Reasoning in the Capacity to Make Medical Decisions: The Consideration of Values

Abstract

Purpose

To examine the contribution of “values-based reasoning” in evaluating older adults’ capacity to make medical decisions.

Design and Methods

Older men with schizophrenia (n=20) or dementia (n=20), and a primary care comparison group (n=19), completed cognitive and psychiatric screening and an interview to determine their capacity to make medical decisions, which included a component on values. All of the participants were receiving treatment at Veterans Administration (VA) outpatient clinics.

Results

Participants varied widely in the activities and relationships they most valued, the extent to which religious beliefs would influence healthcare decisions, and in ratings of the importance of preserving quality versus length of life. Most participants preferred shared decision making with doctor, family, or both. Individuals with schizophrenia or dementia performed worse than a primary care comparison group in reasoning measured by the ability to list risks and benefits and compare choices. Individuals with dementia performed comparably to the primary care group in reasoning measured by the ability to justify choices in terms of valued abilities or activities, whereas individuals with schizophrenia performed relatively worse compared to the other two groups. Compared to primary care patients, participants with schizophrenia and with dementia were impaired on the ability to explain treatment choices in terms of valued relationships.

Conclusion

Medical decision making may be influenced by strongly held values and beliefs, emotions, and long life experience. To date, these issues have not been explicitly included in structured evaluations of medical decision-making capacity. This study demonstrated that it is possible to inquire of and elicit a range of healthcare related values and preferences from older adults with dementia or schizophrenia, and individuals with mild to moderate dementia may be able to discuss healthcare options in relation to their values. However, how best to incorporate a values assessment into a structured capacity evaluation deserves further research attention.

INTRODUCTION

Older adults with cognitive or psychiatric disabilities, such as dementia or schizophrenia, may have difficulty making informed medical decisions for themselves. Questions about an older adult’s capacity to give consent for medical treatment may arise in a variety of clinical settings, particularly when the older patient refuses a recommended treatment, and the family or clinical team suspect that the older patient is cognitively impaired to the extent that decision-making authority should be transferred to a surrogate decision maker.

The concept of capacity to grant medical consent is derived from four legal standards, or decision-making abilities: (1) expressing a choice: the ability simply to convey a relatively consistent treatment choice, (2) understanding: the ability to comprehend diagnostic and treatment-related information, (3) appreciation: the ability to relate diagnostic and treatment information and related consequences to one’s own personal situation, (4) reasoning: the ability to compare treatment alternatives in a rational or logical manner.¹

Historically, clinical evaluations to determine capacity have suffered from unreliability and subjectivity.² In recent years, a number of assessment tools have been developed to help structure the assessment of capacity to give consent,³ with the goal of increasing the reliability and validity of these evaluations in clinical and/or research settings. These tools typically help interviewers to assess the four legal standards: understanding, appreciation, reasoning, and expressing a choice.

While recent research does suggest fair reliability and validity for capacity assessment instruments, different instruments tend not to agree as strongly in their assessments of appreciation and reasoning.⁴ These divergences likely relate to the conceptual complexity of the appreciation and reasoning constructs, as well as to significant variability in how they are operationalized and measured by the different assessment tools. Further, cognitive and rational models of decision-making capacity, derived from legal standards for consent, are at odds with contemporary decision-making science, which emphasizes both rational and more automatic, emotion-driven processes.⁵

Emotional and valuational factors have been under-emphasized in models and methods of assessing capacity.⁶ While the appreciation standard is generally recognized to include emotional and valuational components, and reasoning recognized to include the evaluation of consequences based on an individual’s values,⁷ these emotional and valuational factors have not been explicitly included in structured capacity evaluations. In this article, we first examine the reasoning standard for capacity for consent and its limitations, and then review the role of values in medical decision making. Second, we present data from a preliminary study comparing “rational” versus “values-based” reasoning in medical consent in older adults with dementia and schizophrenia. We conclude by discussing the issues raised by these preliminary findings.

REASONING IN CONSENT CAPACITY

Rational Reasoning

Current methods for assessing reasoning in consent capacity focus on how well rational or logical processes are used to compare the benefits and risks of various treatment options. Several important, though subtle, distinctions are worth noting. First, earlier standards focused on the idea of a “rational choice,” but, more recently, the rationality of the decision process, rather than its outcome, has been emphasized. Second, while it is acknowledged that individuals may not initially follow a rational or logical decision process, it is expected that they should later be able to explain their choices in a manner that reveals the major factors in their decisions and the importance assigned to them.⁸ Third, a purely rational process is not required, but rather an indication that the decisional process is not so impaired by illness or disability that rationality is seriously or negatively influenced,⁹ which is also described as being able to produce “recognizable reasons” for one’s decision. Reasoning includes the ability to evaluate consequences in light of one’s values; however, how to elicit or evaluate those values is not often addressed.

