Informed consent and the aftermath of cardiopulmonary resuscitation: Ethical considerations

Prototypical problem

The following narrative prototype is a composite of multiple, actual, and frequently encountered case scenarios in the intensive care unit (ICU). It illustrates the problems about which the authors are concerned—that informed consent in CPR decision-making too often is not full and informed; that it does not better, and in advance, educate elderly patients and those with serious medical co-morbidities about CPR survival data, post-CPR functionality, and most likely CPR outcomes; that deficient communication between providers, patients, and families about the practices and outcomes of CPR leads to unnecessary physical and emotional suffering; and that delegation of responsibilities for informed consent is sometimes ad hoc or inappropriate, that is, assigned or defaulted to the patient’s nurse. Moreover, in the United States, laws, rules, and policies for advance directives including CPR decisions vary from state to state and even from practice site to practice site. In most cases, an advance directive will be followed in a critical situation, but even that cannot be guaranteed when the patient is incapacitated and family members disagree with the directive. In the case where an advance directive is unavailable or does not exist, a patient is incapacitated, and no healthcare proxy has been identified, US physicians generally order code status in consultation with available family members who are best positioned to have knowledge of the patient’s preferences and critical interests. Again, despite prognosis, presumed consent to CPR is the default position without an advance directive or designated healthcare proxy. ⁵

The case of Mrs M illustrates most of these concerns. Mrs M, an 83-year-old, oxygen-dependent patient with serious co-morbidities including rectal cancer and severe chronic obstructive pulmonary disease (COPD), was found conscious but confused on the floor of her home by a family member. An ambulance was called, and she was admitted in severe respiratory distress to the emergency department of a local hospital in a small to medium-sized, Midwestern, US city. Mrs M’s nasal cannula had fallen off when she fell, and she had probably been without her oxygen for approximately 6 h. In the emergency department, the patient was placed on BiPap to assist respiration; subsequently, she was admitted to the ICU. With no available advance directive, the hospitalist ordered a “full code” after a brief discussion with the frightened and still mildly confused patient about whether she wanted to be “brought back” if her heart stopped. No other discussion took place with either the patient or her family about CPR procedures or possible complications.

Later that evening, Mrs M required intubation, her condition deteriorated, and she became unresponsive and incapable of participating in any further medical decisions. The physician asked the ICU nurse to talk with the family and attempt to confirm the patient’s prior end-of-life wishes. Without any further communication with the physician or any more information about CPR procedures, survival data, or possible post-CPR complications and consequences, the family asserted that resuscitation was what their mother in the past had always said she wanted. The order for full code status remained in place. Later that night, the already ventilated patient suffered a cardiac arrest. An automated external defibrillator (AED), chest compressions, and cardiac medications were applied in full resuscitative effort. Although successful, meaning that a pulse was restored, the client suffered rib fractures and lacerations to her liver. She remained unresponsive and died less than 24 h later after a second cardiac arrest. The family was unprepared for the aggressiveness of the CPR procedure and the devastating post-CPR complications. Mrs M’s suffering at end of life was a source of guilt and moral distress for both the family and their mother’s care providers.

CPR survival rates

Although CPR survival data may not be entirely reliable, investigators have been building the database since at least the late 1980s. An early meta-analysis of literature on survival rates of inpatient CPR used two sets of inclusion criteria—minimal criteria (any adult who received CPR in the hospital) and strict criteria (patients from general medical units and ICUs who received fully sufficient CPR and resuscitation efforts). It documented an overall survival rate of 40.7% and a survival to discharge rate of less than 15%. ¹³ Variables that hindered survival included sepsis on the day before CPR, metastatic and other cancer, dementia, African American heritage, serum creatinine level at 1.5 mg/dL, coronary artery disease, and location of CPR in the ICU. The likelihood of surviving to discharge was only about 1 out of 8 for patients who received CPR and 1 out of 3 for patients whose CPR was successful. ¹³

During the 1990s, other reports of survival to discharge rates following in-hospital cardiac arrest and CPR ranged from 7% to 26%. ¹⁴ By 2014, US, in-hospital, cardiac arrest data showed an overall survival rate to hospital discharge of 23.9% for adult victims of cardiac arrest compared to a rate of 18.4% for similar patients in the United Kingdom. ¹⁴ Advancing age significantly impacted the chance of survival to hospital discharge, reducing survival to discharge rates to a range of 11.6%–18.7% for patients over age 70. ^3,14 CPR was more likely to be withheld as a matter of age independent of co-morbidities. ¹⁵ Even when all else was equal, clinicians were nine times more likely to withhold CPR in a 90-year-old patient than a 60-year-old patient. ¹⁵

