Giving up on the Objective of Providing Goal-Concordant Care: Advance Care Planning for Improving Caregiver Outcomes

Abstract

The merits and effectiveness of advance care planning (ACP) continue to be debated a full thirty years after the passage of the Patient Self-Determination Act. This act gave patients the right to create advance directives, with the objective of ensuring that the care they received at the end of life was consistent with their preferences and goals. ACP has definitely moved beyond the completion of advance directives to encompass the identification of a healthcare agent and the facilitation of communication among patients, surrogates, and clinicians. Nonetheless, the provision of goal-concordant care remains a primary objective for ACP. This article argues that this cannot and should not be the objective for ACP. Patients’ goals change, and the provision of goal-concordant care is sometimes incompatible with other critical determinants of appropriate care. Instead, ACP should focus on the objective of improving caregiver outcomes. Surrogate decision making by caregivers is associated with an elevated risk of post-traumatic stress disorder and other adverse outcomes, and these outcomes can be improved with ACP. ACP focused on caregivers involves helping caregivers to understand how they can help to shape the final chapter in a patient’s life story, preventing caregivers from making promises they cannot keep, and preparing them to use all relevant information at the time decisions need to be made.

A full thirty years after the passage of the Patient Self-Determination Act, which requires healthcare providers to give patients their rights to make advance directives under state law,¹ the medical community is still debating the utility of advance directives and advance care planning (ACP) more broadly. There is marked heterogeneity in the way in which ACP is defined and its processes described.² ACP has definitively expanded beyond the completion of advance directives, designed to allow patients to specify the care they wish or do not wish to receive, to include assignment of a healthcare agent and communication between the patient and caregiver.³ Correspondingly. the objectives for ACP have grown more diverse.⁴ Nonetheless, the provision of care consistent with patients’ preferences and goals remains a core objective of ACP.⁴ Systematic reviews characterize ACP in the following ways: ACP “involves important decision making about future medical needs”⁵ and “is the process whereby patients, in consultation with healthcare professionals, family members, and other loved ones, make individual decisions about their future.”⁶ A recent prominent article arguing against ACP, claiming that it is not effective, states, “The purpose of ACP is to ensure goal-concordant care near the end of life for patients who lack decisional capacity.”⁷ Moreover, considerable time and effort has been devoted to propose and develop methods for measuring goal-concordant care.⁸

The purpose of this article is to argue that it is time to acknowledge that the provision of goal-concordant care cannot and should not be an objective for ACP. The challenges to delivering and measuring care consistent with goals for patients at the end of life have been acknowledged.⁹ However, a strong pull exists to continue to try yet newer and different ways to achieve this objective.¹⁰ If we are able to move ACP away from this objective, a wealth of evidence, clinical experience, and personal stories all support an alternative primary objective for ACP: improving caregiver outcomes.

It has been difficult to give up on the objective of providing goal-concordant care because ACP was developed to promote the preservation of patient autonomy by providing a means for individuals to continue to have control over the care they received even when they lost decisional capacity.¹¹ This was to be accomplished by having individuals express their treatment preferences. Then, the success of ACP was to be measured by assessing whether patients received their preferred treatment. However, from early on, it has argued that the expression of preferences for specific treatment interventions is essentially meaningless because interventions served radically different purposes in different clinical scenarios. Intubation for pneumonia responsive to antibiotics is a different intervention from intubation in end-stage COPD.¹² Further work went on to distinguish between the processes and outcomes of care, with the demonstration that patients make decisions based much more heavily on outcomes than processes.¹³ In other words, it was not the provision of intubation that determined an individual’s preference for receiving this treatment but rather the outcome that intubation would provide. The shift from process to outcomes was signaled by growing references to patients’ goals of care, and the objective of ACP became the provision of goal-concordant care.

This shift does not solve other factors making it impossible to demonstrate the provision of goal-concordant care. People change their minds about the treatments they want to receive and outcomes they want to achieve if faced with a serious illness.^{14, 15} These changes call into question whether individuals can pre-specify their preferences for care. In response to this concern, it has been argued that, while preferences are unstable, individuals hold core values that are stable over time, and that these values inform goals of care and can be used as the basis for end-of-life decisions.¹⁶ However, even these core values may change in the face of serious illness. Core values that inform end-of-life decisions include valuations of quality of life, including attitudes toward physical and cognitive disability and the willingness to accept tradeoffs of worsening disability for life extension. There are definitive studies showing the almost infinite ability of individuals to adapt to declining health and disability, such that states of diminished health viewed as intolerable by those in good health become increasingly acceptable by those whose own health is declining.^{17, 18} Unless measured in the days before they lose the ability to speak for themselves, the assessment of patients’ goals may not reflect this adaptation. If we cannot be sure of patient’s goals, then it is not possible to know, for patients who lose decisional capacity, whether their goals have been met.