Reasoning is operationalized differently on the various instruments used to assess patients’ capacity to consent. The MacArthur Competence Assessment Tool for Treatment (MacCAT-T) rates the patient’s ability to compare treatment risks and benefits and their consequences, to generate everyday consequences of treatment, and the logical consistency of the treatment choice with these statements.¹⁰ The Hopemont Capacity Assessment Interview (HCAI) asks why a particular choice was made, with credit for risks and benefits considered.¹¹ The Capacity to Consent to Treatment Instrument (CCTI) solicits all the reasons for a patient’s choice, and credit is given for the total number and accuracy of reasons provided.¹² Not surprisingly, given the different methods of measurement, these approaches have only modest convergent validity; that is, these different measures of reasoning correlated only moderately when administered to a sample of older adult patients.¹³

“Rational” and “Non-Rational” Decision Making

These capacity assessment instruments define reasoning in terms of the law surrounding capacity and tend to focus on a patient’s cognitive capacity to process and evaluate information. In contrast, theory and research regarding human decision making suggest that humans do not approach decision making in a strictly logical, conscious, or rational fashion.¹⁴ Rather, humans make decisions based on rational, as well as unconscious, automatic, and emotion-driven processes. With life experience, individuals may use a heuristic approach with rules that shortcut the decision-making process.¹⁵ Certainly, in making decisions about negative events such as illness or injury, individuals evaluate the severity and form of their expectations about a current situation based on previous experience and associated schemas, or beliefs and ideas about the world, in a fairly automatic way.¹⁶ Research on aging and expertise suggests that experience may lead to highly efficient decision making when individuals seek and utilize only the information that is most relevant to them.¹⁷ In fact, some postulate that, in more complex decision scenarios, people may make better decisions using a more “intuitive” approach, rather than a more conscious, deliberate decision-making approach.¹⁸

While such decision-making shortcuts may lead to entirely valid decisions, human decision making is also vulnerable to a range of cognitive and emotional biases.¹⁹ Therefore, “non-rational” decision making may lead to either appropriate or inappropriate decisions for an individual. It may be that older adults who perform poorly on the measurement for reasoning during a capacity assessment are, in fact, decisionally impaired, while others may be using “non-rational” strategies to reach an appropriate and valid decision for themselves. Per Marsiske and Margrett, who reviewed the literature on aging and decision making in 2006, “adult problem solving is not always strictly rational. In the real world, when pondering life dilemmas, individuals also consider their own preferences, values, and feelings as well as those of their social partners and cultural context, in addition to purely cognitive considerations …. Thus, scoring adult problem solutions on strictly rational grounds may reveal age differences; however, such differences may not necessarily imply age deficits because they could instead reflect qualitative changes in problems solving ….”²⁰

Values-Based Reasoning

Not all commentary on reasoning in medical consent focuses on rationality of reasoning. For example, the seminal U.S. 1982 President’s Commission for the study of Ethical Problems in Medicine and Biomedical and Behavioral Research defined capacity in terms of the various cognitive standards previously described (communication, understanding, reasoning, deliberation), but also as the possession of a set of values and goals. The commission states that a set of values and goals is foundational to the comparison of treatment alternatives.²¹

Similarly, the American Bar Association’s 2002 Model Rules of Professional Conduct delineates the factors that must be balanced in determining capacity and include “the consistency of a decision with the known long-term commitments and values of the client,” (rule 1.14, comment 6).²² Berg and colleagues reviewed standards for informed consent and concluded that, from an ethical perspective, the most justifiable capacity standard is one that promotes autonomy, protects welfare, and may be employed by practitioners with a minimum of bias. In their words, the decision maker “must first understand and appreciate the risks and benefits of options presented and then weigh those options to make a decision in light of her own values.”²³ Certainly, strongly held values, based in life experience and belief systems, may influence an individual’s reasoning and decision making in both automatic and more explicitly conscious or rational ways. In either case, values-based reasoning appears an important aspect of decision-making capacity that has not been explicitly emphasized in the literature on the assessment of capacity to consent.

