Abstract
Since the 1980s, the four skills criteria have become the most widely accepted mechanism for the assessment of decisional capacity in the United States. These criteria emerged in response to the paternalistic approach to clinical decision-making that had been widely accepted in an earlier era and offered a means of ensuring that physicians honored the rights of capacitated patients to make their own medical decisions. Unfortunately, the criteria are now applied to situations for which they are not suited and in a manner that is often highly inflexible. In an article in this issue of The Journal, Matthew Dernbach and colleagues describe one potential scenario that requires a flexible approach to using the four skills model: situations in which a patient stands at high risk of losing decisional capacity in the near future. Using Dernbach et al. as a starting point, this article offers specific ways in which the four skills model can be improved upon or augmented without abandoning its key principles. These advances include adjusting to empirical evidence, re-emphasizing the importance of autonomy maximization and restorability, and embracing novel conceptual and technological innovations.
The systematization of decisional capacity assessment in clinical settings that developed during the 1980s proved a major advance for psychiatric practice and patient-centered care. Drawing upon the seminal work of Loren Roth and James Drane, Paul Appelbaum and Thomas Grisso advanced a four skills approach to evaluation that was grounded in ethics, intellectually sound, and relatively easy to operationalize.1 Among the most important, and underappreciated, consequences of American medicine’s nearly uniform acceptance of the four skills criteria was an increased respect for patient autonomy.2 The use of the criteria helped ensure that physicians honored the rights of capacitated patients to make their own medical decisions. Forty years later, one all too readily forgets how significant, and arguably radical, a transformation this development was for a field that only a generation earlier had operated according to paternalistic norms.
A major benefit of the four skills criteria was functionality. Its purpose was to protect patients’ rights and to further their well-being. Unfortunately, the ensuing decades have witnessed the use of the approach in ways for which it is poorly suited, often at the expense of patient autonomy and welfare.3 The criteria prove beneficial for situations in which a patient who embraces the values of allopathic medicine possesses decisional capacity to make a particular decision at baseline but then loses that ability as a result of insult or illness.4 Common cases are those of dementia, delirium, psychosis, and intoxication. The goal in such situations is, if possible, to restore patients to capacity, and, if not, to render decisions as they would have wished. As I and others have noted elsewhere, however, the same model may prove far less useful, or even undermine the goals of autonomy and patient well-being, when applied to a broader set of circumstances, such as patients in denial or those who are volitionally unwilling to engage.5,6 In other situations, such as those involving individuals who belong to established communities that reject allopathic medicine at baseline (e.g., Christian Scientists), the model is generally set aside as unsuitable.7 In short, critics of the four skills approach are usually objecting to the breadth of its application. These critics do not call for its complete rejection or even for sweeping revisions but rather a rejection of its universality in favor of limiting its use to the admittedly large subset of cases for which it is reasonably suited and likely intended.
In this issue, an insightful article by Dernbach and colleagues demonstrates how too narrow or inelastic an application of the model can also lead to negative results.8 They describe a scenario that arises frequently in clinical care in which a patient who “may appear to be otherwise capacitated” expresses a decision that is “likely to place the patient at elevated risk of future decisional and functional incapacity” (Ref. 8, p 000). As they saliently note, too rigid an application of the four skills approach risks finding that such patients technically meet these criteria as traditionally understood, all the while placing them in jeopardy of acute or even imminent incapacitation and concomitant decompensation in a manner that neither meaningfully protects their autonomy nor serves their interests. In response, they wisely propose a more robust interpretation of the four skills model that allows for determinations of incapacity based on a lack of understanding of future incapacity and its consequences. As a theoretical matter, whether this approach is best characterized under the rubric of the four skills model or defined as an exception to those principles raises intriguing, albeit academic, questions. In practice, as the vast majority of United States jurisdictions have codified at least portions of the Appelbaum and Grisso rubric into law, Dernbach and colleges have chosen an eminently prudent course.2
Their article also demonstrates three specific ways in which the four skills model can be improved upon or augmented without abandoning its key principles. These advances include adjusting to empirical evidence acquired over more than three decades of assessments; re-emphasizing the importance of autonomy maximization, harm mitigation, and restorability as key components of the model; and embracing conceptual and technological innovations that promise to reshape capacity assessment in the years to come. Each of these topics is discussed further below.
Empirical Evidence
The inherent challenge in developing a clinical assessment mechanism is that one does not have the luxury of knowing a priori how that mechanism will be applied in clinical practice. Over time, experience with a mechanism allows practitioners to fine-tune its use, modify its application, or change course entirely. That recalibration has not yet happened to any significant extent with the four skills criteria. Although the criteria have been in use for 35 years and, as Dernbach et al.8 note, criticism has been raised in the literature, the model, largely unchanged from its original form, has increasingly become ossified into dogma.9 It is not an exaggeration to say that, in the United States, the “Appelbaum criteria” have become synonymous with decisional capacity assessment. Yet the failure to adapt the model in response to years of empirical feedback, both anecdotal and statistical, has done a disservice to the model and to its authors. In identifying one specific, frequently occurring clinical scenario in which the model should be “adapted” (their language) or expanded (my language), Dernbach et al.8 demonstrate the value of such flexibility. Moreover, as they note, their approach will likely undergo “further…evidence-based adjustment,” a crucial process for it to prove most effective.