At any age, survival data are important for full and informed consent in CPR decision-making, but the data are especially important for patients whose chances for uncomplicated survival dim with advancing age. Early on, in 1994, Murphy et al. ¹⁶ interviewed ambulatory patients at least 60 years of age about their CPR preferences. Before they knew that their chances of surviving to discharge ranged from 10% to 17%, 41% of patients said that they would choose CPR in the event of cardiac arrest. After they were given accurate and more detailed information about survival rates, patients’ rate of CPR election dropped to 22%. Later, in 2006, Addams and Snedden ¹² also explored the over-estimation of CPR efficacy among 1000 patients aged 70 years and older and found that 81% of the participants believed their chances of survival were 50% or greater; 23% thought they had a 90% chance or better.

More recently, in 2014, Van Gijn et al. ³ took another look at survival data and examined the probability of post-CPR survival in the inpatient setting after age 70. Results showed that over half of patients who survived CPR died prior to discharge—a survival to discharge rate of 11.6%–18.7%. ³ Survival rates declined with advancing age, revealing a survival to discharge rate of 18.7% for patients aged 70–80 years, approximately 15% for patients aged 80–90 years, and only 11.6% for patients over the age of 90 years. ³ The view of CPR survival from the patient’s perspective was overly optimistic, and the inherent uncertainty of the medical endeavor was highlighted by the finding that physicians had just as much success predicting CPR outcomes with random guessing as they did with review of actual patient cases. ³ Moreover, patients and families often thought that a code status of DNAR meant that no effort would be made to promote quality of life or that loved ones would be left to die with no interventions for pain or hygiene. ³ Across time, over at least two decades, findings are remarkably consistent with respect to the over-estimation of CPR success.

Elderly patients understand discussions about prognosis, and improved communication with them is sorely needed to correct gross misperceptions of CPR outcomes. ¹² Patients and families will fill an information gap with assumptions, expectations, hopes, fears, media narratives, cultural memes, and whatever disparate information might be available to them; but they are better served by carefully conveyed, detailed communication and recommendations in the context of an empathic relationship, based on known survival data and best evidence for functional outcomes post-CPR. ³ Misperceptions of CPR and overly optimistic expectations about CPR outcomes require honest discussion about its brutality, burdens, and limitations. Honest, compassionate, and more detailed discussion will facilitate more informed patient decisions about code status, best interests, and end-of-life options. ¹⁶

Medical futility and the goals of care

Providers have ethical obligations to patients to exercise competence in upholding appropriate standards of care in a context of veracity (truthfulness), fidelity (faithfulness), beneficence (good works), and reliable efforts to advocate for patients’ autonomy (right to self-determination) in CPR and other medical decision-making. ¹⁷ Providers also have ethical responsibilities to facilitate patients’ narratives of identity, relationships, and values in discussions of who they are and what is important to them. ¹⁰ Such narratives weave the threads of patients’ lives—past, present, and future—into whole cloth, which is healing in and of itself and gives patients the opportunity to clarify for themselves how their stories ought to end. While a healthcare provider may have an obvious obligation to aid the sick, he or she has no obligation to propose or advocate for treatment that is futile. ^18,19

Given the value-laden nature of medical decision-making and the inherent uncertainty in using group-based statistics to predict outcomes for a particular person, ⁵ the concept of medical futility defies definition; nevertheless, it can be described in very simple terms as intervention that is unlikely to generate any significant benefit for the patient. ¹⁸ In the current context, medical futility can simply mean “an unacceptable survival rate after CPR.” ²⁰ While medical futility emerged as a descriptive concept that could help providers, patients, and families more easily make treatment decisions at end of life, the concept contains assumptions about the value of continued treatment ²¹ and can pit medical judgments about feasibility and futility against a patient’s right to opt for the possibility of longer life irrespective of its quality. ⁵