VALUES IN CONSENT CAPACITY

While many commentators recognize the importance of values in informing rational reasoning, most commentary ends without elucidating what “values” are and how might they be assessed with regard to decision-making capacity. Values-assessment strategies for the purpose of advance medical and long-term care planning are available,²⁴ but there is no research on strategies for integrating structured values assessment into evaluations of clinical capacity, or even whether this approach would be worthwhile. There are at least three dimensions of healthcare values that may be relevant to assessing capacity: (1) personal perspectives on quality of life in terms of one’s abilities, activities, and relationships; (2) beliefs on preserving quality versus length of life; and (3) preferences for one’s role in decision making (ranging from autonomous, through shared responsibility, through complete deference to others).

Impact of Choices on Valued Abilities, Activities, and Relationships

In evaluating treatment choices, individuals may consider how various treatment outcomes influence their own perceived quality of life; for example, whether a certain outcome might be viewed as a state “worse than death.”²⁵ To a greater or lesser degree, individuals may be concerned about pain, self-sufficiency, cognitive clarity, physical mobility, bodily integrity, relational ability, sexual functioning, and so forth. Such values ratings are predictive of treatment choices.²⁶ Similarly, treatment choices can be based on how they are likely to affect valued relationships. Patients are often very concerned about the impact of their illness and treatment on loved ones. For example, many older adults express concern about being a burden to their families, and the possibility of creating a financial burden for one’s family may influence decisions about receiving intensive care.²⁷

Preserving Quality versus Length of Life

Regarding end-of-life treatment decisions, a variety of cultural, religious, and individual factors contribute to individual differences in preferences for preserving life versus quality of life. Individuals with higher religiosity may favor treatment choices aimed at preserving life. African-Americans and Hispanic-Americans are more likely to prefer life-sustaining treatments compared to Whites. Those who value quality of life over length of life are more likely to refuse life-sustaining treatments.²⁸ Such beliefs about treatment goals could constitute implicit “rules of thumb” that individuals may use – but not necessarily articulate – in complex decisions.

Preferred Role in Decision Making

An individual’s decision-making style may influence the level of motivation to participate in, or even the perceived relevance of, an assessment of his or her capacity to make decisions. Cultural, generational, and personality differences influence the extent to which individuals wish to have control over making their own decisions. Individual autonomy is not a primary value for all people. For example, older cohorts are more likely to assume that family members or physicians will make medical decisions for them and take a less active role in seeking illness-related information.²⁹ Many non-Western cultural groups, including Asian- and Hispanic-Americans, believe that the individual patient should not be burdened with diagnostic or prognostic information and/or that decision-making responsibility belongs to, or should be shared with, the family,³⁰ or left to the will of God. Further, personality issues related to trust, dependency, and risk tolerance may affect an individual’s decision-making style and preferences. Therefore, many patients may not be comfortable reasoning actively about medical treatment decisions.

GOALS OF THIS STUDY

In this study we aimed to develop and explore the concept of “values-based reasoning” and its relationship to classic rational reasoning in older adults with and without neuropsychiatric illness. In particular, we aimed to articulate and investigate ways that values may be incorporated more explicitly into the process of assessing capacity. We hypothesized that values may affect the assessment of reasoning in three ways. First, assessment of reasoning may reflect a patient’s ability to describe how treatment choices are consistent with valued activities and relationships. Second, assessment of reasoning may reflect not only the consideration of specific consequences associated with a treatment choice, but also a belief about the relative values assigned to quality of life versus length of life. Third, assessment of reasoning may reflect an individual’s preferences regarding the deferral of decisions to others, or to religious views (for example, to “God” or to “fate”), and therefore the individual may be less engaged in a “rational reasons” decision-making process.

METHODS

Recruitment

In a study designed to examine a new Assessment of Capacity to Consent to Treatment (ACCT) interview, individuals with dementia (n=20) or schizophrenia (n=20), and a cognitively healthy comparison group (n=19), were recruited from the outpatient clinics of Veterans Affairs (VA) Medical Centers in Boston and Salt Lake City.³¹ Due to the nature of the older VA population in these locations, all participants in this study were male. Inclusion criteria were: (1) age 60 years or more; (2) primary language English, and (3) ability to participate in a one-hour interview. Participants in the dementia group were referred mostly from specialty outpatient geriatric clinics. Participants in the schizophrenia group were referred mostly from specialty outpatient psychiatric clinics. Participants in the cognitively healthy comparison group were recruited from primary care clinics; those who carried a clinical diagnosis of dementia or schizophrenia, or who had a MMSE score lower than 26, were excluded.

