MacArthur competence assessment tool for clinical research (MacCAT-CR); MacArthur competence assessment tool for treatment (MacCAT-T); MacArthur competence assessment tool-criminal adjudication (MacCAT-CA)
The MacArthur Competence Assessment Tools (MacCAT) consist of the MacArthur Competence Assessment Tool for Treatment (MacCAT-T), the MacArthur Competence Assessment Tool for Clinical Research (MacCAT-CR), and the MacArthur Competence Assessment Tool-Criminal Adjudication (MacCAT-CA). Each of these tools measures a different form of decisional capacity; separate instruments have been developed based on the premise that the capacity to make decisions is situation-specific. All three instruments are commercially available through Professional Resource Press (www.prpress.com).
The MacCAT-T (Grisso & Appelbaum, 1998b) measures capacity to consent to medical treatment with a semi-structured interview tailored to the patient’s specific disorder and treatment decision. Patients are provided information about their disorder and are asked to reiterate this in their own words; follow-up questions ensure patients’ understanding of their condition. Patients are then asked if they have doubts about the information and the reasons for their beliefs are explored. The process is repeated for treatment options. Finally, the patient is asked to make a choice; questions are asked to ascertain their reasoning process. The order of the interview is: understanding disorder, appreciation disorder, understanding treatment, understanding risks and benefits, appreciation treatment, alternative treatments, first choice and reasoning, generating consequences, and final choice. Responses are rated by the clinician as 2, 1, or 0 (adequate, questionable, and inadequate), using the manual as a guide. What is considered “adequate” for each response will depend on the specific situation or condition and the clinician’s judgment that the patient’s response was or was not accurate. Summary ratings are generated for four constructs: understanding (rating range 0–6), appreciation (0–4), reasoning (0–8), and expressing a choice (0 or 2). There is no total score; each summary rating can be discussed separately. There are also no cutoffs for the individual summary scores. An inadequate rating on one scale may result in the patient being considered incompetent to consent to treatment. The construct considered the most important in determining competence varies somewhat by state, with most states recognizing understanding, but with variability on use of the other constructs. The MacCAT-T was normed on 40 hospitalized patients with schizophrenia or schizoaffective disorder and 40 controls (matched for age, gender, race, and socioeconomic status). Administration requires 20–25 min. Both a DVD and a VHS videotape are available demonstrating the administration of the MacCAT-T. This tool has the advantages of ease of use in clinical settings and of being tailored to the individual patient’s medical situation.
The MacCAT-CR (Appelbaum & Grisso, 2001) measures capacity to consent to participation in research. Questions are tailored to the specific research study in which the subject is asked to participate. As in the MacCAT-T, four summary ratings are generated: understanding, appreciation, reasoning, and expressing a choice. Responses are rated by the clinician as 2, 1, or 0, using the manual as a guide. Understanding (rating range 0–26) is based on answers to 13 questions that are in reference to five disclosures about the research study. Appreciation (0–6) is based on answers to three questions assessing: (a) the subject’s recognition of the study as not meant for their personal medical benefit, (b) the possibility of their reduced benefit, and (c) their ability to withdraw from the study. Reasoning (0–8) ascertains the subject’s reasoning of their expressed choice (0–2). Administration requires 20–25 min. Norms for the MacCAT-CR are not provided. Like the MacCAT-T, the MacCAT-CR has the advantage of utilizing specific details of the individual’s situation in order to make decisions about competency.