The four skills model is ripe for additional evidence-based adjustments. Much of the evidence for the benefits of such adjustments is anecdotal. For instance, concerns have been raised that the invasiveness of the evaluation process itself may disrupt the therapeutic relationship between the patient and the care team.10 Several difficulties with the model have been studied systematically. First, the four skills approach is subject to considerable variability among evaluators.11 This variation appears particularly pronounced for certain diagnoses, such as mild cognitive impairment in Alzheimer’s disease.12 The model’s assessment process has also shown itself prone to high levels of racial bias.13 In light of these concerns, prudence suggests that the model ought to be adjusted accordingly with the addition of safeguards to minimize bias and enhance consistency. The possibility also exists that the model is simply too imprecise for certain diagnoses, even when applied properly by appropriately trained clinicians, an issue that at a minimum calls for further examination. Dernbach and colleagues offer a strong template for how a particular limitation can be identified and the model then adapted accordingly.
Restoration and Harm Mitigation
In many circumstances, a patient’s loss of decisional capacity will be temporary or reversible. As initially conceived, one of the purposes of clinical capacity assessment was to identify these cases with the goal of restoration. Such an approach serves the underlying ethics objective of upholding patient autonomy. As Appelbaum and Grisso themselves explained, a finding of incapacity “does not end the evaluation process” because “[e]ffective treatment, even for a short time, may restore a patient’s capacities sufficiently” to obviate the need for depriving patients of their decision-making authority (Ref. 1, p 1636). A range of additional tools beyond direct treatment may also further or speed the restoration process, including “mobilizing supports,” “address[ing] communication barriers,” and “optimiz[ing] rapport and trust”(Ref. 14, p 108). In some nonacute circumstances, clinical decisions are best delayed until restoration has been attempted. Overriding a patient’s right to make medical choices is a serious breach of autonomy, and although sometimes doing so is ethically indicated, such a drastic step should be a measure of last resort. In forensic practice, restoration continues to be a “primary goal” of competence assessment.15 Regrettably, possibly as a result of economic pressure in the American health care system for rapid throughput, restoration all too often receives short shrift following clinical capacity assessment. In short, clinicians are quick to turn to third-party agents, such as proxies and surrogates, for guidance in settings where an attempt at restoration may be possible in the time clinically required for a safe medical decision.
Dernbach et al.8 advance an approach to an underappreciated challenge that is highly consistent with the goals of autonomy preservation. They enumerate ways in which providers can offer patients “interventions aimed at reducing their risk of losing capacity” in the first place (Ref. 8, p 000). They also offer alternative mechanisms for identifying future incapacity, including “arranging for family or friends to provide supervision for patients or to check in on them regularly” and “home visits by nurses or community workers,” interventions that may render denying patients decisional authority beforehand unnecessary (Ref. 8, p 000). These efforts at harm mitigation reflect recent trends in the literature that argue for shifting “away from a categorical and exclusionary practice toward a more dimensional and inclusionary concept of capacity” (Ref. 16, p 1571). They serve the same purpose in their enhanced model as restoration does in the traditional model, namely, minimizing the circumstances in which patients must have their autonomy overridden.
Innovation
Any evaluative mechanism is inherently a product of the era in which it was designed. The four skills model is no exception. Arising at the apex of the shift from paternalistic approaches to patient decision-making that prioritized beneficence and to newer, person-centered approaches that placed greater value upon patient autonomy, Appelbaum and Grisso sought to shield patients’ rights against overly protective medical authorities. Opening the prospect of limiting rights based on foreseeable loss of capacity in the future did not further that objective, so the clinical scenarios that might elicit such a need were never raised in the initial papers discussing the four skills model. In doing so at present, Dernbach and colleagues are not only responding to empirical evidence from the field, as discussed above, but also reflecting a significant shift toward foreseeability as a critical aspect of medico-legal decision-making.17 In some areas, such as the duty to warn18 and liability for third-party consults,19 such a shift may be unwelcome by psychiatry, whereas in others, such as physician responsibility to third-party victims of automobile accidents for prescribing decisions, the policy and ethics concerns may prove more complex.20 What is clear is that foreseeability of intervening events has increasingly become relevant to many areas of medical law in a way that it was not 35 years ago, and Dernbach et al.8 are right to apply this important development to capacity assessment.
This subject is just one part of a burgeoning range of novel legal doctrines and technologies that have developed that have significant implications for the field of capacity assessment and the four skills model in particular. For instance, Adrian Owen’s controversial use of functional magnetic resonance imaging (fMRI) to elicit medical preferences from patients who appear to be in persistent vegetative states raises complex uncertainties regarding what one means by voluntarily communicating a clear and consistent choice (i.e., the first of the four skills).21 State statutes permitting medical aid in dying (MAID) have implications for understanding risks, as in MAID, arguably, “the risk is the benefit” (Ref. 22, p 322). For the four skills approach to remain effective, the criteria must prove elastic enough to accommodate these recent technological and legal developments, as well as others not yet anticipated. Dernbach and colleagues offer an excellent example of how the model can incorporate these developments.
Conclusion
Three decades ago, in discussing the merits of the MacArthur Treatment Competence Research Instruments, Appelbaum and Grisso warned users against too rigid a use of these instruments to “provide determinations of legal incompetence to consent to treatment” and emphasized the importance of “contextual factors” in such assessments (Ref. 23, p 170). Marshall Kapp and Douglas Mossman derided the strict use of the rubric as the “capacimeter” approach, one that has now uniformly been rejected in forensics and research settings.24 Ironically, the same inelastic applications that Appelbaum and Grisso feared might befall their research instruments have become entrenched in the clinical uses of their four skills criteria. Dernbach et al.8 offer an impressive example of how clinicians can break free from the calcified status quo. The field of capacity assessment calls for similar reconsiderations in other aspects of evaluation.
Footnotes
Disclosures of financial or other potential conflicts of interest: None.
- © 2024 American Academy of Psychiatry and the Law