Participants in the dementia group ranged in age from 65 to 88 years (M=77.95, S.D.=6.78), with a mean Mini Mental State Examination (MMSE) of 26.42 (S.D.=3.67), out of a possible score of 30. Participants in the schizophrenia group ranged in age from 60 to 93 (M=70.85, S.D.= 8.68), with a mean MMSE of 27.05 (S.D.=1.82). The comparison group ranged in age from 61 to 83 (M=74.42, S.D.=6.55), with a mean MMSE of 28.21 (S.D.=1.44). Age varied significantly between the groups (F=4.58; p=.01); post hoc comparisons indicated that the mean age was significantly lower in the schizophrenic group compared to the dementia group (t=2.88, p=.006), but statistically equivalent in other group comparisons. However, there was no relationship between age and the dependent variables including reasoning scores, and values regarding quality of life and decision making. MMSE scores did not differ significantly between the groups, although there was a trend for group differences (F =2.51; p =.09).

Procedures

The study was reviewed and approved by the institutional review board (IRB) at the VA Medical Center in Boston and at the VA Medical Center in Salt Lake City, and was additionally reviewed and approved by the IRB at Harvard Medical School. All participants provided written informed consent. One patient with schizophrenia had a guardian and, in that case, consent was obtained from the guardian with the patient’s assent. Brief cognitive and psychiatric assessments were completed first, followed by the capacity interview. Primary care physicians or nurse practitioners of the participants with dementia or schizophrenia provided their clinical judgment regarding the participants’ medical decision-making capacity (N=26). Participants were compensated for their time.

Instruments

Cognitive and psychiatric assessment

Cognitive status was assessed with the Modified Mini-Mental State Exam (3MS), an extended version of the MMSE.³² The 3MS yields a conventional MMSE score (maximum score=30) or an extended score (maximum score=100), and includes additional measures of executive function and delayed memory. Psychiatric status was assessed with four subscales from the Brief Symptom Inventory (BSI): anxiety, depression, paranoia, and psychosis.³³

Capacity assessment

Capacity to consent to medical treatment was evaluated with the Assessment of Capacity to Consent to Treatment (ACCT) interview.³⁴ The ACCT was developed based on our previous research comparing several tools to assess medical decision-making capacity.³⁵ The ACCT assesses capacity in two steps. The individual is interviewed about values that potentially may be related to healthcare decisions, and is then asked to consider three hypothetical scenarios and make a hypothetical treatment decision in each case. Decision making is evaluated in terms of four conventional legal standards. A detailed manual was utilized to standardize scoring. Preliminary evidence for reliability and validity was good. Cronbach internal consistency reliability was a=.96 based on all capacity items (56 items) as calculated in the present sample. Internal consistency reliability was a=.82 for reasoning, 12 items across three vignettes.

Three treatment vignettes were developed for the ACCT for research purposes. In this article we describe data for the third vignette only (treatment of a non-healing toe ulcer with amputation or bypass surgery), as only this vignette was administered to the control group sample. The choice for bypass surgery is described as an incision down the leg to insert a new artery. The benefits of this treatment are described as increasing the blood supply and saving the toe; the risks are described as a chance of death during surgery (as with any major surgery), and the need for assistance from others during approximately six weeks of recovery. The choice for amputation is described as a surgeon cutting off the toe, a relatively quick procedure. The benefits are described as getting rid of the infection without major surgery, whereas the risks are described as a need to use a cane and subsequent difficulty with balance.

Assessment of values

Questions were developed to address the three major domains previously identified: valued activities and relationships, focus on quality versus length of life, and preferred role in decision making (see table 1). Items were based on a literature review³⁶ and pilot studies of various forms and items.³⁷ Participants rated their “top three” valued abilities/activities as our pilot studies found that, without such forced choice, many participants selected every ability/activity as important.

Table 1.

Values Interview on the ACCT

Domain	Approach
Valued abilities/activities, and relationships
Valued abilities	Which three things are most important to your life or make your life worth living?1
	To take care of myself (for example, bathing, dressing); not have to depend on others for help with daily life To walk or move around by myself To live at home To think clearly about things To make my own life decisions (for example, about health, finances, housing) To have relationships with family and friends To practice my religion or spiritual life (faith, prayer) To live without significant pain or discomfort To do specific activities or hobbies that I enjoy (for example, reading, TV, gardening)
Valued relationships	State the names/relationships of those people to whom you are most close.2
Preserving quality versus length of life
Quality of life3	The quality of my life is more important than how long I live. Living as long as possible is more important than the quality of my life.
Medical interventions4	If I were very sick, I would like to do everything possible to prolong my life. If I were very sick, I would like to let nature take its course.
Preferred role in decision making
Deference to others5	When you make an important healthcare decision, how much input do you want from the doctor? When you make an important healthcare decision, how much input do you want from family?
Religious beliefs6	When you make an important healthcare decision, how much do your religious or spiritual beliefs influence your decision?

