Department of Surgery, Oregon Health and Science University, Portland, OR, USA
Because of the high mortality rate in elderly patients with unplanned ICU admissions, practitioners need to have the skills to provide optimal end-of-life care. A cultural shift toward including basic palliative care in routine surgical ICU care may improve our ability to effectively care for our patients.
End-of-life care is broad in scope and can only be done successfully with defined goals of care, familiarity with symptom management, and comfort in addressing personal and professional ethics.
Understanding the differences between healthcare power of attorneys, advanced directives, and do-not-resuscitate orders—as well as the roles of palliative and hospice care—is key to engaging in discussions with patients and families.
Given the severity of illness of patients arriving to the intensive care unit (ICU), it is not surprising that mortality of patients admitted to the ICU is high. Approximately 20 % of all deaths in the United States occur each year in the ICU or within days after transfer from the ICU [1–3]. Furthermore, most of these deaths are a result of the decision to withdrawal or withholding life-sustaining therapies [4–8].
When considering geriatric trauma patients, trauma is the fifth leading cause of death in patients over the age of 65 . Furthermore, though geriatric patients represent only 12 % of the overall trauma population, they account for 28 % of the deaths due to trauma . The effects of age begin to impact mortality at age 45, but for each year over age 65, mortality increases by 6 % [1, 11]. Emergency general surgery is also associated with a high risk of death in geriatric patients, with a mortality of two to three times that of nonemergent surgery . Similarly, when compared to younger cohorts undergoing emergency general surgery, mortality of geriatric patients is two to three times greater [13, 14]. Of note, more than 10 % of Medicare decedents have surgery during their last week of life, and more than a third of Medicare patients in hospice were transferred there from an ICU .
Given the challenges in management of elderly surgical patients, it is clear that during a surgeon’s career, he/she will encounter many patients with end-of-life needs. Despite the ubiquity of these issues, there is tremendous evidence that quality of care at the end-of-life, as perceived by patients and their families, needs to improve [3, 16]. Table 42.1 lists common barriers to institution of palliative care measures in surgical ICUs. Some of these barriers to satisfaction and high-quality end-of-life care include inflated expectations of outcomes held by clinicians and families, uncertainty regarding prognosis, delay in attention to palliative needs, and the lack of an integrated approach to end-of-life care . An understanding of palliative care methods and the ethical management of patients at the end of life as well as the needs of their families is necessary to improve the care of the dying patient.
Surgical ICU culture and barriers to palliative care
Surgical ICU characteristics
Barrier to palliative care
Structure and process
No interdisciplinary team rounds or processes of care
Surgical hierarchy prevails
Surgeon has primary relationship with family, not ICU team
No leadership to drive change
No standard processes in place
No one person has responsibility for end-of-life decisions or communication
Interdisciplinary team undermined or devalued
Communication in silos leads to conflict and delays in decision-making
Delay in palliative care consult
Surgical patients and disease
Different disease trajectories: trauma, transplant, surgical oncology, and neurosurgery
Lack of clear prognosis for mortality
Lack of quality of life or functional outcome data
One palliative care approach does not fit all ICU patients
Triggers for palliative care evaluation need to be disease specific
Can’t tell who is dying, and then can’t do palliative care until very late in course
Difficult to predict if patient’s quality of life will be acceptable to patient’s wishes
Sudden or catastrophic illness
High expectation for life-saving surgery (transplant, cardiovascular)
No knowledge of patient’s wishes
Lack of readiness for end-of-life decisions
Unwillingness to accept poor outcome
Family burdened by decision-making
Intensive need for family support
Palliative care = end of life care, only for dying
Death is a failure
Fear of abandonment
Fear of destroying hope
Lack of skill or expertise in communication, pain management
Palliative care delayed until death imminent
Avoidance and delay of family meetings and end-of-life decisions
Refuse palliative care consultation
Avoidance of discussion of bad prognosis
Several barriers hinder a timely and adequate approach to end-of-life care; one of the most difficult to overcome is that of surgical culture. The historical approach in the surgical or trauma ICU has been a culture of rescue. This frames the surgeon as a warrior against disease carrying an absolute responsibility to save the patient’s life . This culture can interfere with open discussions and decision-making. “Admitting defeat” by withdrawing or withholding life-sustaining treatment and allowing a patient to die is an emotional challenge that many surgeons are unwilling to face. As a result, surgeons frequently delay end-of-life discussions and palliative care treatment despite patient or family member requests for these measures. In 2008, Morrell et al. found that surgical services took nearly twice as long as their medical colleagues to establish DNR status in a single-institution review of the management of patients with life-threatening illnesses . Moreover, many patients and families find that continuation of therapies often proceed beyond a time when they feel it is wanted or even acceptable .
