Voluntary informed consent is, with rare exceptions, a necessary, albeit not sufficient, defining precondition of ethical clinical treatment, and it is essential for enrollment in clinical research trials. The validity of an individual's consent is contingent not only on it being given in a voluntary context and with full disclosure of all relevant information (in language intelligible to the recipient) but also on the consenting person having sufficient mental capacity to make a meaningful decision.1,2
Explicit and systematic evaluation of decisional capacity is often thought of in the context of legal competency proceedings, such as when a patient with questionable decisional capacity refuses an intervention that others think necessary to his or her safety and well- being. Whether or not the treating clinician decides to conduct a systematic or explicit assessment of the patient's decisional capacity, he or she nevertheless has an ethical responsibility to ensure that the patient who agrees to a recommended intervention actually has sufficient capacity to understand what will be involved in the treatment.1,2
There are few guidelines that specify when explicit assessment and documentation of an individual's consent capacity is needed. However, such assessment and documentation may be particularly prudent when the individual consents to complex or high-risk interventions, when there are multiple viable treatment options with different or complex risk-benefit profiles, and/or when the patient has marked cognitive deficits or other risk factors for impaired decisional capacity.3,4Decisional capacity
As detailed in the model put forth in 2 studies by Appelbaum and colleagues,5,6 decision-making capacity is generally defined by 4 functional tasks:
- Understanding information as it relates to the choice, such as the nature, purpose, and potential risks and benefits of the proposed treatment as well as alternatives.
- Appreciation of the relevance of that information for one's condition or situation (such as recognition that one has a condition for which effective treatment could be potentially helpful).
- The ability to consider and compare (and reason about) potential consequences of the various viable options (including no treatment).
- The ability to communicate (express) a choice.
These abilities are all framed in reference to a specific decision. The ethically relevant question is not the patient's general (typical) capacity for making decisions, but rather his capacity to make the particular decision at hand.
This 4-component model of decisional capacity was developed, in part, from consideration of the abilities deemed through case law as relevant to establishing competency.5,7,8 However, decisional capacity and competence are not synonymous, the former is a clinical construct and the latter is a legal term. Although courts generally give weight to the evidence an examining clinician provides regarding a patient's decisional capacity, the ultimate determination of competence is made by the court, not by a physician or other clinician.9 In contrast, the treating clinician retains ethical responsibility for ensuring that the patient has the capacity to consent to the treatment(s).Instruments for assessing decisional capacity
Unaided by structured methods, there tends to be low interrater reliability among clinician judgments of decisional capacity.10,11 Fortunately, interrater reliability can be improved with specific guidance on the standards for decisional capacity,12 and a number of structured or semistructured instruments have been developed to assist clinicians to systematically evaluate their patients' decisional capacity relative to such standards.13-15
In a recent comprehensive review we identified 15 published questionnaires or instruments that assess capacity to consent to treatment (as well as 10 instruments for assessing capacity to consent to research).13 Nine of the 15 treatment consent capacity measures provide for assessment of all 4 dimensions of decisional capacity, and evidence of at least adequate interrater reliability (= 0.80) was available for 7 of these 9 measures (Table).13
One primary methodological distinction among capacity assessment instruments is whether patients' decisional capacity is evaluated in reference to an actual treatment option or in reference to hypothetical vignettes. Of the 7 measures, 4 use hypothetical vignettes, whereas the other 3 evaluate a patient's response to actual diagnosis and proposed treatment(s).
The instruments that use hypothetical vignettes have the advantage of standardized methods and facilitate comparison of responses and scores across patients and settings, thus they tend to be ideal for empirical studies of decisional capacity. However, in a clinical context, their predictive validity and generalizability is less clear. On the other hand, tailoring content to the individual's actual choice clearly increases the ecological validity of the assessment but complicates comparison of results from different treatment scenarios. Because the specific information to be understood, appreciated, or reasoned about is variable, the degree to which supporting data on the reliability and validity of the instrument tailored for one treatment scenario generalizes to other decisions is often unclear.
In addition, clinicians who use these instruments should keep in mind the specific item content. Although there appears to be good inter-instrument consistency in the understanding component of decisional capacity, subjective as well as empirical comparison of the subscales from various instruments suggests some discrepancies in the appreciation and reasoning components.23-25
A persistent methodological challenge in establishing the criterion or predictive validity of decisional capacity instruments is the lack of a gold standard against which the current measures can be evaluated. In lieu of an ideal criterion, validity data for these instruments have taken a range of forms, including agreement with expert or other stakeholder judgments, ability to discriminate between cognitively impaired persons and neurologically healthy individuals, and the degree of correlation or agreement with other capacity instruments and/or neuropsychological measures. (Dunn and colleagues13 provide detailed tables summarizing format content, administration time, and psychometric properties for each published instrument.)