In cases characterized by questions of medical futility, and with the patient’s life hanging in the balance, issues of quality of life assume pre-eminent importance. For example, quality of life emerged as the central value underlying the choice for resuscitation in a focus group study of older adults aged 60 and above, which showed that CPR would be elected only if the person had an acceptable quality of life before CPR or reasonably could be expected to have an acceptable quality of life after CPR. ²² Quality of life was determined by (a) medical condition and how much more good life an individual could expect based on diagnosis/prognosis; (b) physical versus mental disability, as mental incapacity, which was associated with communicative inability and cognitive decline, was seen as more destructive of quality of life than physical incapacity; (c) age, since the quality of life of the young was viewed as intrinsically higher than the quality of life of the old; and (d) perception of being a burden on others. ²² Poor quality of life was conceptualized as having terminal and/or serious long-term illness, having mental impairment pre-CPR or as a consequence of receiving CPR, being old, and being a burden or dependent on family or society. ^18,22 Without reasonable expectation of quality of life after CPR, older people did not want it. Thus, full and informed consent to CPR must include quality of life considerations should CPR actually be successful, ²⁰ and providers must engage patients more deeply in CPR decision-making, because quality of life can mean something different to each patient.

Risk versus benefit

Informed consent with disclosure of risk and risk–benefit ratios are basic patient rights. ²³ Difficult as the situation might be, if a DNAR order is determined to be in the patient’s best interest, CPR should not be presented as a realistic choice—and should be communicated by a competent, compassionate, knowledgeable provider as a benefit to overall care. Risk versus benefit conversations between patient and provider must be accurate and complete because the risks of CPR are substantial and survival to discharge (StD) rates vary depending on co-morbidities. ²⁴ For instance, patients with metastatic cancer had a 0% StD rate, while patients with other cancer diagnoses had a 5.8% StD rate. Hospitalized patients with heart disease, who had experienced myocardial infarction, had a StD rate of 24%. Many of the patients who were resuscitated suffered from other effects of the intervention such as fractured ribs, cracked sternum, mediastinal hemorrhage, pneumothorax, aspiration pneumonia, liver lacerations, and tears to other vital organs. ² Irreversible comas were seen in about 10% of resuscitated patients. ²⁴ Patients should be apprised of these potential complications as part of the informed consent process.

Communication and code status

The problem of ineffective communication with patients regarding code status has been recognized for many years. Almost two decades ago, Balaban ²⁵ asserted—and Gawande ⁷ more recently affirmed—that physicians did not receive sufficient education in medical school on communication with patients and families about end-of-life issues. Deficits in communication skills were more evident in older generations of physicians. While efforts to enhance physician communication skills during medical school are beyond the scope of this article, ²⁶ suffice it to say that deeper conversations about suffering and functional loss post-CPR are still the exception even though CPR methods and complications are discussed with patients to some degree, which is important because patients most often accept aggressive medical treatment when they believe the end result is full restoration of previous function, life purpose, and meaning. ²⁷

Other research on communication with patients about code status shows that communication is far from optimal. For example, Gehlbach et al. ²⁸ interviewed patients, their surrogates, and physicians about CPR knowledge, preference for code status, and goals of care. Analysis showed that only half the patients recalled discussing code status and the goals of care with a doctor. ²⁸ Moreover, compared to patients’ recollections, 16 inconsistencies with code status orders were found. Only 5% of patients who preferred full code status could verbalize the basic facts of CPR procedure, and the goals of care as understood by patients and surrogates were inconsistent with those that their physicians considered important. ²⁸

Similarly, Orvos ²⁹ interviewed patients aged 55–79 years with cardiac, advanced pulmonary or liver disease, and patients aged 80 years or older with any acute condition to retrospectively determine the nature and accuracy of communication between provider and patient regarding preferences for life-sustaining treatment. Patients were asked which, if any, family members or proxies knew of their wishes, if the wishes had been discussed with a provider, and if they had been written into an advance directive. Patients’ medical records were reviewed to see if their wishes were reflected in the documentation. ²⁹ Results showed that 37% of participants were subject to medical error—2% of participants who wanted CPR had it withheld, while 35% of patients who wanted a DNAR order would have received instead full resuscitative efforts had a cardiac event occurred. ²⁹ The patient sees the provider as an expert on these issues, and trust is foundational to the patient–provider relationship. Thus, the conditions that led to these remarkable defaults of responsibility should be identified and changed, which leads to Walker’s ¹⁰ ethics of responsibility.

