Advance care planning: It is time to rethink our goals
Chetna Malhotra
Abstract
What is the purpose of advance care planning (ACP)? Does it facilitate goal concordant care,1, 2 improve other patient and caregiver outcomes or curtail the escalating healthcare costs at the end of life (EOL)? In short, is it worth the time and resources spent? These are the questions that have plagued researchers, clinicians, and policymakers internationally,3 thus challenging the increasing efforts to implement and promote the uptake of ACP conversations and its use as a metric to measure the quality of EOL care. Recent systematic reviews have shed some light on this issue.4, 5 These reviews report that a large number of trials conducted among seriously ill but not imminently dying patients have failed to show that ACP improves goal-concordant care at the EOL.5 Further, the trials have failed to demonstrate the impact of ACP on improving patient quality of life. There is also inconsistent evidence that ACP improves patients' mental health, reduces health care costs, or increases home deaths. Thus, ACP does not seem to change patients' EOL trajectory or the care they receive at the EOL.5 A plethora of literature has focused on gaps and challenges in the implementation of ACP. These include, but may not be limited to, lack of appropriate legislative framework, physician training and involvement, and accessibility of ACP documents in medical records.4, 6-8 However, even if endless resources—time, money, and manpower—are spent on addressing all of the documented challenges facing ACP implementation, ACP programs may still not be able to achieve the holy grail of goal concordant care. The reason for this lies within the realms of cognitive psychology. When making decisions for the future, we project our current preferences onto our decisions for the future, even when there is likely to be an incongruity between our current and future experiences and preferences. This phenomenon is termed projection bias.9 Relatedly, people in certain affective states, such as in distress or pain, have difficulties appreciating what their preferences will be when they are not in these states. This is known as the hot–cold empathy gap.9 There are several examples of projection bias in EOL care literature. Chochinov et al. interviewed 168 cancer patients, admitted to a palliative care unit, at frequent intervals to assess their will to live. They found that patients' will to live fluctuated with their levels of dyspnea and distress, indicating that physicians' clinical decisions should not be based on a momentary assessment of patients' will to live.10 In a prospective cohort of patients with advanced heart failure, surveyed every 4 months over 2 years, patients' preferred type of EOL care (aggressive vs. non-aggressive) and preferred place of death was reported to change over time.11, 12 Interestingly, the change was not consistent in one direction and was influenced by patients' prognostic beliefs and their quality of life at the time of eliciting their preferences. Similar results were reported in studies with patients with advanced cancer,13 and in a study of caregivers of people with advanced dementia. In summary, patients' and caregivers' goals and preferences for the future change over time with changes in their circumstances and prognostic beliefs. What does all of this mean for the ACP process and clinical decision-making? While it may be feasible to align current medical care with patients' current goals and preferences, planning for future EOL care decisions in the face of ever-changing goals and preferences is challenging. In other words, if goals and preferences are transient, then goal-concordant EOL care is not an achievable aim for ACP programs. There is however some promise. First, many studies show that patients do not understand enough about their illness due to gaps in patient–physician communication,9 and as their health condition and prognosis evolve. Relatedly, even when the prognosis is clearly disclosed to patients, many patients may refuse to acknowledge it.14, 15 Their prognostic beliefs can also change over time with changes in their symptoms, even if the two are unrelated. This indicates some cognitive dissonance among patients.16, 17 ACP conversations have the potential to educate patients and their caregivers about their (patients') illnesses. These conversations should happen continually to address patients' evolving health conditions, prognosis, and cognitive dissonance. Second, ACP allows patients to clarify, express and share what is important for them with their healthcare providers and surrogates, thus paving the way for them to engage in shared decision-making and providing them an opportunity to make decisions for their current care.18 What is important to patients changes over time as their health condition and prognostic beliefs evolve. This again suggests that ACP conversations need to be revisited time and again. Third, even though the actual EOL decisions may be influenced by specific clinical scenarios and patients' and caregivers' goals and preferences at that point in time, ACP conversations prepare patients, caregivers, and providers to make those future in-the-moment EOL decisions.19, 20 Thus “preparation” rather than “planning” can be an objective for ACP.5 For