Reconsidering Gold Standards for Surrogate Decision-Making for People with Dementia

Introduction

Dementia is a common neurodegenerative illness that is characterized by a progressive decline in cognitive domains such as memory, attention, executive function, and language that ultimately affects independent functioning with respect to activities of daily living (e.g., driving, managing finances, bathing, toileting).¹ As there are a number of underlying pathologies that can drive a dementia process such as the relatively common Alzheimer’s disease, Lewy body disease, and vascular disease, the clinical course of dementia even within a single diagnosis can be relatively heterogenous, though the advanced stages are more uniform with severe cognitive and functional impairments. Given the heterogeneity in illness progression, there tends to be a variable course of impact on decision-making capacity for people with dementia, particularly those persons at mild to moderate stages. With an inexorable decline in cognitive functioning, however, surrogate decision-making becomes increasingly important and prevalent with advancing illness.²

For those deemed to lack decisional capacity for a decision at hand, there are a number of available standards to guide surrogate decision-making that can be applied in a hierarchical fashion.³ For some decisions, a person’s known wishes are available through either a previous informal discussion with a surrogate or codified formally in an instructional advance directive; if there are no known previously-stated wishes, then a substituted judgment can be made based on a person’s previously held values and preferences; if these values and preferences are unknown, then a best interests approach can be used.⁴ A key central assumption in applying a known wishes or a substituted judgment surrogate decision for a person with dementia is the presumption that these wishes, values and preferences are stable over time and still relevant to the decision at hand.^5,6

Unlike other incapacitating illnesses where there may be a clearer distinction in the transition from capacity to incapacity (e.g., a catastrophic stroke), dementia is unique in that there may be a more drawn-out graying of decision-making capacity over time; indeed, median survival after a diagnosis of dementia may be up to 12 years.¹ Available data suggests that people with dementia, even those in more advanced stages, however, are generally able to reliably communicate their values and preferences and want to participate in the decision-making process, even if a surrogate is ultimately in the role of primary decision-maker.² There is then the opportunity, in some situations, to directly compare surrogate assessments made by proxies and those made by persons with dementia themselves to see if proxies “get it right,” so to speak.

The aims of this paper are to highlight findings on the accuracy of surrogate assessments made in a variety of domains relevant to proxy decision-making for persons with dementia such as care, everyday psychosocial experiences, research participation, and quality of life. The importance of “getting it right” in proxy assessments with respect to autonomy for persons with dementia in applying advance directives and substituted judgments and reconsiderations of the gold-standard approach in surrogate decision-making for persons with dementia will also be explored.

Accuracy of proxy assessments for persons with dementia

As functional dependence increases with disease progression, aspects of personal care become a larger consideration in the day-to-day experiences of persons with dementia, even outside of more sentinel decisions such as transitions to long-term care. A number of instruments have been used to assess the values and preferences of persons with dementia. For example, the Values and Preferences Scale (VPS) was developed as a means to articulate care-related values and preferences for a person with cognitive impairment about current and future care needs.⁷ In a study population of persons with mild to moderate dementia, it was found that family caregivers consistently and significantly underestimated the importance of care-related values (i.e., autonomy, burden, control, family, safety) in their proxy assessments of the VPS relative to the self-reports of the study participants with dementia.⁸

Although the focus on care experiences may expand as dementia progresses, persons with dementia continue to live in a world of everyday experiences such as daily routines, social pursuits, and leisure activities. The Preferences for Everyday Living Inventory (PELI) is a validated psychosocial preferences assessment tool for older adults comprising items across domains such as autonomous choice, social engagement, personal growth, and keeping a routine.^9,10 In looking at the PELI as a whole in a study population of older adults with spouses as proxies, there was significantly poorer proxy agreement with self-reports of persons with dementia (Clinical Dementia Rating [CDR] score = 1) compared to those without dementia (CDR = 0).¹¹ Honing in further on specific PELI domains, a significant underestimation by proxy care partners was noted in the importance ratings of “social engagement” preferences (e.g., regular contact with family, meeting new people, volunteering) for persons with clinically significant cognitive impairment (CDR ≥ 0.5).¹⁰

In addition, in assessing responses to treatment and other interventions for persons with cognitive impairment, either in a clinical or research setting, proxy reports are often used to clarify the impact of an intervention and the overall health status of a person with cognitive impairment.¹⁴ There appears, however, to be significant discrepancy in in these dyadic ratings across a variety of different domains.¹⁵ For example in a study population of persons with Alzheimer’s dementia, proxy ratings by caregivers were noted to significantly underestimate quality of life ratings and overestimate ratings of suffering (i.e., psychological, existential/spiritual, and physical) relative to the self-reports.¹⁴ In looking at neuropsychiatric symptoms as well as quality of life in a sample of persons with mild cognitive impairment and Alzheimer’s dementia, proxy ratings by caregivers were noted to significantly overestimate depressive symptoms and apathy and underestimate quality of life and functional abilities relative to the self-reports.¹⁵ Taken together, these studies suggest that social engagement, routines, and subjective experience and care values are critical to ethical and effective treatment planning for persons with dementia.

