Mrs. C is an 86-year-old woman with metastatic colon cancer, scheduled for surgery for bowel obstruction. She is hypotensive and tachycardic. She agrees to invasive monitoring but does not want to be resuscitated if her heart stops in the operating room. Her surgeon argues not to place invasive monitors because she is a “no code.”
Mr. D is scheduled to be executed for the rape and murder of a child in the state of Missouri. He appeals his sentence on the grounds that lethal injection subjects the prisoner to potential prolonged suffering during the execution process. A court rules that an anesthesiologist must be present to assure unconsciousness before administration of the paralytic agent and potassium. The ruling states that the anesthesiologist must personally mix the drugs and administer them, or directly supervise their administration.
Conflicts concerning patient choices and autonomy are particularly challenging in anesthesiology, in part because anesthesia care routinely alters patient consciousness, interferes with patient competence, and restricts or abolishes physical autonomy. Anesthesiologists are at times expected to use their knowledge and skills for the very purpose of abolishing patient resistance. Such expectations present conflicts with core values in the ethical practice of anesthesiology, and of medicine itself. Navigating the complicated course among ethical principles governing patient choice, fulfillment of beneficent intentions, and preservation of professional integrity requires understanding of ethical values and principles such as respect for patient autonomy, beneficence, non-maleficence, preservation of human dignity, promotion of patient safety, and safeguarding of professional integrity. When patients refuse resuscitation in the operating room (OR), anesthesiologists may experience conflicts between the ethical principle of respect for patient autonomy and choice, and professional imperatives to act beneficently. In the matter of executions, anesthesiologists must answer the question of whether they should engage in acts that superficially resemble medical care but require a personal moral transformation embodying the very antithesis of the medical profession’s philosophies of valuing human life, respecting individuals, and taking moral responsibility for their actions.
Requesting a patient’s permission for a medical procedure implies that the patient can and will sometimes deny that permission. Physicians are obliged to honor informed and competent refusals, lest the “consent” process be devoid of actual choice or autonomy. The tension between “informed consent” and “informed refusal” is especially great when one course of action would likely sustain life, and another would likely lead to death.
Physicians often argue that “benefits” outweigh “harms” of cardiopulmonary resuscitation (CPR) in the OR. In-hospital CPR is associated with low overall survival rates of 10–15% (Brindley et al., 2002; Myrianthefs et al., 2003; Abella et al., 2005) and devastating neurological injuries (Zandbergen et al., 2003). In contrast, survival rates for CPR in the operating room approach 90% (Reis et al., 2002). The difference reflects underlying causes of cardiac arrest; in-hospital arrests are often caused by severe underlying disease. When an arrest occurs in an unmonitored situation, resuscitation may be delayed, contributing to poor outcomes. In the OR, underlying causes of cardiac arrest are often identifiable and reversible, and intervention occurs immediately (Olsson and Hallen, 1988; Sprung et al., 2003).
Allowing a patient to die in the OR from a cardiac arrest that is highly treatable does not appear to many OR physicians to uphold the principle of beneficence. But patients may not believe that resuscitation from cardiac arrest is “beneficial” if they are left with significant physical impairments or are merely revived to die in a short time of a preexisting terminal illness. In weighing the beliefs of the patient and their physicians, physicians are ethically obliged to give priority to patients’ perception of benefits and harms over their own (Clemency and Thompson, 1993, 1994; SUPPORT Principle Investigators, 1995; Wenger et al., 1997).
Some anesthesiologists claim that anesthesia is “ongoing resuscitation,” and that the two cannot be separated from one another. This unfortunate statement implies that anesthesia is riskier than it actually is, and that cardiac arrest is more likely to occur in the OR than other hospital settings, when in fact the opposite is true. Although anesthetic care and CPR share some common techniques, they are almost always easily distinguishable from one another. Would any anesthesiologist seriously argue that a patient undergoing mechanical ventilation during an elective laparoscopy is being “resuscitated?” Or that CPR is a routine part of anesthesia care? Although some life-sustaining procedures, such as assisted ventilation, are necessary during some surgeries, CPR is not integral to surgery, is rarely needed in the OR, and only then to treat rare complications which OR physicians strive to avoid.