The MacCAT-CA (Poythress et al., 1999) measures a defendant’s competency to stand trial. It is a structured interview of 22 questions, organized into understanding, reasoning, and appreciation constructs. Responses are again rated by the clinician as 2, 1, or 0, using the manual as a guide. Unlike the previous MacArthur tools, the MacCAT-CA features a hypothetical vignette, from which questions for the understanding and reasoning scores are derived. Understanding score questions determine defendants’ comprehension of the roles of attorneys, judge, and jury; the charged offense and any lesser offense; consequences of conviction and pleading guilty and rights waived when pleading guilty. Reasoning score questions assess judgments by the defendants of the relative importance of facts, as well as their decision process about two pleading choices. The appreciation score comes from questions about the defendants’ legal situation, and specifically by the explanations of their choices (to rule out implausible or delusional thinking). Summary scores are generated for understanding (rating range 0–16), reasoning (0–16), and appreciation (0–12); there is no total score. Cutoff scores, while provided, are intended to be used comparatively and not as absolutes. Norms were derived from 283 hospitalized defendants whom were deemed incompetent, 249 jailed defendants whom were treated for psychiatric reasons but deemed competent, and 197 jailed inmates presumed competent. Subjects were between 18 and 65 years of age, and 90% were male. This tool has the advantages of potential use in both clinical and research settings and of having a large normative sample. However, it has the disadvantage of employing a vignette and not being completely tailored to the individual defendant’s actual situation. The applicability to females has also been questioned (Grisso, 2003).
Neuropsychologists are occasionally asked to provide information regarding an individual’s decision-making capabilities, which can be used in judgments of competency. Competence is a legal construct involving one’s right to make decisions for oneself; adults are deemed to be competent unless proven otherwise. Simply having a mental or cognitive disorder or diagnosis is not enough to be considered “incompetent.” There must also be evidence that individuals are unable to understand relevant information, engage in reasoning based on provided information, and make and express decisions regarding their wishes.
Determinations of competence are not all-or-nothing: there are different types of competencies; each involving different skill sets (see Grisso, 2003). While executive functioning skills, and decision-making ability in particular, are important for demonstrating competence, evidence of competency is derived from the performance of various functional tasks (e.g., managing finances and taking medication). An important distinction is made between decisional capacity (the ability to decide) and executional capacity (the ability to carry out the decision) (Collopy, 1988). Individuals with only decisional capacity may still be deemed competent, as they may be able to carry out tasks with assistance and/or they can instruct others to perform tasks in accordance with their decisions. Moberg and Kniele (2006) discuss the ethical evaluation of competency by neuropsychologists, and advocate integrating specific competency measures into interview and neuropsychological testing data, when available. The MacCAT are one such example of competency measures, each evaluating a different type of decision-making competency. All three measures were created through funding from the MacArthur Research Network on Mental Health and Law.
The MacCAT-T was derived from three instruments that were part of the initial research initiative in the 1990s, the MacArthur Treatment Competence Study: the Understanding Treatment Disclosures (UTD), Perceptions of Disorder (POD), and Thinking Rationally about Treatment (TRAT) instruments (see Grisso, 2003). These three instruments assessed decision-making ability (understanding, appreciation/acknowledgment, and reasoning, respectively), but were lengthier (60–90 min each), used hypothetical vignettes rather than patients’ actual situations, and had detailed scoring criteria. A shorter, more flexible, and more relevant instrument was sought for use by clinicians. The MacCAT-CR was derived from the MacCAT-T.
The MacCAT-CA is derived from the longer (47-item) MacArthur Structured Assessment of Competencies of Criminal Defendants (MacSAC-CD), which was a research tool as part of the MacArthur Research Network on Mental Health and Law to develop measures of competence to stand trial.
Each of the three MacArthur tools is standardized with regard to procedure (i.e., structure of interview questions), but not with regard to content, as each utilizes at least some, if not all, of the subject’s individualized situation in the questions. The MacCAT-CA, utilizing a vignette and a highly structured interview, has the best standardization of the three.
For the MacCAT-T, interrater reliability was 0.99 for understanding, 0.87 for appreciation, and 0.91 for reasoning (Grisso and Appelbaum, 1998b). This was determined by correlations among three raters on 20 patient and 20 community control protocols. Retest reliability was not performed. Given the purpose of the instrument, a review of test items revealed that the measure has good face validity. Tests of convergent validity with physician ratings have been mixed (Sturman, 2005). For example, in a German study (Vollman et al., 2003), there were moderate associations between MacCAT-T scores and physician ratings for the schizophrenia and dementia groups, but not the depression group.
