Ethics of Research and Treatment for Neuropsychiatric Disorders: Empirical Findings

On February 25, at an afternoon symposium, 5 speakers addressed issues related to the decision-making abilities and behavior of neuropsychiatric patients and their caregivers.

The Voice of the Afflicted: The Caregiver and Patient Perspective on Treating Alzheimer’s Disease

“The key factor that drives caregivers’ willingness to slow Alzheimer’s disease (AD) is how they rate the patient’s quality of life (QOL),” said Jason H. Karlawish, MD, assistant professor of medicine, Alzheimer’s Disease Center, Center for Bioethics, University of Pennsylvania, Philadelphia.

Given the nature of AD, patients are often unable to make treatment decisions for themselves, placing caregivers in key decision-making roles. Often, the caregivers’ own concerns and issues (e.g., burden, depression and financial strain) impact the decisions they make on behalf of the patient.

Dr. Karlawish and colleagues interviewed 102 caregivers of mild-to-severe-stage AD patients, and 48 mild-to- moderate AD patients, to measure both capacity to make treatment decisions and willingness to use risky and non-risky AD medicines. Results demonstrated substantial difficulty among AD patients in making reliable treatment decisions, as evidenced by varying responses over time and poor performance on capacity measures (e.g., understanding, appreciation, and reasoning). On the other hand, caregivers received high scores on measures consistent with reliable decision-making.

Findings indicated that both patients and caregivers were willing to treat AD. The vast majority (83% of caregivers and 79% of patients) were amenable to slowing the disease when no risk was involved. Nearly half of the caregivers (49%) as well as 44% of the patients were even willing to assume risk (3% risk of GI bleeding and the need for a hospital transfusion) to slow disease progression, though patients’ responses proved less reliable over time.

Dr. Karlawish and colleagues also explored predictive factors associated with caregiver and patient treatment attitudes. For risky treatments, the most important univariate predictors of caregiver treatment preferences were caregiver assessments of patient QOL and caregiver race. Acceptance of risky treatments was only slightly decreased with increased severity of patient dementia, and in situations of caregiver depression and economic difficulty.

In “no risk” scenarios, the single predictor of caregiver willingness to accept treatment was patient QOL. Again, higher QOL ratings were associated with greater willingness to treat. Factors that predicted an unwillingness to accept treatment were: increased caregiver age, financial burden, caregiver depression, and increased severity of patient dementia. (In the patient sample, no factors emerged as significant predictors of treatment decisions.)

A QOL sub-scale analysis indicated that high caregiver ratings for patient’s overall mood, overall self, and “life as a whole” served as the best predictors of willingness to treat in “no risk” situations. A high rating of patient’s overall physical health was the one predictor associated with caregiver willingness to accept a “risky” treatment.

In conclusion, results demonstrated that decisions not to treat AD are largely associated with the caregiver’s loss of the patient as a whole person, as judged on three dimensions (e.g., self, mood and QOL), as well as negative assessment of patient health. Findings also reinforce the need to take a bio-psychosocial approach to AD treatment, and incorporate non-medical considerations. “It’s not just how you rate their QOL, but it’s also how are you going to pay these bills if they live longer,” said Dr. Karlawish.

Racial differences in willingness to accept AD treatment risk suggest possible cultural differences in concepts of dementia, surrogate decision-making, and risk tolerance. Further research is indicated.

Impaired Decisional Capacity and Research Participation Preferences
of Persons with Alzheimer’s Disease

“In our culture, we tend to put increasingly more and more decision-making on the patient or subject,” said Scott Y. Kim, MD, PhD, assistant professor, Psychiatry; director, Program in Clinical Ethics, University of Rochester Medical Center, Rochester, New York. “This creates a special burden for geriatric psychiatrists because considerable compromise in decisional capacity exists even with relatively early AD,” said Dr. Kim.

Preferences of incapacitated persons are ethically and legally important in our society. In the research setting, even an incapable subject’s objection must be respected and assent (versus consent) is required. It is important to know whether the preferences of incapacitated persons reflect their preserved values or represent impaired judgments not indicative of their values.

Assessment of preferences is difficult in persons who are already impaired, yet researchers can explore whether people with AD and normal elderly individuals make similar research partici- pation choices. Studies can also assess whether a person’s risk tolerance is impacted as their decision-making and cognitive impairment advances.

Dr. Kim and colleagues evaluated research participation preferences of 34 AD patients with mild dementia and 14 normal elderly controls by recording subjects’ willingness to participate in four hypothetical treatment vignettes of varying risk. Decisional abilities were measured using the clinical research version of the MacArthur Competency Assess-ment Tool (MacCAT-CR). Results indicated that, except for the clinical drug trial vignette (where willingness was greater for normal controls) there were no significant differences between patients and normal elderly in terms of willingness to participate. In both groups, willingness declined as risk increased.

Univariate and multivariate analyses of AD patient data revealed that willingness to participate in moderately risky research scenarios (e.g., blood draw, clinical drug trial, and Challenge Study/PET scan) went up as decision-making abilities increased. Conversely, the more impaired the respondents were, the less likely they were to volunteer for these types of studies. Age, gender, education, or Mini-Mental scores did not correlate consistently with willingness to
participate.

Findings demonstrate that in early AD, even with decisional impairment, risk-related preferences may be relatively preserved. “ People who have very mild disease may have an even higher rate of willingness to parti-cipate, which makes sense,” said Dr. Kim. “If you have the disease, you want to be helpful to research. But as a person’s disease progresses, and understanding becomes more difficult, their gut reaction may be for self -preservation. It is reassuring that increasing impairment does not appear to make patients more vulnerable to agreeing to participate in higher risk studies.”