Traditional model of moral inquiry

Kantian ethics (and Enlightenment moral philosophy in general) gave rise to the moral theories used most often by bioethics and biomedicine—and consequently, by Nursing, which historically has been subordinate to the goals and interests of biomedicine. ^8,30 Biomedical ethics is based on deontological ethical theories that assume universal laws and a core of moral knowledge that can be accessed through the exercise of reason. ¹⁰ It locates morality in a few central moral principles that are understood, balanced against each other, and tested by individual reflection. ¹⁰ With a principle-based biomedical ethics, the person(s) faced with an ethical decision, for example, to forego aggressive, life-saving treatment like CPR at the end of life, weighs these central moral principles—beneficence, nonmaleficence, fidelity, justice, and autonomy—applies them to the ethical situation at hand, and uses the resultant moral knowledge to come to a rational decision about what is right or good to do. This model of ethical inquiry, which Walker ¹⁰ has called theoretical and juridical is largely free from context in the sense that having the right action-guiding system, like a principle-based ethics, ensures that one can figure out what is right or good to do in any similar ethics scenario or patient situation. ¹⁰ With this kind of action-guiding system, a provider confronted with a patient’s decision to refuse aggressive treatment at end of life might approach the problem by balancing the ethical principle of autonomy, that is, the patient’s right to decide his own best interests, against the ethical principles of doing good or at least of doing no harm and then applying the ensuing moral knowledge to the ethical decision at hand.

However, context is inescapable. Survival-to-discharge rates apply differently to one patient—for example, to one with serious co-morbidities and advanced age—than to another. Family members of one patient have different narratives of relationship to their loved one than do family members of other patients, and these differing narratives influence their decisions as proxies or surrogate decision-makers. Practices of responsibility with respect to CPR decision-making are applied differently in one clinical setting than in another, and different healthcare occupational groups have different priorities and different degrees of delegated responsibility. The eradication of disease and the preservation and prolongation of life have long been among Medicine’s top priorities; patient advocacy and holistic care have long been among Nursing’s top priorities. Nursing priorities also have included imperatives to move beyond the medical diagnosis, especially in the context of the withdrawal of aggressive care at end of life, to understand who the patient is as a person, to help families more accurately see the deteriorating condition of their loved ones, to imagine and act on moral possibilities for better end-of-life scenarios, and to facilitate saying goodbye and seeing the patient through to the end of his or her story. ³¹

Alternative model of moral inquiry

Because context is inescapable, a group of people attempting to make a moral decision typically engage in a narrative process that strives for consensus, unless they are structured to do otherwise. Narrative (story) and analogy (comparison) are the forms that moral reasoning takes. ¹⁰ Different people with different knowledge of the patient tell their stories from their own perspectives. They compare their own story to others’ stories and to similar stories from the past or from another context. The narrators work inductively, from all the particulars of the patient’s unique context and from all the particulars of various persons’ disparate moral knowledge to a moral understanding that is broadly shared. Emotion, intuition, and empathy are factors; and it is not entirely or even primarily a matter of rational reflection. The principles at stake are not usually identified at the outset, although analysis of applicable principles can inform decisions and supplement other forms of moral knowledge.

An alternative model of moral inquiry for healthcare ethics encapsulates and describes the narrative, socially negotiated process that occurs in practice. Walker’s ¹⁰ alternative model, which she labels expressive and collaborative, as opposed to theoretical and juridical has given rise to the ethics of responsibility. This model holds that morality consists of a group’s or a system’s actual social practices and not moral theories. ¹⁰ In an ethics of responsibility, the practices characteristic of morality are the practices that implement commonly shared understandings about how responsibilities, for example, for identification and implementation of informed consent practices and code status, are to be divided and assumed. ¹⁰ These are commonly shared understandings about who is obligated to do what for whom and who might be privileged to avoid or even remain unaware of these kinds of obligations. ¹⁰ Thus, for example, in the US healthcare system, it is the physician’s role-defined responsibility to write DNAR orders and to obtain the full and informed patient consent that provides the ethical and legal foundation for the order. However, all too often, the discussion is deficient in the ways already discussed. Not only is the signature on a consent form delegated to others—nurses, most often—but the discussion that would clarify the patient’s and family’s best interests and true intent is deflected as illustrated in the narrative prototype. From Walker’s ¹⁰ alternative model of moral inquiry, these delegations and deflections of responsibility constitute the practices that create and sustain a system’s morality. Identification, examination, critique, and correction of such practices provide the leverage for new moral understandings among those who inhabit the social space. ¹⁰

