Executive control functions have been defined as "those processes which orchestrate relatively simple ideas, movements, or actions into complex goal-directed behavior"(1). Without them, behaviors important to independent living can be expected to break down into their component parts; patients become overdependent on environmental cues and are easily distracted and perseverative (2,3).
Executive functions have been associated with three prefrontal-subcortical subtypes, the dorsolateral prefrontal, the orbitofrontal, and the anterior cingulate (mesiofrontal) (4,5,6). Damage to the dorsolateral subtype impairs abstraction and hypothesis generation. Orbitofrontal lesions lead to impaired emotional control. Irritability and mood swings in the absence of pervasive mood disorder are common sequelae. Mesiofrontal lesions lead to apathy, indifference, and impairment of goal-directed attention. Patients become passive and disorganized (7,8). In 1994 the American Psychiatric Association added executive control functions to the list of cognitive domains that should be considered when making a diagnosis of dementia (9).
Impairments in executive control functions are not limited to dementia, where they contribute greatly to patient morbidity, caregiver burden, and institutionalization (3,10,11). Nor are they limited to traumatic brain injury of the frontal lobes. Specific systemic illnesses that appear to be related to declines in frontal lobe abilities include hypertension, diabetes, and chronic obstructive pulmonary disease (12). Impaired executive control functions can also be recognized in major depression, Parkinson's disease, subcortical strokes, and psychotic disorders (3,13,14).
Royall and associates (15) have suggested that the prevalence of impairments in executive control functions increases significantly with age. If this is the case, then psychiatrists caring for an ever-growing aging population will inevitably confront clinical and ethical challenges, some of which are discussed in this paper. Patients with impaired executive control functions do not necessarily exhibit objective impairment of memory or other nonexecutive cognitive skills. Although impairments in executive control functions are better recognized today, few tests have been developed to help identify and discriminate between these clinical disorders (3). Individuals with impaired executive control functions often do well on many standard cognitive assessments and function adequately in a structured setting.
The purpose of this paper is to identify the ethical implications of impaired executive control functions for clinical practice. We examine a philosophical concept of autonomous decision making relevant to individuals with impaired executive control functions in order to provide a practical, ethically justified framework for making treatment decisions for patients with impairments in these functions.
Autonomous decision making
Faden and Beauchamp (16) explained that the concept of autonomy means that a person is substantially self-governed in thought, speech, and behavior. Collopy (17) identified two functional realms of autonomy: making decisions, or decisional autonomy, and carrying out those decisions, or physical autonomy. Our focus will be mainly on decisional autonomy because of its clinical and ethical significance in providing care for patients with impaired executive control functions.
Three components of decisional autonomy
Faden and Beauchamp (16) identified three components of decisional autonomy: understanding, intentionality, and noncontrol or voluntariness. Although many brain disorders have clinical and ethical implications for patient autonomy, different neurocognitive syndromes can adversely affect the three components of autonomy to different degrees. Impaired executive control functions are distinctive in their clinical and ethical implications for intentionality and voluntariness.
Understanding means that a patient knows both that he or she is authorizing a clinical intervention and what that particular intervention involves (16). This task requires the patient's cognitive understanding of the nature of the disease, the elements of the intervention, and the benefits and risks of the intervention. White (18) has shown that this task also involves affective or evaluative understanding—that is, weighing the benefits and risks in terms of one's values and goals.
Appelbaum, Grisso, and associates (19,20,21,22) have successfully operationalized this component of autonomy into measurable subcomponents: understanding of treatment disclosures, perceptions of disorder, thinking rationally about treatment, and expressing a choice. Although patients with impaired executive control functions can experience diminished abstraction and insight and other problems that could impair evaluative understanding, these deficits are typically less salient.
Intentionality means an action "willed in accordance with a plan" (16). In the clinical decision-making process, patients are asked to commit themselves to clinical care as an organized process, or plan, that involves a health-related goal and to the orderly clinical steps required to achieve this goal. Patients also bring to the clinical setting their own personal goals and values, which can range widely across health-related domains. Intentionality in the clinical setting means making and implementing plans of care on the basis of such valued goals.
Voluntariness means that the patient's actions do not result from either coercion or substantially controlling manipulation (16). Voluntariness can be described from the decision maker's internal perspective or from an observer's external perspective. Patients with severe psychotic symptoms, such as delusions of thought control, command hallucinations, or the automatic obedience of catatonia, may have severe impairment in the voluntariness component of autonomous decision making from an internal perspective. Impaired executive control functions may also diminish voluntariness from an external perspective in ways not immediately obvious, as we explore below.
Patients with executive dysfunction exhibit distinctive deficits in two components of decisional autonomy: difficulty with formulating and carrying out plans, including alternatives to plans proposed by clinicians—the intentionality component—and difficulty with making voluntary decisions, or the voluntariness component. Consequently, they participate in the clinical decision making with, at best, impaired decisional autonomy. We now analyze the clinical and ethical implications of these impairments for the general management of patients with executive dysfunction.
Impairments of intentionality
Patients, especially in the outpatient setting, should be able to cooperate in carrying out their plan of care, monitor themselves for significant clinical changes, and to alter the plan as necessary. Executive dysfunction can impair patients' ability to carry out these tasks, thus placing them in jeopardy if they refuse the proffered plan of care.
These impairments of intentionality are ethically significant for two reasons. First, the informed consent process is predicated on the ethical and legal assumption that patients possess intact decision-making capacity for the particular task or life function called into question (23). Possession of this capacity involves all three components of autonomy. In the management of chronic illness, the patient's cooperation is essential, with success or failure depending on that cooperation. Second, the concept of respect for refusal by competent patients makes an implicit appeal to the additional assumption that competent patients are able to fend for themselves: they are able to carry out alternative plans of care.