NOTES

¹Participants are asked to select three items as this helped to prioritize the activities for later comparison.

²This question was added to the ACCT after pilot testing. In the initial study this information was elicited more informally.

³The item was rated on a five-point scale ranging from “very true” to “very false.” The item was asked in two directions, and a summary score was created reversing scoring on the second item.

⁴Ibid.

⁵The item was rated on a five-point scale ranging from “I want to make the decision myself” to “I want my doctor (or family) to make the decision entirely.”

⁶This item is rated on a five-point scale ranging from “not at all” to “completely.”

Assessment of reasoning

For each vignette, reasoning is assessed with four items (see table 2). After naming choices and selecting one choice, the individual restates key risks and benefits for him-or herself. In our measure, full credit was given if participants could identify at least two risks, benefits, or reasons for the preferred choice. The individual then describes how the choice would impact valued abilities or activities, and, next, relationships. Finally, the individual demonstrates the consistency of the choice by explaining how the selected option is better than the alternative, in terms of the answers just provided. The first and last items represent classic approaches to assessing “rational reasoning,” whereas the second and third items attempt to capture “values reasoning.” Items are scored 0, 1, or 2. Full credit responses are detailed in table 2. One point is given for a vague but technically accurate answer such as “the benefits are greater than the risks” (without providing additional information despite prompting). A score of 0 is given for answers that are incorrect or non-responses such as “I cannot name any risks or benefits.”

Table 2.

Reasoning Items

Aspect of Reasoning	Description	Item*	2-Point Response
Rational reasoning	Providing risks and benefits	1. What risks and benefits did you consider when making that decision?	States 1 risk and 1 benefit.
	Providing comparisons	4. You think that [choice] is best. Tell me why _____ seems better than _____.	Compares choices stating at least 1 difference.
Values reasoning	Explaining impact of choices on valued activities	2. Looking at the list we made earlier, what are the ways, if any, that [choice] could impact things critical to you?	Explains which activity might be impacted by choice and how; or if not, why not?
	Explaining impact of choices on valued relationships	3. How, if at all, would [choice] affect the people who care about you?	Explains how relationships may be impacted by choice; or if not, why not?

^*Item number represents suggested order of administration on the ACCT.

RESULTS

Values: Group Comparisons

Valued abilities, activities, and relationships

Participants were asked to select the three most important valued abilities or activities from a list of nine. Selecting the top three made it easier to ask participants about values-based reasoning later on. When all groups were combined, the most highly rated items were “to have relationships with family and friends” (48 percent), “to live without significant pain or discomfort” (44 percent), and “to take care of myself” (41 percent). Groups differed significantly on the items of “living at home” (50 percent of dementia group chose this item, compared to 15 percent or less of the other groups) and “having relationships with family and friends” (only 20 percent of schizophrenic group chose this item, compared to 55 percent or more of the other groups). Top choices for participants in the schizophrenia group were “living without pain” (50 percent) and “practicing religious/spiritual life” (40 percent).

Quality versus length of life

Participants rated their views on the importance of quality versus length of life with two items, which were combined for analysis. Overall, participants rated “quality” as more important than “length of life” (mean ratings for both items falling in the “mostly true” range). Participants in the primary care group rated “quality” most highly, followed by participants with dementia, and lastly by participants with schizophrenia. Differences between the primary care group and the schizophrenic group were statistically different (p<.01) in post hoc comparisons. Ratings for the desire for medical intervention to prolong life did not differ between the groups, with ratings for both items falling in a middle range. Higher ratings for “quality of life” were associated with higher MMSE scores (r=.32, p<.05), but not BSI scores.

Preferred role in decision making

Participants rated preferences for autonomous, shared, or delegated decision making on a scale of 0 to 4, with 0=“I want to make the decision myself,” 4=“I want my doctor (or family) to make the decision entirely,” and 2=“I want to make the decision together.” There were no differences between the groups in rating how much input they desired from doctors or families. Mean ratings reflected a general preference for collaborative decision making with doctors (overall mean of 2.02) and family (overall mean of 1.55). However, 5 percent of the entire sample wanted to make decisions entirely alone (without input from doctors) while 15 percent wanted doctors to make the decision mostly or entirely. Decision-making preferences were not associated with MMSE scores or BSI scores.