Many physicians feel that it may be appropriate to allow a patient with a do-not-resuscitate (DNR) order to die from a terminal disease, but that they are ethically obliged to “rescue” a patient whose cardiac arrest is a consequence of the physician’s actions (Casarett and Ross, 1997; Casarett et al., 1999). Examples include arrests caused by drug reactions, hemorrhage, or arrhythmia provoked by surgical manipulation. Often, however, complications are not attributable to only one action or cause, nor are they usually the result of negligent care. Primary therapies and treatments for the complications of primary therapies both have their origins in the problem for which the patient sought medical care, and it is difficult to reason that the patient may refuse one but not the other (Ross, 2003).
Finally, withholding surgery that a patient desires, such as surgery to relieve a bowel obstruction, unless the patient also agrees to submit to unwanted procedures that are not integral to the surgery, such as CPR, is coercive and, therefore, unethical. As Walker (1991) stated: “Surgery may provide palliative treatment for otherwise untreatable disease. Suspension of DNR orders in the perioperative period places the patients in the unfair position of having to weigh the benefits of palliative treatment against the risks of unwanted resuscitation.”
Founding principles of the medical profession prohibited killing, but those principles must now be reconciled in modern cultures that accept physician participation in pregnancy terminations, physician-assisted suicide, and euthanasia. Arguments favoring physician involvement in executions often cite the principle of respect for “autonomy” by helping a prisoner to have their “desired” mode of death (Baum, 2001; Clark, 2006). Prisoners, however, are among the most vulnerable of society’s constituents. This is reflected in efforts to regulate how the state or medical researchers can treat prisoners. The Nuremburg Code (1949) restricting the use of human subjects in medical experimentation arose out of experiments on prisoners during the second, World War. The United Nations (UN) Third Geneva Convention mandated humane treatment of prisoners of war, and protections against violence, intimidation, public curiosity or insults, torture, or coercion (UN, 1949).
A state of imprisonment is one of severely restricted autonomy. But execution represents the ultimate destruction of autonomy. Prisoners do not usually seek capital punishment of their own free will, and when they do, mental incapacity is cited as a defense against their execution (Blume, 2005). With regard to autonomy and rights, a prisoner’s situation is less like that of an autonomous adult than that of a vulnerable older child, who possesses intellectual capacity, without many legal rights. Arguments seeking to justify physician participation in execution could theoretically cite the ethical value of preserving a prisoner’s dignity, but not respect for non-existent autonomy.
Beneficence-based arguments that physician executioners are needed and best qualified to relieve suffering are based on flawed concepts of “suffering,” as those physical sensations experienced by the prisoner during execution, and a limited concept of beneficence as being measured by only one aspect of the execution process. Capital punishment causes suffering in many persons, including victims’ families, convicts’ families, and prison staff (Osofsky and Osofsky, 2002), none of which is resolvable by medical means. There are also no data to support contentions that competent technicians are less capable of efficient, painless, or “humane” executions than physicians. Forensic examination reveals that all execution methods are fraught with complications, even when physicians are involved. The most frequent problems associated with lethal injection are difficulties obtaining vascular access, and painful subcutaneous infiltration of medications, delaying onset of unconsciousness (Khan and Leventhal, 2002). A recent study suggests that blood levels of hypnotics administered during lethal injections are inadequate to assure unconsciousness in almost half of cases (Koniaris et al., 2005). There is no evidence that these problems would be significantly lessened in the hands of physician–executioners, when complications with intravenous access and awareness during anesthesia also occur during the course of routine medical care.
Arguments invoking the principle of beneficence to justify physician involvement in criminal executions have historically been associated with subsequent “slippery slope” justifications for physician involvement in the killing of persons who have never faced an accuser or had a fair hearing, including those with physically or mentally handicaps or other social “flaws” (Hinman, 1944; Jonsen, 1993; Pelligrino and Thomasma, 2000). If physicians accept a role in executions based on beneficence arguments, it becomes harder to draw the line at participation in other state-sponsored activities, such as torture, coercion, and “medical incarceration,” because they too are often defended as being “beneficial” to society (Silver, 1986; van Es, 1992; Pelligrino, 1993; Pelligrino and Thomasma, 2000).
