Ethics in Cancer Care Delivery—Do Not Resuscitate

Introduction

The National Cancer Act, also known as the “War on Cancer” Act, was signed into law in 1971. Thanks to the Act’s support for ongoing cancer-related research, the number of cancer survivors in the United States has been progressively increasing. Today, nearly 70% of those diagnosed with cancer will be alive after 5 years. It is predicted that in 2026 there will be 20.3 million persons in the United States who have survived cancer.

The cure rates for patients with localized cancer have steadily improved, and many patients with metastatic disease are now able to live for years with their malignancy as a chronic illness (e.g., prostate, breast, and colorectal cancers). Many unprecedented discoveries within the last few decades, most notably in the fields of molecular genetics, cancer cellular biology, and immunology, have led to the development of many novel anticancer therapeutic modalities, from molecular targeted agents to bioengineered cellular treatments. Due to these therapeutic advancements, cancer patients with advanced stage disease, once considered to be an imminently fatal diagnosis, are anticipated to have a longer life expectancy.

Thus for many patients, a diagnosis of cancer does not imply a prognosis of imminent death. Certain complex therapeutic interventions in cancer patients such as surgery, or rescue procedures such as cardiac resuscitation, are not necessarily futile or contraindicated under the appropriate circumstances. Evidence-based risk assessments, including anticipated prognosis of the malignancy, are therefore important considerations in treatment planning. To establish appropriate goals of care, these issues should be discussed with the patient and their family well in advance of acute care interventions.

In this chapter, we will review the evolving evidence on the outcomes of cardiac resuscitation as well as palliative surgery in cancer patients. We will also discuss the regulatory requirements for ensuring that patients’ rights to do not resuscitate (DNR) decisions are safeguarded, and the ethical dilemmas that such decisions could generate for clinicians.

Cardiopulmonary Resuscitation

“Closed-chest cardiac massage” was introduced into medical practice in 1960 by Kowenhoven, Jude, and Knickerbocker when they published in the Journal of the American Medical Association a series of 20 selected patients who each had an in-hospital cardiac arrest (IHCA) event that was witnessed. Their cardiac arrest events were reversed through the application of a technique of external chest compressions, accompanied by external electrical shocks in some cases. Fourteen of the 20 patients survived and left the hospital, leading to their conclusion: “The use of this technique on 20 patients has given an over-all permanent survival rate of 70%. Anyone, anywhere, can now initiate cardiac resuscitative procedures. All that is needed are two hands.”

The application of this technique, which was renamed “cardiopulmonary resuscitation,” or “CPR,” quickly caught on, and began to be applied literally “anywhere,” both within the setting of hospitals and in the community at large. Within a few years, outcomes in large series of cases were being published. One of the largest early in-hospital CPR case series included over 500 patients who had experienced an IHCA at the Royal Victoria Hospital in Montreal Canada. Published in 1967, they defined resuscitation as successful if the patient survived 24 h or longer. By this definition, resuscitation was successful in 32% of the cases, but only half of these patients (15%) recovered completely and were discharged from the hospital. These results were vastly different from those of the original publication and, until recently, have remained remarkably constant across institutions and across time. In 1987, nearly 30 years after the original study, McGrath conducted an extensive literature review of published in-hospital CPR outcomes in studies that included at least 100 patients. Forty-two publications met his criteria and in total included 12,961 patients. The summary of results demonstrated essentially the same outcomes as the early Canadian series: on average, 15% of patients had long-term survival after in-hospital CPR. An additional observation was that patients with coronary artery disease, who likely had developed an arrhythmia or asystole due to cardiac ischemia, were the group that most seemed to benefit from CPR.

Law, Regulations, and Guidelines

Rescuing patients from what previously had been fatal events, regardless that the numbers were relatively small, was considered clinically significant, and CPR became widely used. The results, however, were consistently showing that in the setting of a general hospital patient population, most patients with IHCA were not benefitting from CPR, and some who survived were left with late effects, including cognitive deficits and even in vegetative states of existence. Due to concern that CPR was not appropriate for all IHCA patients, the American Medical Association in 1974 issued a recommendation that patients be informed of risks/benefits of this procedure, and the patient’s decision regarding their preferences for CPR (or “code status”) be documented in the medical record. Nearly a decade later, in 1983, the President’s Commission for the Study of Ethical Problems in Medicine published their opinion that consent for CPR be considered implicit, unless otherwise explicitly stated with an order issued to withhold CPR. This stemmed from a position of moral reasoning that a person would rationally want to be given an opportunity, even if small, that they would survive a cardiac arrest. Cardiopulmonary resuscitation thus became the only intervention for which patient consent is required to NOT have the procedure, and an order must be written to NOT perform it.