Participants also rated to what degree religious or spiritual beliefs influence their decisions, with a rating of 0 being “not at all” and a rating of 4 being “completely.” There were no group differences for the impact of religious or spiritual beliefs on treatment decisions, with mean ratings falling somewhere between “a little” (1) and “somewhat” (2); all three groups showed considerable variability on this item. Eight (14 percent) of the participants said that religious or spiritual beliefs would influence their medical decisions “completely.” Higher ratings for the influence of religious or spiritual beliefs on medical decision making was associated with higher BSI total scores (r=.27, p<.05) and BSI paranoia subscale scores (r=.33, p<.05), but not MMSE.

Reasoning Ability

Group comparisons

Participants were rated on two items that assess reasoning as classically defined — that is, providing rational reasons or comparing choices. As depicted in table 3, individuals in both the schizophrenia group and dementia group performed significantly worse than the primary care comparison group on these items. Participants were also rated on two items that assess reasoning in terms of values – that is, to justify their choices in terms of the valued activities and relationships. With respect to reasoning in terms of valued abilities/activities, participants in the schizophrenia group were impaired compared to both of the other groups, but dementia and primary care groups were equivalent on this item. In contrast, participants in the dementia and schizophrenia groups were impaired compared to the primary care group on reasoning in terms of valued relationships.

Table 3.

Mean Group Differences on Reasoning Items

Item	N	df	Schizophrenia Group		Dementia Group		Primary Care Group		Between Groups F
			Mean	S.D.	Mean	S.D.	Mean	S.D.
Rational
Providing risks/benefits	59	2,56	1.35	.88*	1.35	.75*	1.89	.32	3.98	p=.024
Rational
Providing comparisons	59	2,56	1.15	.88*	1.40	.88*	2.00	.00	7.06	p=.002
Values
Impact on activities	58**	2,55	0.89	.81* +	1.50	.83	1.89	.46	9.28	p<.001
Value
Impact on relationships	59	2,56	0.95	.89*	1.10	.91*	1.84	.50	7.01	p=.002

Note: each item has a potential range of 0–2.

^*Dementia or schizophrenic patient group mean significantly less than primary care comparison group mean in post hoc analyses.

^**One patient in the schizophrenic group was unable to answer this question.

⁺Schizophrenia group mean significantly less than dementia group mean in post hoc analyses

Relationship between reasoning and values variables

Overall, 63 percent of the participants chose amputation for treatment for the (hypothetical) non-healing toe ulcer; 27 percent (n=16) chose bypass surgery, and 7 percent (n=4) were unable to decide. Of the nine valued abilities/activities, two would hypothetically be impacted by the treatment choice. A decision for amputation would make walking more difficult, but taking care of oneself (during the recovery period) easier. Of those who chose amputation (n=37), 19 percent (n=7) rated walking as one of the three most important life activities, compared to those who chose bypass surgery (n=16), of whom 44 percent (n=7) rated walking as one of the three most important (X² = 3.54, p=.06). However, the groups who chose different treatments did not differ on the ratings of the importance of taking care of oneself.

Those with higher ratings for “quality of life” versus length of life, had higher rational reasoning (r=.48, p<.001) and values reasoning (r=.28, p=.05).

The desire for more input from doctors was not associated with reasoning scores, whereas the desire for more input from families was modestly correlated with higher values reasoning (r=.28, p<.05). The importance of religious or spiritual beliefs in making medical decisions was not correlated with either measure of reasoning. However, comparisons of eight individuals who said that religious or spiritual views inform their decisions “completely” to the remainder of the sample revealed lower values reasoning (t=2.19, p<.05).

DISCUSSION

The conceptualization and evaluation of capacity to consent have been greatly aided by the work of numerous scholars.³⁸ According to these seminal contributions, a hallmark of capacity to consent is reasoning, defined by the ability to convey a rational or logical decision-making process. The research literature, as well as our own clinical experience, however, suggest that an assessment of reasoning that is based on the ability to articulate a logical decision-making process may underestimate the capacity of individuals whose decisions are guided by strong underlying values and preferences. Clearly, people make difficult decisions about medical treatment in accordance with their unique values, beliefs, fears, hopes, and social and cultural contexts. Ideally, reasoning entails a process of weighing the risks and benefits of different options in accordance with all of these influences. Research on decision making finds, however, that long life experience and/or strongly held values or emotional reactions provide a basis for “intuitive” decision making rather than patently conscious, deliberate, or “rational” processes. Further, individuals differ greatly — according to cultural, cohort, or personality influences — in their motivation to participate in autonomous medical decision making and may thus appear to demonstrate poor reasoning ability. To date, structured assessments of patients’ capacity to consent have not explicitly included evaluation of the values and beliefs that individuals bring to the process of making complex medical treatment decisions.