Physician participation in executions produces many harms and is, therefore, not “non-maleficent.” It causes harm through the “medicalization” of a non-medical and distasteful act in order to defuse moral objections and render it more acceptable to an increasingly skeptical public. In this regard, it parallels historical misuse of psychiatric diagnoses and mental illness-based incarcerations to manage non-medical social or political problems (Rood, 1979; Gluzman, 1991; Adler et al., 1992). Agreeing to participate in executions transforms the physician into a deceptive “double agent,” who is acting on behalf of the state while appearing to act on behalf of the “patient” (Silver, 1986; van Es, 1992; Pelligrino, 1993). It erodes public respect and trust (Sikora and Fleischman, 1999). It also sometimes undoubtedly engages the physician in the killing of innocent persons (Hinman, 1944; Dieter, 2004; Gross et al., 2005).
Physicians have typically tried to divorce the issue of physician participation in executions from the question of whether capital punishment itself is moral. But the morality of engaging in an activity simply cannot be completely separated from moral aspects of the activity itself (Thorburn, 1987; Hastings Center, 1996). The practice of medicine involves consideration of the principle of justice, and the extent to which capital punishment intrinsically fulfills that principle is relevant to whether physician participation in executions is consistent with professional integrity.
DNA technology has proven that innocent people are wrongly convicted of capital crimes and sentenced to die, and that innocent people almost certainly have actually been executed (Langer, 2001; Dieter, 2004; Gross et al., 2005). Capital punishment for comparable crimes is applied unequally across racial and socioeconomic groups (Baldus and Woodworth, 1997). Studies consistently demonstrate that it does not deter violent crime (Sorensen et al., 1999; US Department of Justice, 2003; Rosenfeld, 2004; Berk, 2005). It unfairly consumes economic resources, because it is much more expensive than lifelong incarceration (Cook et al., 1992; Dieter, 1992; Forsberg, 2005). It fuels ethically objectionable proposals to curtail existing “safeguards” – the appeals process – in order to cut costs (US Senate, 2005). It does not appear to provide the closure that victims’ families seek (Lithwick, 2006; Schieber, 2006).
There is disturbing evidence that executioners undergo a process of “moral disengagement,” or a kind of moral degradation (Osofsky et al., 2005). Executioners avoid self-condemnation by dehumanizing the convict, devaluing his or her life, and deflecting personal moral responsibility away from themselves by blaming juries, judges, governors, and “the law” for the prisoner’s execution, and not themselves.
The courts unambiguously support the rights of competent patients to refuse life-sustaining interventions and to have treatment refusals honored while they are unconscious (In the Matter of Quinlan, 1976; Barber v. Superior Court, 1983; Cruzan v. Director MDH, 1990). Because anesthetic care usually interferes with a patient’s ability to make and express decisions, courts have found that anesthesiologists have legal obligations to protect the patient in the OR from unwanted intrusions (Schloendorff v. Society of New York Hospital, 1914; Kroll, 1992). Recently, physicians have been found liable when an unwanted resuscitation resulted in survival but significant morbidity (Anderson v. St. Francis–St. George Hospital, Inc., 1996; Osgood v. Genesys Regional Medical Center, 1997).
In the USA, persons who participate in legally sanctioned executions are protected from criminal charges or civil penalties. In almost all cases, anonymity is promised to participants, making it difficult to know exactly how many executions are carried out with the help of physicians. Although lethal injection was developed as a “humane” method of execution, reports of complications and undue suffering of prisoners has called the constitutionality of lethal injection into question. Several state courts have recently ruled that lethal injection must be carried out under the supervision of a physician. In the state of Missouri, a court has ruled that an anesthesiologist must mix and personally administer the drugs, or supervise their administration (Michael Anthony Taylor v. Larry Crawford et al., 2006). Physicians are not legally compelled, however, to participate in the execution of prisoners.
Automatic suspension of DNR orders in the OR does not appropriately recognize patient rights to refuse medical therapies during the perioperative period. Guidelines established by the American Society of Anesthesiologists (1993), the American College of Surgeons (1994), the Association of Operating Room Nurses (Murphy, 1993), and the Joint Commission on Accreditation of Health Care Organizations (1996) require that DNR orders be rediscussed in the setting of surgery and anesthesia and state that automatic suspension of DNR orders in the perioperative setting is unethical. Discussion of DNR orders in patients undergoing anesthesia and surgery should include an explanation of the risks and benefits, including the more favorable outcomes for CPR in the OR, and the medical staff should document either the patient’s goals for treatment or specific treatments the patient refuses or accepts.