Professional responsibilities

According to the American Nurses Association (ANA), confusion about DNAR orders persists even with efforts to help patients, families, and proxies make informed choices. ³² While physicians currently bear the responsibility for DNAR and code status orders—and any inadvertent or inappropriate delegations and deflections of responsibility aside—nurses have legitimate professional obligations to advocate for discussions about code status and to play an active part in them. This role is part of the comprehensive nature of nursing care, which is outlined in Provision 1.3 of the ANA Code of Ethics for Nurses. ³³ Of course, just because life-sustaining treatments are available does not mean that they should be recommended, advocated, or used. Advancements in healthcare technologies can prolong life but become a matter of profound ethical concern at the point that such advancements also prolong suffering. ³⁴ The conditions that produce and prolong suffering can contribute to emotional and sometimes unstable medical decision-making processes especially with respect to code status, where decisions to resuscitate or not—and at what level of effort (full resuscitation, AED only, no AED, chest compressions only, cardiac medications only, ventilation only, or DNAR)—require skilled communication and a stable, well-rationalized, ethical plan. ⁴

A study particularly relevant to this article’s purpose explored nurses’ moral knowledge in situations of perceived, overly aggressive medical care, of which CPR can be a particular type. ³⁵ The work illustrates the application of Walker’s ¹⁰ approach to ethical inquiry and thus receives more focus than other selected literature. Particularly illustrative is the tenet that the way work is divided in a social group establishes how moral responsibilities are understood and shared by differently situated people, for example, by hospitalists and ICU nurses, whose role-related responsibilities define the scope of their moral agency, shape their professional identities and values, and determine who is accountable for CPR outcomes. ¹⁰ Results showed that nurses, physicians, and families experienced much uncertainty and moral distress in the process of negotiating their moral responsibilities in scenarios much like this article’s prototype. ³⁵ Roughly one-third of narratives represented the deflection of the medical team’s responsibility to adequately respond with appropriate communication to medical futility or threat of death. ³⁵ Lack of communicative competence and comfort among clinicians created uncertainty among patients and family members and contributed to the choice of aggressive treatment, questions of medical futility notwithstanding. ³⁵ Moreover, the medical emphasis on cure, which ensued as a consequence of how responsibilities for informed consent typically are divided, for example, between Medicine and Nursing, led to the prolongation of aggressive care beyond that which was in patients’ best interests. ³⁵

In Walker’s ¹⁰ model, physicians’ primary responsibility to preserve and prolong life is central to understanding the physician’s moral identity, values, and relationships. Opting for a team-oriented intervention like CPR may dissipate the physician’s individual responsibility for the death of a patient, for example, through a DNAR order, and may deflect responsibility onto an entire social network of clinicians whose well-defined roles in the delivery of aggressive care in this often-experienced medical context serves to make death more socially acceptable. ^35,36 Despite any differences suggested between nurses’ and physicians’ responsibilities in the area of CPR decision-making, both groups share difficulties communicating with patients and families when aggressive care is no longer effective. ³⁵

While more investigation is needed to clarify why clinicians of all types fail to develop the communicative and moral competencies needed to reliably find opportunities for the more detailed, in-depth conversations that can better support CPR decision-making, the practical and moral failure is not merely that of individuals. Viewed from the perspectives of Walker’s ¹⁰ model of moral inquiry, CPR practices, communicative and moral competencies, role requirements, and divisions of labor in informed consent processes are socially organized practices of responsibility that are responsive to changes in social arrangements. Those practices of responsibility change when organizational and institutional norms, structures, processes, and practices change, for example, in the organization of healthcare or in education, law, and public policy as social institutions. Thus, close examination of practices of responsibility and how they are working in a given healthcare context and can show where and how to exert leverage for social and institutional change. For example, what if close examination of practices of responsibility in CPR showed that the responsibility for informed consent to CPR should fall in some greater measure to Nursing given its proximity to patients, its expertise in therapeutic communication, and its ethos of advocacy and care? When social understandings shift, moral understandings do as well. ¹⁰