The relationship between the surgeon and patient has been described as both a covenant and an implicit contract made preoperatively between a patient and their physician . In an analysis of physicians’ opinions regarding the contract between patient and physician, Schwarze et al. found that important negotiation about postoperative participation occurred during preoperative surgical planning and discussion. It is during this time that an implicit contract is created that has important consequences in the postoperative time period .
It is important to note that during these preoperative discussions for planned procedures, the person who would be the patient’s surrogate decision-maker may not be present. This has significant implications for end-of-life decision-making. It has been reported that 64 % of surrogates of critically ill patients did not believe physician predictions that ongoing treatment was futile . Rather than relying on the medical team for prognostic information, 98 % relied on nonmedical sources (including patient character and will to live, as well as spiritual beliefs) to determine prognosis . The arrival of a surrogate who was not part of preoperative discussions, and who has a vision for goals of care different from the surgeon, can lead to feelings of betrayal and create conflict . Thus, it is well advised to encourage the participation of family in preoperative planning for high-risk procedures.
It is encouraging to note that structured interventions tailored to individual surgical ICUs can successfully effect cultural change with resultant changes in patient outcomes. Mosenthal et al. reported such a program that resulted in earlier DNR orders (without an associated change in death rate) . This allowed family members and loved ones more time to say goodbye and increased the ICU team’s focus on pain control.
30-day mortality is a publically reported, commonly cited quality metric used across the country. These data are intended to provide patients and community members with valuable information about safety of operations at a given institution or with a particular provider. While these data are easy to capture, they can be misleading when the patients or procedures involved are particularly high risk. Recent work by Talsma et al. showed that 30-day mortality significantly underrepresented risk of death when compared with 90-day mortality. Further, they found that 90-day mortality was better at capturing deaths related to the surgical procedure, rather than patient comorbidities . A desire to keep 30-day mortality measures low may motivate providers to continue futile or unwanted interventions that keep patients alive for the duration of this measure . Similarly disturbing, it may prompt discriminatory practices where high-risk patients are avoided. 30-day mortality also fails to capture important information regarding palliative procedures; relieving an obstruction in the face of non-survivable cancer may be aligned with the patient’s goals of care and may be quite effective at alleviating suffering .
Vignette: The surgeon on call at a tertiary care facility is paged by a rural hospital requesting transfer of an 85-year-old demented man with a small bowel obstruction. The patient has emesis not controlled by nasogastric tube and systemic inflammatory response syndrome (SIRS). The surgeon accepts the patient, who upon arrival is tachycardic and having feculent emesis. The patient’s son (who is his medical decision maker) advocates for surgical exploration. In the operating room the obstruction is relieved, but the patient is found to have a non-survivable amount of infarcted small bowel. Goals of care are adjusted, the patient is made comfortable, and the following morning life-sustaining treatment is withdrawn in ICU after the patient’s family has arrived and said good-bye.
Does 30-day mortality capture patient safety in this case? Does it capture the surgical team’s efficacy at meeting goals of care?
Frailty vs. Age
Although age has a well-described impact on outcomes in the surgical and trauma ICU [12–14], frailty may be more informative about prognosis. Frailty can be understood as a decreased physiological reserve and resistance to stressors that results in increased vulnerability to poor outcomes . Joseph et al. showed that in trauma patients older than 75 years, frailty index was a better predictor than age for mortality, adverse discharge, pneumonia, and urinary tract infection . There are many competing validated frailty indices; unfortunately most require the input of too many factors to be calculated easily or routinely in clinical practice. Easy to assess metrics that may provide an admittedly incomplete measure of frailty include unintentional weight loss >10 lbs in the past year, slowed walking speed (>7 s to walk 15 ft), and degree of independence prior to hospitalization [27, 28].
Establishing Goals of Care
The transition of care from curative to palliative does not need to be a dramatic shift in strategy  (Fig. 42.1). A patient-centered approach to care should include early discussions with the patient and family regarding goals of care and treatment limitations. These discussions should occur prior to the acute situation when decisions are difficult to make and the patient frequently is no longer able to voice his or her opinions. In situations where the admission to the hospital or ICU is the initial acute situation, the discussion should occur as soon as possible—ideally, during the first conversation with the patient and/or family . Communicating clearly and establishing an open dialogue are critical to setting expectations and allowing those with the most at stake to be active participants in healthcare decisions. This approach also ensures that no matter the goals of treatment, care will not be withdrawn. Certain treatments may change, but the patient and family are provided care and support throughout their illness.