In 1988 the Joint Commission issued a standard that hospitals must have policies regarding CPR. The Patient Self-Determination Act of 1991 further required that hospitals receiving Medicare or Medicaid funds were to inform patients of their right to self-determination, including the right to refuse life support procedures such as CPR. The American Society of Anesthesiologists (ASA), in alignment with the Patient Self-Determination Act, first published guidelines in 1993 stating that the practice of automatic perioperative suspension of DNR orders was in direct conflict with patients’ rights to self-determination. This principle has been reiterated in more current updates to the ASA’s guidelines, with support from the American College of Surgeons.

Intraoperative CPR Outcomes

Despite the time elapsed since the first publication of the ASA guidelines, there have been conflicting points of view with regard to the suspension of CPR intraoperatively, due to several considerations. First, some argue that because cardiopulmonary failure may be secondary to the processes of anesthesia itself, there is a responsibility to attempt to reverse this adverse effect. Second, the published outcomes of CPR in the intraoperative setting have been generally more successful than those in the hospital at large. An example of these outcomes is a published study by Keenan and Boyan in 1986 in which they focused on identifying anesthesia-related causes of cardiac arrest in over 100,000 anesthesia cases within a single institution in the United States. They reported an incidence of 1.7 events/10,000 anesthesia cases. Among 27 anesthesia-related cardiac arrests, they identified that the most common cause of the arrests was failure to ventilate appropriately, followed by overdose of anesthetics, and third by hemodynamic instability. Forty-eight percent of the patients who had an intraoperative cardiac arrest were rescued and survived to hospital discharge.

Technological advances in intraoperative monitoring devices, newer anesthetics, and pain control agents, as well as checklists, have improved anesthesia-related safety in the operating room (OR) environment. Surgical procedures have also progressively moved to the ambulatory setting, making the hospital OR a selective environment in which the more complex and higher-risk surgical procedures are performed. Thus, the number and type of cardiac arrest events within the OR have also changed. A more recently published analysis of events in a Portuguese series of 122,289 anesthesia cases identified 62 cases of intraoperative cardiac arrests over a period of 7 years. In this case series, the most common causes of the arrests were due to underlying comorbid conditions, and not the anesthesia itself. The incidence of anesthesia-specific related events was 0.74/10,000 anesthesia cases. The identified risk factors for a cardiac arrest were ASA PS score higher than 3, underlying cardiac disease, and use of vasopressors. Nine cases were attributable to the anesthesia process itself. All nine survived the arrest and left the hospital, while 43% of the 62 patients who had an intraoperative cardiac arrest survived to discharge. Thus the overall survival expectation from intraoperative CPR seems to be holding above 40%, and in this particular series all anesthesia related cardiac arrest events were reversible.

Outcomes of CPR in Cancer Patients

Until recently, very little was known about the outcomes of CPR in cancer patients who had experienced an IHCA, other than studies stating that the diagnosis of cancer portended a poor outcome, with a predominant focus on patients with advanced cancer. A recently published multiinstitutional study compared survival outcomes post-CPR for IHCA in patients with a diagnosis of cancer that is not metastatic versus a propensity-matched cohort of patients without cancer. The data source is the National Inpatient Sample dataset, compiled by the Agency for Healthcare Research and Quality (AHRQ). The data for this analysis is from 2003 to 2014, with a total of nearly 2 million hospitalizations complicated by an IHCA event. Of these, 112,926 occurred in patients with a history of cancer, not metastatic. Overall the CPR to discharge survival was better for the noncancer cohort (46%) compared to the nonmetastatic cancer patients (31%), even when controlling for general prognosis and comorbid conditions. Of note is that from 2003 to 2014 the overall survival to discharge improved for both the cancer patients and their matched cohort: from 23% to 31% in the cancer cohort and from 40% to 46% in the noncancer cohort. These survival to discharge rates for CPR are thus significantly higher than published in the latter decades of the 1900s.

Additionally, the study compared postarrest procedure utilization for each of these groups and determined that the cancer cohort statistically was less likely to receive postarrest procedures, such as coronary angiography, percutaneous coronary intervention, or targeted temperature management. The authors hypothesize that the less intense post event management of the cancer cohort contributed to the lower survival outcome and imply that biases may exist against implementing aggressive postarrest management in those with a cancer history, even though there was a potentially good cancer prognosis in most cases. They conclude: “Additional research is needed to clarify the role of patient-physician perceptions of cancer diagnosis and selective applications of post-resuscitation care.”

A second recent study focused on the post-cardiac arrest CPR outcomes of patients with advanced (metastatic) cancer. The data source for this analysis was similarly a large database, the Get With The Guidelines Resuscitation registry. The data encompassed the time period of April 2006 to June 2010, included 369 hospitals, and 47,157 adult patients with an IHCA. Of these, 6585 IHCA events occurred in patients with advanced cancer, and 1143 of these in turn occurred in cancer patients who were hospitalized for a surgical intervention (177 were cardiac surgeries). Independent of other risk factors, the survival to discharge for patients with advanced cancer was much lower than those without cancer (7.4% vs. 13.4%). The study did not report on the outcomes of the surgical subset of advanced stage cancer patients.

