate that only specific resuscitative measures be used (e.g., vasopressors but no CPR or mask ventilation but no endotracheal intubation); refuse surgery unless the DNR order is maintained throughout the procedure, thus accepting the risk of potentially reversible causes of death. If the patient demands the last option, both surgeon (or in our patient’s case, psychiatrist) and anesthesiologist must decide if it is proper to proceed without the ability to institute life-support measures should a complication occur. Although physicians are not obligated on moral or legal grounds to operate under these constraints, they do have a duty to try to secure other physicians to care for the patient, so as to avoid potential charges of patient abandonment.

The American Society of Anesthesiologists10 and the American College of Surgeons11 have drafted guidelines for the management of the patient with a presurgical DNR order. Although both societies have adopted a policy of “required reconsideration,” neither offers recommendations for either postoperative resumption of the DNR order or discontinuation of life support should intraoperative CPR be necessary. Recommendations for weaning or terminating life support in this setting have been proposed by Franklin and Rothenberg.12 Should resuscitation be required in the operating room, necessitating life support either in the form of mechanical ventilation, inotropic agents or vasopressors, then postoperatively, the patient should be transferred to an appropriate care area such as the intensive care unit, and an attempt to wean the patient from support should commence. This type of management should probably not take place in a post-anesthesia care unit where the discussion and enactment of possible termination of life support may be distressing to other recovering patients. Weaning should ideally be achieved within a reasonably short period (i.e., 24 hours) should the etiology of the arrest be iatrogenic or unrelated to the patient’s underlying disease. If successful, the DNR order may be reinstated and the patient transferred to an appropriate care area.

Should the weaning process fail, then provisions for terminating life support should be considered. Prior preoperative discussions in this regard are essential so that patients and their families may be assuaged of their fears of being left on life support systems indefinitely. Appropriate medical consultation to ensure that the patient’s condition is irreversible and to support the decision to withdraw support is encouraged. Constant communication with the patient and/or surrogates and documentation of these discussions in the medical record is also mandatory.

The actual termination of life support should be done in accordance with well-established guidelines and be conducted with the sole purpose of minimizing patient suffering.

Case resolution

Our patient agreed to undergo ECT and suspend his DNR order during the time surrounding his treatment and the immediate recovery period with the stipulation that should he require endotracheal intubation and mechanical ventilation, these supportive modalities would be discontinued in an expeditious fashion (24 hours). Appropriate sedative/analgesic therapies would be instituted if tracheal extubation would likely lead to respiratory and subsequent circulatory arrest, assuring comfort at the end-of-life and in keeping with the principle of the “double effect”. The patient was able to successfully complete a series of eight treatments and was discharged from the hospital with a DNR to be maintained at his chronic care facility.

Key points

• DNR orders may be written if CPR would be physiologically futile, or at the request of patients who feel that CPR would result in poorer quality of life.

• Pre-hospital DNR orders to prevent futile or unwanted CPR in the field are becoming increasingly common and are likely to be widely inacted.

• DNR orders in the setting of anesthesia and surgery should undergo “required consideration” in which the risks and benefits of resuscitation in the OR are reviewed and patient desires determined.

• When a patient desires that a DNR order be maintained during anesthesia and surgery, providers who feel they cannot proceed nevertheless have an obligation to try to secure alternative physicians who are willing to respect the DNR order.

• Plans should be formulated for postoperative resumption of the DNR order, if suspended, or for discontinuation of life support after intraoperative CPR if appropriate.


1* Mills, J.S. (1978) On Liberty. Indianapolis, IN; Hackett Publishing, p. 75.

2 Schloendorff v Society of NY Hospital, 105 N.E. 92, 1914.

3* Chadwick JMann WN (eds). The Medical Works of Hippocrates. (1950). Boston, MA: Blackwell Scientific.

4* Council on Ethical and Judicial Affairs, American Medical Association (1991). Guidelines for the appropriate use of do-not-resuscitate orders. JAMA265, 1868–71.

5* Schneiderman, L.J.Jecker, N.S., and Jonsen, A.R. (1990). Medical futility: Its meaning and ethical complications. Ann Intern Med112:949–53.

6* Brody, H. (1994). The physician’s role in determining futility. J Am Geriatr Soc42, 875–8.

7* Venneman, S.S.Narnor-Harris, P.Perish, M., and Hamilton, M. (2008). “Allow natural death” versus “do not resuscitate”: three words that can change a life. J Med Ethic34, 2–6.

8 Omnibus Budget Reconciliation Act of 1990, Public Law 101–508, Section 4751.

9* Vayrynen, T.Kuisma, M.Maatta, al. (2007). Medical futility in asystolic out-of-hospital cardiac arrest. Acta Anaesthesiol Scand52, 81–6.

10* Committee on Ethics, American Society of Anesthesiologists: Ethical guidelines for the anesthesia care of patients with do not resuscitate orders or other directives that limit treatment. American Society of Anesthesiologist Directory Members, Park Ridge, IL, ASA, 1994, pp. 746–7.

11 Committee on Ethics. (1994). American College of Surgeons: Statement on advance directives by patients: Do not resuscitate in the operating room. Am Coll Surg Bull79, 29.

12* Franklin, C.M. and Rothenberg, D.M. (1992). Do-not-resuscitate orders in the presurgical patient. J Clin Anesth4,181–4.

Further reading

Bishop, J.PBrothers, K.B.Perry, J.E. and Ahmad, A. (2010). Reviving the conversation around CPR/DNR. Am J Bioeth10(1), 61–7.