One of the first steps is clearly delineating goals of care. Goals of care include the expectations of the patient and surgeon if everything goes as planned as well as if there are significant complications encountered. These goals should include discussions regarding optimal treatment outcomes and what the patient would consider unacceptable outcomes. Many patients have opinions regarding quality of life limitations that they have considered privately but never spoken about to family or healthcare providers. This conversation should include not only goals of care for while in the hospital but also opinions about disposition. In particular, it is important to identify patients for whom a skilled nursing facility (SNF) would be unacceptable either as a short-term discharge destination or as a long-term disposition.
Treatment limitations include setting limits on the aggressiveness with which to sustain life. These limitations may be made considering both the potential for survival and/or the subsequent quality of life. Treatment limitations might include any number of treatment modalities. These limitations are best set while the patient is lucid and able to rationally consider all treatment options.
Terms and Definitions
Advance directives, advance healthcare directives, medical directives, and living wills fit into this category. They are defined as “written expression of how you want to be treated in certain medical circumstances.” They can address any number of scenarios in specific or general terms. Examples include use of feeding tubes, mechanical ventilation, discharge to nursing facilities, and desire for cardiopulmonary resuscitation.
Do Not Resuscitate (DNR)
A DNR is a physician order regarding the implementation of therapy if the patient were to suffer a cardiopulmonary arrest. On the surface, DNR seems simple to understand and implement; however, there are many nuances to the DNR order that complicate execution of the wishes of the patient. Frequently, DNR orders are not discussed during the preoperative stages of care. Furthermore, most hospitals have rules informed by the American College of Surgeons and American Society of Anesthesiologists’ guidelines regarding DNR order holding while the patient is in the operating room. Highlights of these guidelines are as follows:
Institutional policies that automatically enforce or cancel pre-existing DNR orders do not sufficiently support a patient’s right to self-determination.
Physicians should discuss and document DNR status with the patient or surrogate decision-maker prior to proceeding with a planned operation.
Options include suspending the DNR for a defined amount of time in the perioperative period, continuing DNR in the perioperative period, and limiting specific interventions (such as chest compressions but not intubation) in the perioperative period.
DNR orders may be complicated to understand as many individuals do not want a portion of resuscitation but would allow other treatment modalities if necessary. For example, a patient may not want chest compressions but would allow intubation or performance of electrical defibrillation. There are stigmata surrounding various resuscitation treatments held by healthcare providers as well as patients. Most patients do not understand the risks and benefits of undergoing CPR in the hospital setting. Similarly, many healthcare providers are uncomfortable “allowing” a patient to die without intervention through heroic measures despite data showing poor outcomes in patients who suffer a cardiorespiratory arrest during hospitalization.
The use of these “partial DNR” orders should be discouraged . These orders are intended to provide a tailored approach to anticipated patient needs and desires; however, because it is impossible to predict all potential patient scenarios, these types of orders frequently lead to non-beneficial or harmful interventions . For example, a patient may not want chest compressions, but the use of medications without chest compressions during an arrest is extremely unlikely to be of any benefit to that patient. Patient education and physician response to patient emotions and questions are critical . It is up to the physician to provide a recommendation based on medical evidence and patient priorities when establishing code status .
Further complicating the implementation of DNR orders is the poor understanding of CPR outcomes seen in surrogate decision-makers. Although many decision-makers have taken CPR classes, it is common for their understanding of the outcomes of CPR to be based on depictions in TV or movies. In a recent study, more than 70 % of surrogate decision-makers thought that survival after CPR was >75 %, 20 % understood that brain damage could be present afterward, and only 2 % understood that the patient could be dependent on life support afterward . This knowledge gap is a significant source of unrealistic expectations.
Healthcare Power of Attorney
These are defined as “durable power of attorney specifically designed to cover medical treatment.” These designate a person who will make decisions on the patient’s behalf, although they do not specify what decisions should be made. State laws will dictate the hierarchy of designating medical decision-makers if this person has not been appointed. If no person exists, a court-appointed designee may be necessary. Involvement of an institution’s ethics board may be helpful in this situation.
There will be times when despite open conversations with patients and their family members, there will be disagreement over the use of certain life-sustaining therapies. Medical futility is a concept that refers to the appropriateness of medical treatment when the expected success of that intervention in poor. Quantitative futility defines a situation in which there is a <1 % expected chance that the intervention will be success. Texas and California have statewide futility policies, but the majority of the country relies on individual institutional policies. Physicians are not legally, ethically, or professionally required to provide any medically futile therapy as defined by the standard of the medical community.
Planned vs. Emergent Admission ICU
Vignette: A frail 76-year-old woman with terminal colon cancer presents with cholangitis. She is in severe pain and has an associated mental status change. She wrote an advance directive in the past specifying her wish for limited intervention in the event of terminal disease. Her husband, who is her surrogate decision maker, is unsure if they should proceed with surgical treatment of her biliary tract disease: he wonders if this should be seen as acute and reversible or as a fatal event.