Apart from this measurement issue, which means that achieving goal concordant care cannot be the objective for ACP, there are three important reasons why this should not be its objective. First, there are instances in which no treatment choices exist that can meet patients’ goals. For example, patients with cancer who seek life-prolongation no matter the adverse effects of treatment may simply have no options for such therapy.¹⁹ Second, patients’ best interests may change in ways dictating that treatment decisions making balance consideration of these best interests against previously expressed treatment goals. Symptoms often worsen at the end of life and be exacerbated by interventions being delivered in the attempt to achieve a patient’s goal for life extension. When patients can no longer speak for themselves but manifest signs of physical distress, the imperative of clinicians is to provide symptomatic relief, particularly as the ability to provide life extension becomes less likely. Caregivers’ best interests may also come to play an important role in decision making. Many geriatricians have encountered caregivers who, in accordance with the wishes of their loved one never to go to a nursing home but who now has dementia, are suffering adverse effects on their own health by taking care of that person at home. Given the nearly universal desire of individuals to spare their loved ones the burdens of their illnesses,²⁰ it is unlikely that they would want care at the expense of their caregivers’ health and wellbeing.

This is not to say that there are not individuals who have firm, unchanging values that can and should guide the treatment decisions made on their behalf at the end-of-life without coming into conflict with other, competing, considerations. For this most likely small group of persons, ACP will accomplish the objective of providing care consistent with a patient’s goals. Rather, the more important question is whether there is a role for ACP for the many more patients for whom this is not the case. The answer is a resounding “yes.” Decisions at the end of life are too fraught and too complex not to demand preparation. Caregivers feel the weight of being asked to participate in decisions that appear to them as causing the death of their loved one, and many carry the weight of this decision for years in the form of guilt or anxiety.²¹ In one study of caregivers for patients in an intensive care unit, the overall prevalence of post-traumatic stress disorder, measured at nine months after death or discharge, was 33% and the prevalence among caregivers who participating in decision making on behalf of the patient was 60%.²² Family members fight over decisions,²³ and this conflict can lead to complicated bereavement.²⁴

There is compelling evidence that ACP improves caregiver outcomes,² particularly if it focuses on facilitating patient-caregiver communication.^{5, 25, 26} It is possible that this communication helps caregivers to understand whether and how patients are adapting to declining health and whether they have reached a point in their illness when they are no longer willing to trade quality for quantity of life. However, even if this communication does not allow surrogates make treatment decisions based on their knowledge of “what the patient would have wanted,” it serves a critical purpose. Surrogates can be prepared to participate in choosing a broader plan of care that helps to bring the story of the patient to an end in a way that is consistent with its earlier chapters and/or writes the final chapter in a way that is consistent with how the patient wanted it to be written. This approach, the narrative interest standard, has been proposed for surrogate decision making in dementia,²⁷ but is applicable to surrogate decision making more broadly. In addition, surrogates can consider the role that their own interests and concerns will play and prepare to respond to the changing best interests of the patient. This complex mix of caregivers’ representing the voice of the patient in the context of their own needs and the patient’s best interests all contribute to the experiences caregivers have with end-of-life decision making. Caregivers’ psychological outcomes following surrogate end-of-life decision making, including complicated bereavement, PTSD, depression, and anxiety, can all be measured in straightforward ways, and these outcomes capture the influence of the multiple factors on caregiver experience.

Because these outcomes can only be measured after the death of a patient, it will be a challenge for studies of ACP in non-seriously ill populations to demonstrate effects on these outcomes, because of the need for long follow-up times. It would be ideal to have an intermediate caregiver outcome by which the success of ACP could be measured more proximately, such as caregiver preparedness. However, it is unlikely that existing intermediate measures will reliably predict adverse caregiver psychological outcomes following surrogate decision making. Caregivers frequently overestimate their preparedness to serve as a surrogate decision maker²⁸ and only recognize after the death of a loved one the conversations they wish they had and the information about the patient they wish they knew.¹⁹ It is possible that, as we learn more about how to conduct ACP with a goal of improving the psychological outcomes of caregivers after making end-of-life decisions, we will be able to develop intermediate outcome measures.

ACP will require substantial modifications in order to maximize its beneficial effects on caregiver outcomes. The very act of promoting communication between patients and their loved ones, regardless of the specific content of that communication, may be beneficial because it can contribute to a sense of closure¹⁹ and prevent the regret of things left unsaid. Table 1 provides an outline of the ACP components needed to improve caregiver outcomes. These components are an extension of the recommendations made in an earlier article describing the goal of ACP as helping patients and caregivers make “in-the-moment” decisions.⁹ However, focusing the goal of ACP squarely on improving caregiver outcomes means a de-emphasis of trying to anticipate how patients’ goals will inform specific treatment decisions and a heightened emphasis on providing the caregiver with some crucial tools for future decision making.