In the USA, physician organizations have consistently held that physician participation in executions is unethical (American Medical Association, 2000; American Psychiatric Association, 2003). However, there are no reported cases of disciplinary action against a physician or expulsion from a professional society for such involvement. In part, anonymity provided to executioners prevents many professional organizations from even being able to identify which, or how many, of their members aid in executions.
Physician involvement in euthanasia and executions concerns anesthesiologists in particular; to the uninformed, their skills appear to make them ideal candidates for duties that involve killing (Jonsen, 1993; Truog and Berde, 1993). The American Society of Anesthesiologists (ASA) has, therefore, addressed the issue in several ways. The ethical guidelines of the society specifically support those of the American Medical Association, which prohibit physician involvement in executions (American Medical Association, 2000; American Society of Anesthesiologists, 2003). A statement by the President of the American Society of Anesthesiologists in 2006 (Guidry, 2006) concluded that, “physicians should not participate in executions, either by direct action or by performing ancillary functions. This includes making recommendations about drugs to be used. Physicians are healers, not executioners. The doctor–patient relationship depends upon the inviolate principle that a doctor uses his or her medical expertise only for the benefit of patients.”
The favorable prognosis for resuscitation in the OR obligates anesthesiologists to revisit DNR orders with patients in the perioperative period, because the patient may make a different decision in the specific setting of surgery. Presenting a patient with an exhaustive “consent check list” of possible invasive or resuscitation measures, however, can be intimidating, confusing, and coercive at a time of significant stress. Some authors have, therefore, suggested a “goals-directed” approach to DNR in the OR that focuses on patient desires regarding outcomes rather than on specific techniques to try to meet those desires (Jackson and Van Norman, 1999; Truog et al., 1999). The patient may agree to have obvious, reversible problems addressed in a medically appropriate fashion but wish withdrawal of life support after surgery if other complications arise.
There is probably no single “best” way of approaching the patients with DNR orders who are scheduled for surgery. The process of informed consent and refusal depends on individual patient–provider relationships and conversations (Jackson and Van Norman, 1999). By the same token, DNR orders do not constitute permission to stop “caring” for the patient in other ways. Invasive monitoring, for example, may help the anesthesiologist to prevent the cardiac arrest that cannot subsequently be treated.
Current ethical standards of most medical professional organizations either explicitly prohibit or discourage the involvement of physicians in the execution process, although such standards are often vague as to what constitutes “participation.” Most organizations appear to define participation as being present during executions, directing medical procedures involved in the execution, and prescribing further measures when the prisoner does not immediately die. The appropriateness of physician participation in declaring death (which may indirectly lead the physician to “prescribe” further measures if death has not resulted), or formulating methods of execution or recipes for lethal injection, remains controversial. In recent statements from the American Society of Anesthesiologists, however, involvement in such “ancillary” aspects of lethal injection is also discouraged (Guidry, 2006).
Although it is tempting to believe that physician participation in executions is somehow “merciful” or “beneficial,” such activities nevertheless cause tremendous harms, both to persons involved and to the medical profession at large. Physicians should, therefore, not participate in executions.
The first case considers a do not resuscitate order in the operating room. Mrs. C was competent to refuse life-saving therapy. She was at moderate risk for cardiac arrest in the OR. It became obvious through discussion that she hoped to survive her surgery, obtain relief from pain caused by the bowel obstruction, and rejoin her family. She wished to avoid prolonged mechanical ventilation, particularly if things appeared “hopeless.” She did not see any purpose in resuscitation if her heart stopped beating during surgery, but expected medications to be administered to “prevent it from stopping,” if possible. Despite her surgeon’s objections, the anesthesiologist placed invasive monitoring lines to facilitate hemodynamic management and reduce the risks of cardiac arrest in the OR. He also planned placement of the patient in the intensive care unit postoperatively. During the surgery, dopamine was initiated to support blood pressure and cardiac output. She remained mechanically ventilated postoperatively. Dopamine was gradually weaned and she was extubated on the third postoperative day. After 10 days in the hospital, she was discharged to a nursing facility, and returned home to her family two weeks later. She died at home after four months from the effects of metastatic cancer.
The second case deals with an execution. Despite a court order that lethal injection could only be carried out if a physician was present to supervise the execution, no physician could be found who agreed to participate. In February of 2006, executions by lethal injection in the state of Missouri were placed on an indeterminate “stay” until the issues could be resolved. Mr. D remains on death row awaiting execution.