At least in part because there are currently no disease modifying treatments available for persons with dementia and given the severity of the functional, behavioral, and financial impact of dementia, there is also tremendous pressure to elucidate better symptomatic as well as disease-modifying treatments.¹² Clinical trials are a critical means to discover novel treatments but require enrollment of persons with dementia, which can complicate the informed consent process given the actual and potential impact of cognitive impairment on decision-making capacity. As such, proxy consent has become a common practice in clinical trials involving persons with dementia to enable trial participation.¹³ Within the dyad of participant with dementia and the proxy decision-maker, however, there appears to be discrepancies throughout the decision-making process ranging from how the enrollment decision was made, to who ultimately made the decision to enroll, to the reasons for enrolling in the trial.⁴

Impacts of discrepancy in proxy assessments

The emerging evidence about person/patient-proxy decision maker discrepancy in dementia raises questions about how to understand the impact of discrepancy in surrogate decision-making and its clinical and ethical implications for people with dementia. In other words, does this discrepancy matter and should we care about it? From a theoretical perspective, we should care about discrepancy in surrogate decision-making. The driving bioethical principle underlying use of known wishes or a substituted judgement in proxy decision-making for a person with dementia is that a person’s autonomy is then preserved into the future, even in situations in which a person may be deemed incapable of making the decision at hand.¹⁶ On this account, fundamental respect for persons and humanity would require that more be done to align decisions made on behalf of incapacitated persons with their actual preferences, values, and wishes. In addition, a stronger sense of autonomy appears to be an important predictor of life satisfaction for older adults as well as a predictor of lower levels of depression and agitation,¹⁷ thereby adding additional clinical evidence in support of this ethical stance. Thus at least from a general, theoretical, and statistical standpoint, processes that reduce the sense of autonomy do have negative impacts on perceived quality of life as well as neuropsychiatric experiences for persons with cognitive impairment.

That being said, it is not always clear to what extent, if any, a statistical relationship will correlate with an appreciable clinical difference. It does appear, however, that discrepancies in proxy assessments are associated with significant clinical differences in some situations. For instance, a higher level of discrepancy between proxy ratings and self-reports on the importance of “social engagement” preferences was significantly associated with increasing depressive symptoms for persons with clinically significant cognitive impairment.¹⁰ With respect to socioemotional care preferences (i.e., companionship, activities, going out, emotional support), discrepancy between proxy assessments and self-reports for persons with dementia was also significantly associated with reports of greater dyadic relationship strain as well as worse mood for the person with dementia.¹⁸ Additionally, caregiver burden was associated with discrepancy between proxy assessments and self-reports of quality of life for persons with Alzheimer’s dementia,¹⁴ highlighting the importance of focusing on all members of the the patient-caregiver dyads and caregiver teams to align care and treatment.

Approaches toward optimizing surrogate decision-making for persons with dementia

Although a substituted judgment is considered the gold standard approach to surrogate decision-making and, in some instances the legally recommended (and even required) approach, proxy decision-makers don’t necessarily follow these guidelines in decisions for persons with dementia.¹⁹ For instance in a study of proxy decisions made by family members for persons with advanced Alzheimer’s dementia, proxies appeared to preferentially use a best interests approach instead of a substituted judgement approach in medical decision making.²⁰ Thus in addition to the discrepancies in accurately applying a substituted judgment as described above when it is applied, there appears to be a disconnect in the theoretical recommendations, ethical guidance, and legal standard and the practical approaches taken by proxies for persons with dementia.

In order to faithfully implement and optimize surrogate decision-making for persons with dementia, then, one could consider several avenues to address the discrepancies in assessment of values and preferences as well as the method of making proxy-decisions. First, in keeping with the core assumption of preferential status for a substituted judgement approach because it most closely approximates the person’s own autonomous decision were the person competent, efforts could be made to improve the fidelity in proxy assessments. From a purely theoretical standpoint, a substituted judgment would be made as follows: a decision is to be made, a person with dementia is deemed to lack capacity to make that decision, a surrogate decision-maker is sought, the surrogate uses an understanding of the person’s past preferences and values to make the decision, thus preserving the autonomy of the person with dementia into the future. This approach could be strengthened by recognizing that a proxy decision for a person with dementia does not necessarily have to be focused on the past or a grab back into a vacuum of some prior conversation or experience. Instead given the heterogeneity of dementia processes and their clinical courses, there may be ample opportunities for a proxy to review values and preferences with a person with dementia over time and into the future, particularly given that values and preferences may fundamentally change over time in the face of a life-altering diagnosis and illness.⁶