This study is limited by a number of factors, including a small and demographically non-representative sample including only male veterans. The participants with dementia were in early stages of the illness, and our observations about their performance cannot be generalized to a broader group of individuals with dementia. However, the preliminary results raise a number of interesting possibilities and suggestions for further research. In the introduction we posited three ways that the issue of values may arise in the assessment of capacity for medical consent. We will consider each of these in turn, noting what we learned about the feasibility, potential validity, and limitations of assessing values as part of an assessment of reasoning specifically, and consent capacity more broadly.

Impact of Choices on Valued Abilities, Activities, and Relationships

A clinician might approach an assessment of reasoning by first learning what activities or relationships are most important to the individual, and then asking the individual to explain why a choice is preferred in terms of its impact on those valued activities or relationships. We found it was possible to elicit this information from older adults, including those with mild dementia or schizophrenia, consistent with our previous research. ³⁹ Of note, even in this relatively homogeneous sample, there was substantial individual and group variation in valuation of basic activities to daily life. We found it possible to use this values knowledge to inform an inquiry into reasoning.

Ideally, an assessment of values-based reasoning would demonstrate that an individual’s treatment choice was consistent with her or his values regarding life abilities/activities and relationships. In this preliminary study, we were limited in our ability to demonstrate this link, given that the one hypothetical vignette explicitly addressed only two of the valued activities listed (and individuals could choose only their top three). Even with this limitation of the data, we did find that individuals who chose amputation of the toe (which would cause enduring problems with balance) were significantly less likely to list ability to walk as one of their top three valued activities.

As expected, we also found that participants with schizophrenia or dementia had more difficulty with both “rational” reasoning and “values-based” reasoning, compared to a comparison group, with one exception. Older adults with dementia were as able as the controls in the study to explain their hypothetical treatment choice in terms of valued abilities and activities; those with dementia had greater difficulty with “rational” reasoning, compared to the comparison group, as expected. While this finding is preliminary, it is intriguing to consider that adults with mild to moderate dementia might have trouble reasoning through complex treatment scenarios, but may be better able to participate in medical decision making by focusing on the values that are most important to them and explaining treatment decisions in terms of those values.

Operationalizing the assessment of values-based reasoning is a complex challenge for a structured capacity tool — a method that enhances reliability, and is necessary for research on capacity, but lacks the flexibility and fluidity of a clinical interview. Our attempt to simplify and structure the consideration of valued abilities/activities for the patient restricted the possible menu of values that an individual might point to in the first place, and resulted in a cognitively challenging task. Future efforts to incorporate values assessment into a structured capacity assessment might allow patients to add salient “other” personal values, and help patients to “discard,” early on, those treatment risks and benefits that have no consequences in terms of valued activities or relationships.

One challenge for explicitly eliciting and incorporating values into an evaluation of capacity is the question regarding the authenticity or “competence” of values. Sometimes expressed values are symptoms of mental illness (for example, delusional distrust and desire to distance oneself from a family member, or belief that one is worthless and should die) or are ethically unacceptable (for example, a desire to rape others).⁴⁰ In a discussion of the evaluation of capacity to consent in anorexia nervosa, in which some values that influence treatment decisions could be considered “pathological,” questions arose as to what extent the appreciation standard was able to capture the illness-related beliefs and values about body image and, ultimately, life and death decisions.⁴¹ These conceptually and ethically complex considerations about the authenticity of values points to the limits of an oversimplified inclusion of values into a capacity evaluation and the importance of considering a structured capacity evaluation in the context of a thorough clinical evaluation.

Preserving Quality Versus Length of Life

We also considered that individuals might have general “rules of thumb” that would influence their decision making — in particular, strong views about quality of life being more important than length of life (or vice versa). We have observed in our own work that, as some older patients become more medically infirm and more functionally disabled, they are less inclined to desire medical interventions. ⁴² This reluctance to seek further curative treatment can create conflict when caring, impassioned physicians want to intervene to help these patients. Such patients may refuse medical intervention without “rational reasons,” instead citing a general view such as, “I just don’t want any more surgical interventions.” On the other hand, there are patients and/or families whose goals are clearly to prolong life as long as possible, due to religious beliefs, other moral considerations, or difficult emotional/interpersonal considerations.