Table 1:

Preparing Caregivers for Surrogate Decision Making to Prevent Adverse Caregiver Outcomes

Task	Words to Use
Identify and overcome barriers to communication	[PT]: Are you able to talk with [CG] about your concerns/hopes/wishes … …if your heath were to get worse …about the end of your life Can we arrange for [CG] to come to your next appointment? [CG}: I’d like to use today’s visit so that [PT} can start to talk to you about things that are important to [PT]. While these can be hard to discuss, knowing these things will make it easier if you ever have to make a decision on [PT’s} behalf.
Encourage [PT] to tell a story of how the final chapter of life might look, exploring effects on [CG]	[PT]: Have you thought about how you would like things to go if you got very sick and it looked like you might not recover? Where care would be delivered Consideration of loves ones’ needs [CG}: I need you think about what [PT] has just said about (wanting to die at home) (never wanting to go to a nursing home). Can we talk about what could get in the way of your being able to make that happen?
Anticipate the factors that might enter into surrogate decisions	[PT]: At a time when you get very sick, the doctors might make recommendations based on the symptoms you are having or the likelihood that treatments will help. How do you feel about [CG[ using these recommendations? [CG}: How do you think you can use what you know about [PT] and what the doctors will be telling you if you need to make a decision for [PT]?

PT = Patient CG = Caregiver

After identifying and addressing the many potential barriers that prevent patients and caregivers from being able to discuss these issues, patients and caregivers need to discuss the end of a patient’s story. How does the patient want to be remembered? How much is this influenced by the very end of his/her life? Does the patient understand what the end of life looks like in an intensive care unit versus home or a hospice unit? Patients and caregivers also need to reflect on caregivers’ needs. Caregivers should not be allowed to make promises to patients they may not be able to keep, such as having the patient at home throughout their illness trajectory. If the caregiver is going to have a hard time letting go, how does the patient feel about this? What about allowing the caregiver to acknowledge the burden of caregiving? And finally, the patient and caregiver need to understand how the patient’s objective needs may influence care decisions. What will be the influence of worsening symptoms? Or worsening disability? These suggestions are only a starting point for understanding the optimal content for ACP communication. Because studies of ACP have been so heterogeneous in terms of definitions and composition of ACP and the outcomes being measured,^{2, 5} it has been impossible to draw any but the most general conclusions about associations between process and outcomes. However, it should be possible to study and compare different approaches to and content of caregiver-focused ACP communication in order to establish evidence supporting best practices.

Other questions about how best to accomplish the activities outlined in Table 1 will also need to be addressed. The most appropriate timing for conducting these activities is not yet known. On the one hand, multiple observational studies support the notion that patient and caregiver experience with serious illness and decision making serves as a strong facilitator of engaging in ACP.^{29, 30} Moreover, thinking about bringing a life’s story to its end requires that most of the life story already been written. This argues for delaying ACP until a patient is diagnosed with a serious illness. On the other hand, introducing ACP earlier, into routine care, helps overcome many of the most common barriers to engaging in ACP.³¹ It allows patients and caregivers to establish open communication prior to the added stressors imposed by serious illness. These considerations support an approach of beginning an ongoing process of ACP in late middle or early older ages and increasing the specificity of its content as a patient’s illness progresses. Also unknown is the extent to which clinician involvement will be required to facilitate patient-caregiver communication. While there are tools that patients can use outside of the clinical setting to engage in ACP,^{32, 33} the effects of these tools on overcoming barriers to patient-caregiver communication and providing caregivers with kinds of information outlined in Table 1 has not been directly studied.

Regardless of the role of the clinician in ACP, a focus on caregiver outcomes emphasizes the importance of more widespread dissemination of clinician training in serious illness communication. The best communication-focused ACP cannot help caregivers if clinicians do not skillfully participate in the difficult task of engaging caregivers in end-of-life decision making. Not only do many caregivers experience their decisions as ones resulting in the patient’s death,³⁴ but also point to uncertainty about prognosis and outcomes.³⁵ Clinicians must enter into a partnership with the caregiver to change the caregiver’s experience. This is perhaps the most important clinical scenario for “shared decision making,” since the manner in which decision making is done allows the clinician to shift the burden off of the caregiver. Clinicians can separate the process of choosing specific treatment approaches from writing the final life chapter for the patient. The surrogate is asked to help identify what the chapter should look like, and the clinician can then present the different treatment options in terms of how likely they are to write the desired chapter. Just as ACP should prepare caregivers for understanding the roles of their own needs and the patient’s best interests, the clinician must recognize these factors and describe how they are affected by different treatment options. Many of these concepts are included in the evidence-based VitalTalk communication skills program.³⁶ A caregiver-focused approach to ACP coupled with dedicated training will enable clinicians to enter into a true partnership with caregivers for surrogate decision making at the end of life.

Key Points:

• Despite an expanded scope, advance care planning continues to focus on the provision of goal-concordant care.

• This objective neither can nor should be an objective for advance care planning.

• Instead, advance care planning should focus on improving caregiver outcomes.

What does this matter:

• Advance care planning focused on caregiver outcomes looks different from current approaches.