A Prototype Instrument

Daniel C. Marson. JD, MD.

Kellie K. Itigraiii

Heatlici A. Cody

Linda E. Harnell, MD,PhD

Objective: To assess empirically the competency of patients with Alzheimer's disease (AD) to consent to medical treatment under different legal standards (LSs).

Design: Comparison of normal older subjects and patients with AD on measures of competency to consent to medical treatment.

Setting: University medical center.

Subjects: Normal older control subjects (n=15) and patients with probable AD (n=29) 15 with mild and 14 with moderate AD.

Main Outcome Measures: Two specialized clinical vignettes were developed that test a subject's capacity to consent to medical treatment under five well established LSs for this competency: LS I, evidencing treatment choice; LS2, making the reasonable choice; LS3, appreciating consequences of choice; LS4, providing rational reasons for choice; and LS5, understanding treatment situation and choices. Performance on the LSs was compared across control and AD groups using Student's t test, x2, and analysis of variance. Demented subjects were categorized as competent, marginally competent, or incompetent under each LS by using a cutoff score derived from normal control performance.

Results: No differences between groups emerged for LSI and LS2. Control subjects performed significantly better than patients with mild AD on LS4 and LS5, and significantly better than patients with moderate AD on LS3, LS4, and LS5. Patients with mild AD performed significantly better than patients with moderate AD on LS4 and LS5. With respect to competency status, patients with AD showed a consistent and progressive pattern of compromise (marginal competence or incompetence) related to dementia severity and stringency of the LS.

Conclusions: A reliable prototype instrument validly discriminated the competency performance and classified the competency status of control subjects and patients with mild and moderate AD under five LSs for competency to consent to medical treatment. While the groups performed equivalently on minimal standards requiring merely a treatment choice (LS or the reasonable treatment choice (LS2), patients with mild AD had difficulty with more difficult standards requiring rational reasons (LS4) and understanding treatment information (LS5), and patients with moderate AD had difficulty with appreciation of consequences (LS3), rational reasons (LS4), and understanding treatment (LS5). The results raised the concern that many patients with mild AD may not be competent to consent to treatment and supported the value of standardized clinical vignettes for assessment of competency in dementia.

COMMENT

With the increasing prevalence of dementia, issues of competency have taken on heightened importance. While concerned about maintaining the autonomy and well being of its aging citizens, society is equally interested in protecting them (and others) from risks and dangers caused by their declining capacities. These competing social interests are particularly prominent in the treatment of AD. While patients with AD benefit from advances in treatment and research, their progressive cognitive impairment also makes them vulnerable to potential abuses. The question of whether a patient with AD is competent to consent has, therefore, become increasingly important. There is a critical need at the present time for standardized instruments to guide competency assessment of patients with dementia. Such an instrument must be reliable and valid, relatively easy to administer, and acceptable and comprehensible to patients, families, and health care and legal professionals. Such an instrument must also specifically test the functions integral to the capacity to consent, in the appropriate context for such consent. A diagnosis of AD, which conveys no information specific to the capacity to consent, cannot be dispositive of competency questions. Similarly, cognitive screens like the MMSE or more discrete neuropsychologic tests cannot serve as direct measures of competency because they do not yield sufficient capacity specific information. Neuropsychologic testing can, however, help determine the degree of cognitive impairment, reveal cognitive threats to capacity, indicate specific interventions to support capacity, and be used to investigate competency loss.

Clinical vignettes represent an innovative approach to competency assessment. Although hypothetical, a clinical vignette can approximate the "real life" physician patient assessment situation," and can satisfy face, content, and construct validity to a high degree. Vignettes can also incorporate different LSs, or thresholds for determining capacity to consent. The availability of different thresholds is important, insofar as it allows decision makers the flexibility to vary the standard for competency, depending on a variety of factors, particularly the risks and benefits of different treatments. The use of accepted LSs in clinical competency assessment also represents an important bridge between advances in medicine and law.