In our study we found that all participant groups could describe their views on quality versus length of life. Individuals who placed a higher emphasis on quality over length of life tended to receive higher reasoning scores, although the reasons for this relationship are not clear. Perhaps individuals tuned into maximizing life quality more strongly engage in a process that asks them to weigh various treatment options, whereas individuals with strong preferences to prolong life are guided in a more direct or intuitive manner to a treatment choice that seems consistent with that belief. Again, the vignette presented in this study did not present a tension between quality versus length of life; a vignette focusing on a life-sustaining treatment decision may have shown a significant relationship in beliefs and treatment choices regarding quality of life versus length of life. Regardless, it may be helpful to understand patients’ perspectives on these issues as part of a capacity evaluation including, to the extent possible, the basis for strongly held beliefs about prioritizing quality versus length of life (for example, religious or cultural beliefs, or long-standing struggles with illness).

Preferred Role in Decision Making

A third way in which values may relate to the assessment of reasoning is that individuals who defer to others or to religious beliefs in decision making may demonstrate lower levels of reasoning. The capacity assessment model assumes that an individual is an independent decision maker, and asks him or her to make a decision independent of the usual supportive and collaborative “shared decision-making” paradigms, despite research that individuals vary considerably on this dimension. In our clinical experience, we find that many patients are perplexed by even being asked, “Which treatment would you want?,” as they presume that the doctor is the one with the medical expertise, and because they were socialized during a time when making decisions about treatments were entirely within the hands of a doctor.

We found it quite feasible to ask patients about their preferred role in decision making. However, these ratings were not associated with their reasoning scores, in part because most of our sample preferred collaborative decision making. This issue may be best approached as an “outlier” phenomenon, meaning that a clinician should be mindful of an individual who, for historical or cultural reasons, prefers not to make independent medical decisions, and the ways that this may influence the capacity assessment process. For older adults who prefer others (doctors, family) to make decisions for them — certainly a valid choice if the elder is capable of understanding the responsibility he or she is giving to another person — we should remain aware that such patients may be at risk for accepting treatments (including investigative treatments) without fully understanding potential risks.

We found that individuals were able to express the extent to which they believed religious views would influence healthcare decisions and that, even in this homogenous sample, the influence of religious beliefs varied widely. Overall, these ratings were not correlated with reasoning, although individuals at the extreme end (religious beliefs inform decisions entirely) had lower scores on one measure of reasoning. It is not necessarily clear whether these individuals had poorer reasoning ability, or whether our assessment of reasoning failed to capture the valid religious reasons that underlie treatment decisions.

A real challenge for considering the role of religious values in evaluating capacity to consent is the individual whose strong religious beliefs may be associated with an underlying psychiatric condition. In this sample, higher ratings of religious beliefs influencing treatment choices were associated with higher levels of BSI paranoid symptomatology. For example, one participant with schizophrenia refused a hypothetical treatment intervention, stating, “I don’t believe in all that science and pills curing. I think it is witchcraft. I believe God will cure me.” This example underscores that capacity assessments must consider the individual and his or her values, the social and cultural context, the decision-making process, and the individual’s medical and psychiatric diagnoses and their impact on symptoms and functioning. While assessment instruments may help to provide an understanding of each of these factors, the integration of the data requires a full clinical evaluation and, ultimately, professional clinical judgment.⁴³

CONCLUSION

This preliminary study illustrates the feasibility, challenge, and potential usefulness of integrating values assessment into the evaluation of reasoning for the capacity to consent. Further research may examine expanded methods for assessing the values that are relevant for the capacity to give medical consent as well as elucidate the relationships between values, emotions, personality, cognition, and decision-making capacity. Additionally, further study may help to clarify both the appreciation and reasoning constructs for consent capacity, the influence of emotion and values in each process, and how appreciation and reasoning are related. For example, appreciation, or appraisal of the personal relevance of diagnostic and treatment information, likely involves emotional processing, and may be necessary for values-based reasoning.⁴⁴ Further research may also explore how the values of the clinical evaluator may influence the capacity evaluation process.⁴⁵ Ongoing research on consent capacity has the potential to illuminate the highly complex concepts involved in capacity assessment — concepts informed by interesting and sometimes confounding intersections of the law, biomedical ethics, the psychology of decision making, and socio-cultural influences.