providers, all of this implies that they should not be making EOL decisions based on a momentary assessment and documentation of patient preferences, as these preferences are likely based on patients' situation at the time of documentation rather than a future hypothetical scenario. Documented goals and preferences should be considered only if patients have expressed the same goals and preferences again and again over time, under a variety of health conditions, clinical scenarios, and settings. What are the indicators to assess the success of this redefined ACP? As mentioned earlier, ACP is unlikely to change the EOL trajectory for patients, including goal concordant care, patient quality of life, and mental health or reduce health care costs in a substantial way. I, therefore, recommend that ACP's success should be measured in the context of its three objectives (Figure 1). The success of illness education can be assessed by measuring patients' and caregivers' understanding of their (patients') illness and prognosis. The success of value-elicitation can be assessed by measuring the extent to which patients share their values/goals with their providers and caregivers, whether providers and caregivers/surrogates understand patients' values, whether it strengthens patient–provider and patient–caregiver relationships, whether patients participate actively in decision-making, and the extent to which current treatments are concordant with patients' values and goals. Whether ACP prepares patients for future EOL decisions can be assessed by asking patients and caregivers if they feel prepared for future EOL decisions, and its impact on caregivers' bereavement adjustment. ACP is unlikely to facilitate the provision of goal-concordant care, reduce healthcare costs, or change patients' physical or mental health. Therefore, the current approach of asking patients to document their goals and preferences for their future is unlikely to be useful, unless done for imminently dying patients. A new approach to ACP is required. In this new approach, ACP can have three main objectives—(1) illness education, (2) elicitation of values and goals to guide current decisions, and (3) preparation for future EOL decisions. The success of ACP should be measured in relation to its three objectives. It is time that we rethink the objectives of ACP and define new indicators for measuring its success. “It is time that we rethink the objectives of ACP and define new indicators for measuring its success” Chetna Malhotra conceptualized and drafted the article. The author declares no conflicts of interest. None. We thank Dr. Malhotra for her thoughtful and thought-provoking Commentary about Advance Care Planning (ACP) and Advance Directives (AD). One of the clinical scenarios that got me interested in a career in Geriatric Medicine when I was a medical resident was seeing very frail older patients who were obviously at the end of life come to the Emergency Department and be admitted to the Intensive Care Unit on full code status. I wondered why such futile, uncomfortable, and expensive care was being done. By necessity I learned all I could at the time about ACP and AD, and since my experience and understanding have increased exponentially. While I do not disagree with Dr. Malhotra's thoughts and recommendations, I do have some of my own that are complimentary. First, if you have ever witnessed a 90-pound, 90-year-old woman who is bedridden and unable to recognize or communicate with her loved ones, getting CPR in a nursing home, you know how brutal and futile it is. Numerous ribs cracking with virtually no chance of meaningful survival. People must be educated about being in this situation and document their choices for it. Second, I have had many families ask me over the years: “What do you think we should do Dr. JO?” Many physicians will say something to the effect: “it is your decision,” thereby abrogating their responsibility to be helpful at an extremely stressful time. My approach, with due respect to the ethical tenet of paternalism, is to tell them what I would do for my own parents (now what I actually did for my own parents). This can help relieve the family or other surrogates from stress and a lifetime of potential guilt for “pulling the plug.” Third, I am skeptical about order forms such as the POLST. I am not convinced that those without a medical background or previous experience who sign such forms truly understand the risks and benefits of specific treatment options. In addition, decisions may change depending on clinical circumstances. For example, a patient who has elected not to have antibiotic treatment at the end of life may develop a painful abscess or cellulitis that requires an antibiotic for comfort. Thus, I tend to favor more general advance directives such as the one I have that simply states that if I am in a condition with very limited life expectancy, I do not want any intervention that will cause me and my family more distress or discomfort than the potential benefit. With this directive, my surrogates and health care professional can make decisions that are in my best interest at the time. I am certain that many JAGS readers have other thoughts, and I encourage those who due to submit them for review and potential publication a Commentary or Letter. -Joseph G. Ouslander, MD