Similarly to the nonmetastatic cancer patients, the postarrest care for the metastatic cancer patient cohort was significantly less intense, with a higher rate of DNR orders within 48 h of the arrest compared to the noncancer patients. Of note, the authors of this study comment that despite the low rate of survival to hospital discharge for the advanced cancer patient group, in their opinion this rate is not so low as to be considered futile and merits discussion in setting goals of care for patients with advanced cancer. However, compared with patients with nonmetastatic cancer, and noncancer patients, the CPR to discharge survival rate for patients with disseminated cancer is clearly very low ( Table 57.1 ).

Table 57.1

CPR to Discharge Survival for Patients with Inpatient Hospital Cardiac Arrest (IHCA)

Location	Patients without Cancer	Patients with Disseminated Cancer	Patients with Nondisseminated Cancer
Hospital	15%–46% , ,	<10%	31%
Operating room	43%–48% ,	–	–

Outcomes of Palliative and Rescue Surgery in Cancer Patients

For centuries, surgery has been the primary intervention for many types of malignancies. It continues to be a primary and potentially curative treatment modality for most solid tumor malignancies presenting with localized or limited stages of disease, whether as a single treatment strategy or in combination with chemotherapy and/or radiation. While the benefit of surgery in those circumstances is not questioned, the benefit of surgical interventions for patients with disseminated (metastatic) disease has been controversial. In many instances, surgery can have significant benefit in palliating symptoms, improving quality of life, and even prolonging survival in these patients. On the other hand, there are unique and higher risks of morbidity and mortality associated with surgery in patients with metastatic cancer, and therefore assessing risk/benefit is a critical element of treatment planning.

The National Surgical Quality Improvement Program database of the American College of Surgeons (ACS NSQIP) has made it possible to analyze outcomes in large cohorts of patients across multiple institutions. Thus, risk analysis nomograms have been published for many surgical procedures. Tseng et al. formulated a predictive nomogram of morbidity and mortality specifically for patients with disseminated cancers who undergo surgery. The authors queried the ACS NSQIP database from 2005 to 2007 and identified 7447 patients with disseminated malignancy diagnoses who had undergone surgery. The overall morbidity and mortality in this cohort of patients were 28.3% and 8.9%, respectively. They analyzed in total 133 factors, including preoperative, intraoperative, demographic, postoperative, as well as surgical type factors for their significance in the risk of morbidity and mortality and identified 14 factors as significant for risk: DNR status prior to procedure, weight loss >10% of body weight, dyspnea at rest, functional dependence, the presence of ascites, chronic steroid use, active sepsis, elevated serum creatinine, low serum albumin, high white cell count, low hematocrit, surgical acuity (emergent procedures), type of procedure, and patients’ age. These factors are ranked and are weighted by point scores in the nomogram, and when totaled for the individual patient, a percentage risk can be calculated for that unique individual. While it is complex given the number of risk factors that are taken into account, the nomogram can help assess relative risk for an individual patient in an objective manner, and support treatment decisions.

A more recent outcomes analysis in a larger ACS NSQIP database cohort of 21,755 patients with disseminated malignancies, who had a surgical procedure from 2006 to 2010, noted that more than half of all the surgeries (54%) were either bowel resections or other type of gastrointestinal procedures, and an additional 10% were multivisceral resections. The authors examined the trends of morbidity and mortality across the 5-year period of the analysis and identified improvements in the outcomes: morbidity decreased from 33.7% to 26.6%, and mortality decreased from 10.4% to 9.3%. They attributed the improvements to fewer emergency procedures, fewer patients with acute sepsis, as well as improvements in presurgery performance status and body weight.

Another situation to consider is that of patients who are undergoing chemotherapy that can lead to leukopenia and while under treatment, develop acute complications that require surgical rescue. A prior study looking at the risks of emergency surgery in 956 cancer patients (ACS NSQIP database, 2005–2008) who had received chemotherapy within 30 days of surgery showed an increased risk of morbidity and mortality for the cancer patients when compared to a matched noncancer cohort. The morbidity risks were 44% vs. 39%, and mortality risk 22.4% vs. 10.3%, respectively ( Table 57.2 ). Thus, unless surgery is absolutely emergent, medical management of neutropenic patients until counts are recovered should be considered. If surgery must be performed, the patients and their families or surrogates should be aware of the relatively high morbidity and mortality risk of such procedures.

Table 57.2

Only gold members can continue reading. Log In or Register to continue