Results are consistent with previous studies which demonstrated that even incompetent patients who have trouble understanding treatment purposes and procedures may still have reasonable preferences. In addition, findings support the current research practice of requiring proxy consent/subject assent. “The preferences of incompetent persons still matter,” said Dr. Kim. “Incompe-tence doesn’t mean that what a person says doesn’t have true moral and ethical input,” Dr. Kim concluded.

Decision-making Capacity in Middle-Aged and Older Patients with Psychotic Disorders

“Informed consent is a central tenet of ethical clinical and research practice,” said Barton Palmer, PhD, assistant adjunct professor, Department of Psychiatry, University of California, San Diego. Components of informed consent include voluntary choice, disclosure of relevant information, and decision-making capacity (e.g., the patient’s ability to understand and use information to make meaningful choices). Moreover, patient decision-making capacity (DMC) incorporates four dimensions: understanding, appreciation, reasoning, and expression of a choice.

Some older patients with psychiatric disorders may be at risk for having greater difficulty making meaningful treatment choices. Reasons include mild-to-moderate cognitive deficits and insight difficulties that can accompany these disorders, normal cognitive changes of aging, and the complexity of treatment-related decisions facing some older patients. Problems in understanding may also arise from insufficient delivery of disclosure information to patients. On the other hand, psychiatric patients demonstrate considerable heterogeneity with regard to decision-making.

Historically, DMC has not been formally assessed in most clinical settings except in situations involving legal questions of competence. More formal scales (e.g., via structured interviews) are emerging as a means of assessing DMC. Grisso and Appelbaum developed the MacArthur Competence Assessment Tool for use in treatment (MacCAT-T) and clinical research (MacCAT-CR) situations. These instruments provide assessments of all four dimensions of DMC.

Dr. Palmer and colleagues used the MacCAT -T to assess the range of treatment-related DMC in a small sample of middle -aged and older, non-demented patients with schizophrenia or schizoaffective disorder, and to determine which patient characteristics predict DMC. Results demonstrated a wide range of specific decision-making skills and abilities, similar to patterns seen in prior research involving younger patients. Few significant correlations emerged between patient characteristics and DMC, though higher education and higher performance on the Mattis Dementia Rating Scale (a measure of cognitive deficits) correlated with greater reasoning ability. Age did not appear to impact DMC on any of the four dimensions. No significant correlations were apparent for severity of symptoms, though an expected negative directional pattern was evident (e.g., more severe symptoms were associated with decreased DMC).

Significant improvements in scores over a series of repeat presentations of information using another scale (the Hopkins Competency Assessment Test) suggest that patient understanding of disclosed information can be improved, and may relate closely to the manner in which information is disclosed to the disclosee.

Future MacCAT-T and MacCAT-CR studies are ongoing, involving larger samples and more patients over age 65 (to determine whether effects of older age are more significant). Compre-hensive assessments of neuropsy- chological abilities, insight and psychiatric symptoms have been incorporated into these studies.

“Neither the presence nor absence of decision-making ability should be assumed in this population. It needs to be considered on a patient-by-patient basis, without unfairly stigmatizing people with a psychiatric diagnosis,” said Dr. Palmer. “Structured measures (like the MacCAT-T) may be helpful when evaluating a patient whose decision-making capacity is questionable,” Dr. Palmer concluded.

The “Appreciation” Component of Capacity to Consent to Psychiatric Research

“There are problems in measuring the component abilities within capacity, and problems measuring appreciation are especially significant,” said Elyn R. Saks, JD, Orrin B. Evans Professor of Law and Psychiatry and the Behavioral Sciences, University of Southern California Law School, Los Angeles.
Capacity is important in the context of psychiatric research because basic law requires that medical consent or refusal be informed, voluntary and competent to be valid. The four components or capacity are understanding, appreciating, reasoning, and evidencing a choice.

Various problems are associated with the numerous existing instruments for measuring capacity. Some tools lack uniformity in describing the four component abilities, leading to different definitions among researchers. Others either fail to draw the line between capacity and incapacity, or present problems in measuring component abilities, particularly appreciation (i.e., the formation of adequate beliefs about how information applies to oneself).

Professor Saks and colleagues developed a new instrument, the California Scale of Appreciation (CSA), that is the first to focus strictly and comprehensively on measuring appreciation. The concept of a patently false (i.e., grossly improbable) belief, rather than agreement with doctors and researchers (as in some previous instruments), provides the pivotal measure of adequacy of subjects’ beliefs. The CSA sets a cut-off score for capacity (versus incapacity) based on measures of the appreciation component.

Results from a pilot test of the CSA among older subjects with psychosis, as well as normal controls, demonstrated good interrater agreement and reliability of results on re-test. A relatively small percentage of patients (7.7%, 10.3% and 12.8%) received ratings of “incapable,” based on their acceptance or rejection of a list of beliefs presented to them. All normal subjects were deemed “capable.”

Total CSA scores were positively correlated with Mattis’ Dementia Rating Scale (DRS) Total Scores, as well as the DRS Attention and Conceptualization sub-scale scores. No significant correlations were apparent between CSA total scores and the Positive and Negative Syndrome Scale (PANSS) or Hamilton Rating Scale for Depression (HAM-D) scores.

Future studies should apply the CSA to larger patient samples and younger patients, as well as to subjects suffering from other medical conditions and
psychiatric disorders (including more acutely ill patients). Comparisons of preliminary CSA results to measures of other components of capacity (e.g., reasoning or evidencing a choice) are also indicated, as are attempts to correlate CSA scores with demographic and clinical characteristics.

“Development of instruments such as the CSA will hopefully enhance researchers’ ability to assess capacity,” said Professor Saks. “Ultimately, panels of psychiatrists, psychologists, lawyers, judges, consumers, and consumer advocates will be studied using the CSA data to determine capacity cut-offs.”