Results of the present study suggested that our vignette instrument reliably and validly tested the competency of patients with AD at different legal thresholds. The instrument elicited information relevent to treatment consent capacity and a reliable scoring system clearly discriminated the performance of the normal Control and AD groups as well as the control, mild AD and moderate AD subgroups on the three most stringent standards (LS3 through LS5). Assignment of competentcy status based on cutoff scores resulted in a consistent pattern of competentcy compromise among patients with AD which related to both dementia severity and stringency of LS. We feel that the instrument has construct validitv and represents an important first step toward standardized evaluation of competency in patients with dementia.

The five LSs for competency appeared to be hierarchical in difficulty. Control subjects. Patients with mild AD, and patients with moderate AD performed equivalently on two minimal standards: simple expression of a treatment choice (LS1) and making the reasonable treatment choice (when the alternative is manifestly unreasonable) (LS2). These findings suggested that even patients with moderate AD retain the capacity to elect a treatment choice and to select a reasonable treatment option from a manifestly unreasonable one. On more challenging standards, however, patients with AD demonstrated increasing difficulty. On LS3, which tests the capacity to appreciate the emotional and practical consequences of a treatment choice, over 60% of the patients with moderate AD and 33% of the patients with mild AD demonstrated compromise (marginal competency or incompetency). On a standard requiring rational reasons for treatment choice (LS4), over 50% of the patients with mild AD and virtually all patients with moderate AD demonstrated compromise. Generating even one coherent reason for a treatment choice appears to be a very challenging task for most patients with moderate AD and many patients with mild AD. Finally, on a standard requiring fairly comprehensive understanding of the treatment situation, alternatives and associated risks and benefits (LS5), both patients with mild and patients with moderate AD demonstrated 100% compromise. This standard, at least as used in this study, appears to require a level of information retention and processing beyond the capacity of most patients with AD. It is possible that, with rehearsal or different forms of information disclosure patients with AD could perform better on this standard.

The hierarchical difficulty of the five standards was also reflected in the pattern of competency outcomes for individual demented subjects. Competency compromise for individual patients with AD occurred in a consistent pattern, with compromise on less stringent standards (LS1 and LS3) almost always associated with compromise on more stringent standards (LS4 and LS5). The pattern was particularly pronounced for patients with moderate AD, reflecting the general relationship of dementia level to competency status. Future studies might examine whether this hierarchical pattern of competency loss is related to the pattern of progressive neuropsychologic decline characteristic of AD.

Our findings raise the concern that. depending on circumstances and the standard to be applied, many patients with mild AD my be incompetent to consent. Particularly in situations where a treatment decision involves a significant risk or action (for example a new experimental treatment or a do not resusitate order). and where a court would apply a more stringent standard, relatively few patients with AD may be competent to consent. Accordingly it is important that competency be tested in a formal and standardized manner for all patients with AD, and that the applicable LS or threshold be specified at the outset. General guidelines for determinations of capacity to consent in AD have recently been set forth in a position paper by representatives of AD centers nationally.

This study supports the potential value of vignettes as an objective clinical measure of competency to consent to treatment. We wish to emphasize, however, that our prototype instrument is intended to assist, but certainly not displace, the clinician. A single instrument and score can never take account of the variety of medical, legal, ethical, and other factors that inform a competent decision. In addition, a clinical vignette is a limited measure of consent capacity, since the process of understanding and responding to a hypothetical medical problem will always differ from that associated with a personal medical problem . Nonetheless, a vignette instrument can provide the standardized basis for systematically evaluating important abilities integral to the competentcy to consent at different legal thresholds. In this vein, we believe that normatively derived cutoff scores are an essential element of such instruments. Although the use of "expert based" competency cutoff scores has received criticism, cutoff scores based on normative samples appear valid and are arguably indispensable in clinical research and practice.

In summary, this empirical research represents an initial effort to study competency in a well characterized dementia population. We found our two vignettes to be a reliable and valid prototype instrument for discriminating the competency performance and classifying the competency status of normal older control subjects and patients with AD under different LSs. In the future, we hope to evaluate the criterion validity of the vignette scoring system using external physician competency judgments. Future research might also investigate the relationship between a physician's choice of LS and the level of treatment risk. Overall, we feel that clinical vignettes with well operationalized and normatively based scoring systems have great promise as a competency assessment instrument in dementia.