Akin to the process of advance care planning more generally, a review of the values and preferences of a person with dementia can (and should) be an iterative process over time, allowing flexibility and evolution in the setting of changing life circumstances²¹ and the ability to assess and modify treatment plans and goals fo care over time. With respect to health care decisions, a dementia-specific advance directive has been suggested to focus on goal-directed discussions in the context of various clinical stages of a progressive illness.²² This recommendation could inform and support an iterative, flexible, and person-centered approach to decision making withing person-proxy dyads and systems to promote both the autonomy and the quality of life for people with dementia. Beyond healthcare decisions, the importance of individual values has also led to a greater focus on incorporating psychosocial preferences into the everyday experiences of people with dementia, particularly those in long-term care settings, using assessments similar to the PELI to maximize person-centered approaches.²³ The upshot here is the recognition that substituted judgments and proxy decision-making are best conceptualized and can be optimized through a forward-looking, dynamic, and inclusive approach of refining preferences assessments using structured tools that rely on direct contributions from persons with cognitive impairment and a present-informed view towards the future rather than through over-reliance on historical data and perspectives..

A further consideration in optimizing surrogate decision-making for persons with dementia would be a deemphasis of the standard hierarchy of proxy approaches all together. For instance, some proxy decision-makers see no substantive difference between a substituted judgment and best interests approach in medical decision-making for people with dementia.²⁰ In looking at proxy medical decision-making more generally outside of cognitive impairment per se, it appears that this standard hierarchy of proxy approaches is often ignored by proxies, clinicians, and even patients themselves.³ The question remains then whether these standard approaches for proxy decision-making are meaningful and useful at all in the unique context of dementia. Instead of parsing through standards with apparent limited utility and effectiveness, attention could be more constructively and ethically focused on concepts and methods that have actual clinical and quality-of-life import and maximize respect for the changing perspectives and experiences of people with dementia during the course of their illness through even limited participation in decision-making.

Previously I have argued that a dementia-specific standard for surrogate decision-making is needed given the inflexibility of substituted judgments and the subjectivity in best interests assessments for people with dementia.⁶ A narrative interest standard has been advanced for surrogate decision-making for people with dementia in which there is a focus on the risks/benefits of a decision at hand while incorporating a narrative review with space for the direct contributions of a person with dementia to provide further context and flexibility for values and preferences over time.⁶ The foundational piece of this approach is the opportunity for the person with dementia to participate in the decision-making process, particularly in those settings where a proxy is the designated primary decision-maker.

Conclusions

As the cognitive impairments inherent to dementia progress, surrogate decision-making becomes an increasingly active part in the life of a person with dementia. By convention, there is a hierarchical approach to proxy decision-making with known wishes, then a substituted judgment standard, then a best interests standard. For people with dementia, it appears that discrepancy in proxy assessments is not only common but also associated with negative behavioral outcomes with respect to mood, relationship strain, and caregiver burden. In thinking about optimal approaches to proxy decision-making for people with dementia, we offer options that encourage the participation of persons with dementia either through iterative tools such as dementia-specific advance directives and preferences assessment or standards that explicitly rely on consideration of longitudinal changes in values and preferences over time for persons with dementia.

Key Points.

1. With decline in cognitive functioning for persons with dementia, surrogate decision-making becomes increasingly important with advancing illness.

2. There are a number of available standards to guide surrogate decision-making that can be applied in a hierarchical fashion: a person’s known wishes; if there are no known previously stated wishes then a substituted judgment can be made based on a person’s previously held values and preferences; if these values and preferences are unknown, then a best interests approach can be used.

3. For people with dementia, it appears that discrepancy in proxy assessments is not only common but also associated with negative behavioral outcomes with respect to mood, relationship strain, and caregiver burden.

4. In thinking about optimal approaches to proxy decision-making for people with dementia, we offer options that encourage the participation of persons with dementia either through iterative tools such as dementia-specific advance directives and preferences assessment or standards that explicitly rely on consideration of longitudinal changes in values and preferences over time for persons with dementia.

Synopsis.

As dementia progresses and cognitive function declines, surrogate decision-making becomes increasingly prevalent. By convention, there is a hierarchical approach to proxy decision-making beginning with known wishes, followed bythen a substituted judgment standard, and then a best interests standard. For people with dementia, it appears that discrepancy in proxy assessments is not only common but also associated with negative behavioral outcomes. Therefore, optimal approaches to proxy decision-making for people with dementia should instead prioritize and implement options that encourage direct participation of persons with dementia and standards that explicitly rely on consideration of longitudinal changes in values and preferences over time for persons with dementia.

Clinics Care Points.

• It appears that people with dementia, even those in more advanced stages are generally able to reliably communicate their values and preferences and want to participate in the decision-making process, even if a surrogate is ultimately in the role of primary decision-maker.

• It appears that discrepancy in proxy assessments for persons with dementia is not only common but also associated with negative behavioral outcomes with respect to mood, relationship strain, and caregiver burden.

• In thinking about optimal approaches to proxy decision-making for people with dementia, we recommend options that encourage the participation of persons with dementia either through iterative tools such as dementia-specific advance directives and preferences assessment or standards that explicitly rely on consideration of longitudinal changes in values and preferences over time for persons with dementia.