The MacCAT-CR’s interrater reliability in two studies of persons with schizophrenia and Alzheimer’s disease (Carpenter et al., 2000; Kim et al., 2001) was 0.98 and 0.94 for understanding, 0.84 and 0.90 for appreciation, and 0.84 and 0.80 for reasoning, respectively. Retest reliability in patients with depression (Appelbaum et al., 1999) was 0.26 for understanding, 0.36 for appreciation, and −0.15 for reasoning.
For the MacCAT-CA, interrater reliability was very good to excellent, with intraclass correlations of 0.90 for understanding, 0.85 for reasoning, and 0.75 for appreciation (Otto et al., 1998). Regarding internal consistency, Cronbach’s alpha was 0.85 for understanding, 0.81 for reasoning, and 0.88 for appreciation. Within each scale, inter-item correlations were 0.42 for understanding, 0.36 for reasoning, and 0.54 for appreciation. Significant correlations were also found between MacCAT-CA scales and WAIS-R Full Scale IQ (particularly between understanding and reasoning) and psychoticism (as measured by the Brief Psychiatric Rating Scale (BPRS) and the Minnesota Multiphasic Personality Inventory-2 (MMPI-2)) in the original study (Otto et al., 1998). A more recent study examining the factor structure of the MacCAT-CA using the original normative dataset (Zapf et al., 2005) found the instrument to be normatively sound and fitting best with a three-factor model (understanding, reasoning, and appreciation). Correlations between MacCAT-CA scores and tests of intelligence (estimated WAIS-R Full Scale IQ) and psychopathology (BPRS) were 0.42 and -0.36, respectively, indicating moderate to strong covariation.
In a review of the measure, Grisso (2003) stated, “A strength of the MacCAT-T is its use of constructs that are based on legal analysis of competence and that have proved useful in studies with the parallel research instruments that influenced the development of the MacCAT-T (p. 425).” All three MacArthur tools utilize the constructs of understanding, appreciation, and reasoning. The MacCAT-T and MacCAT-CR offer clinical portability and situation-specificity at the expense of exactness and standardization. The MacCAT-CA offers better standardization, but loses some situation-specificity. The lack of a total score on these instruments underlies the concept that competence is multifaceted; the absence of cutoff scores (on the MacCAT-T and MacCAT-CR) indicates that competence is not “all-or-nothing.” Even though the MacCAT-CA provides cutoff scores, the manual cautions against using these as absolutes. Rather, they should be used as part of the evaluation for competence.
The MacCAT-T has been used to assess competence to consent to treatment in persons with dementia, schizophrenia, and depression (Vollman et al., 2003). Patients with dementia were found more impaired than those with schizophrenia, who in turn were more impaired than those with depression. In medical inpatients, Raymont and others (2004) utilized the MacCAT-T to examine decision-making capacity in acutely ill medical patients of various diagnoses.
Regarding the MacCAT-CR, studies have shown that cognitively impaired individuals with Alzheimer’s disease (Kim et al., 2001) and mild cognitive impairment (Jefferson et al., 2008) have more decision-making impairment compared to controls, and therefore would have impaired capacity to provide informed consent for research. The MacCAT-CR has also been used to assess capacity of patients with schizophrenic and HIV to consent to clinical drug trials (Moser et al., 2002).
According to the manual, the MacCAT-CA can be used for felony and misdemeanor defendants in inpatient, outpatient, or forensic settings and can be used to assess treatment. The MacCAT-CA is not recommended for individuals who are cognitively impaired (e.g., those with limited intellectual abilities). It also does not include an effort measure (explicit or embedded) to rule out feigned incompetence or retardation (Rogers et al., 2002). For a screening tool with an embedded measure for feigned incompetence, the Evaluation of Competency to Stand Trial-Revised (ECST-R; Rogers et al., 2004) is favored (see Marcopulos et al., 2008). Finally, the MacCAT-CA has limitations in non-English speaking individuals as well as those suffering from delusions (Pinals et al., 2006).
The MacCAT-CA has recently been tested on an adolescent population (Grisso et al., 2003). It has also been adapted to legal requirements for competence to stand trial in England and Wales (Akinkunmi, 2002). This measure, the MacArthur Competence Assessment Tool-Fitness to Plead (MacCAT-FP) evidenced good internal consistency and interrater reliability, and significantly differentiated a prison group from a hospital group.