By impairing intentionality to different degrees, executive dysfunction calls into question the presumption of patients' ability to make such plans and thus to control the outcomes of the consent process. This impaired intentionality has the significant ethical implication that when the treatment team concludes that a patient's wishes are not in that patient's best short-term or long-term health-related interests, the patient's formulation or refusal of a plan of care need not always ethically override the team's judgment, as the presumption of autonomy usually requires.
Impairments of voluntariness
Some patients with executive dysfunction can respond nonvoluntarily to verbal and behavioral cues or prompts in their environment, even when they understand their circumstances. When patients respond to such cues with diminished control over their response, they exhibit "stimulus-bound" behavior. Such patients may also be unable to cooperate with outpatient treatment without appropriate cues. This diminished ability may be further affected by defects in understanding, such as diminished abstraction or poor insight. When those cues or prompts (or their absence) are contrary to or inconsistent with medically appropriate plans of care for the patient, these plans are often disrupted.
The capacity for autonomous decision making of such patients is impaired in its voluntariness. Moreover, patients' expressed preferences can be overridden by others, particularly clinicians in an inpatient setting who have power to control the discharge planning process. Clinicians can induce "acceptance" of a care plan, which is ethically troublesome because patients appear to be competently agreeing when, in fact, they are substantially controlled by the verbal and behavioral cues of the clinician.
Anterior cingulate (apathetic) frontal lobe syndrome
Mr. A was a 74-year-old man with a history of coronary artery disease, hypertension, and non-insulin-dependent diabetes mellitus. A provisional diagnosis of dementia was made because his persistent noncompliance with treatment and self-neglect continued to result in frequent visits to the medicine clinic for exacerbated medical problems.
Mr. A was involuntarily hospitalized in the past because of medication noncompliance and gross self-neglect. He had been living alone in squalid conditions; for example, he hoarded cartons of rusting cans of spoiled food and stacked them several feet high in his driveway. He had persecutory delusions that his neighbors and son were trying to steal his belongings. During a previous admission for severe hypertension caused by noncompliance with medication, the medical team was in favor of initiating guardianship proceedings, but the psychiatric consultant did not find Mr. A to be incapacitated. He was committed to the geropsychiatry unit, and guardianship was eventually pursued successfully.
On admission to this unit, Mr. A presented as an elderly, unkempt white male who was cooperative but stated that he should not be in the hospital. His affect was mildly blunted, but he reported his mood as normal and not depressed. His speech was coherent, although circumstantial and often irrelevant. He denied suicidal or homicidal ideation. He had the previously described persecutory delusions but denied having hallucinations. He was well oriented and scored 28 of 30 points on the Mini Mental State Examination. Neuropsychological evaluations, including the Dementia Rating Scale, Trail Making A and B, and the comprehension subtest of the Wechsler Adult Intelligence Scale—Revised, as well as an occupational therapy assessment using the Kohlman Evaluation of Living Skills indicated that he could not return to independent living because he exhibited variable executive abilities. He was able to adopt an abstract attitude but showed poor judgment and was a safety risk. An electroencephalogram was found to be within the normal range. A computerized tomography (CT) scan of the head revealed an area of encephalomalacia in the right mesial inferior frontal lobe consistent with an anterior cerebral artery infarct.
Mr. A's medical condition was stabilized, and he was started on low-dose risperidone, which resolved his delusional thinking. He spent most of his free time in the day area. He would never begin any activity on his own but would participate appropriately once he became involved at someone else's initiation. The treatment team recommended nursing home placement, but Mr. A strongly disagreed with that discharge plan. On the other hand, he did not formulate any alternative plan, and he eventually was transferred to a nursing home without incident.
Dorsolateral frontal lobe syndrome
Mr. B was a 78-year-old married retired longshoreman with a history of coronary artery disease and hearing impairment; he was voluntarily hospitalized because of psychosis and his inability to care for himself and his wife. Eight months before admission, Mr. B's wife had a stroke and required almost total care, which Mr. B was providing at home. He began to believe that his wife's home health nurses were not administering the correct care, whereupon he fired them, threw his wife's enteral pump into a ditch, and did not give his wife her medications. In addition, he became less able to care for himself and could not maintain the house or their finances.
Two weeks before Mr. B's admission, he was persuaded by a friend to place his wife in a nursing home. Just before his admission, Mr. B was getting into frequent arguments with the nursing home staff and was planning to take his wife out of the nursing home.
On admission Mr. B was pleasant and cooperative, with a euthymic mood. He had some tangentiality and persecutory delusions, but no hallucinations. His attention and memory were good, and insight and abstraction were fair. His score on the Mini Mental State Examination was 28 of 30 points. A CT scan of the brain showed changes within normal limits for his age. The Kohlman Evaluation of Living Skills indicated that Mr. B had difficulty with money management and with recognition of dangerous home situations.
Mr. B was started on 1 mg of haloperidol. After several days on the unit, his thoughts became more coherent, and his delusions abated. Neuropsychological testing detected intact memory, but identified "moderate impairment in executive self-regulatory functioning as reflected in difficulty adopting an abstract attitude, changing cognitive set, and using feedback to change a problem-solving strategy."
Mr. B was not able to adequately formulate a coherent treatment plan without the treatment team's help. The team recommended that he return home and visit his wife in the nursing home, but that he should not take on any major caregiving responsibilities for her. The staff also recommended that someone help him with his finances and that someone check on him regularly. He was discharged home to live independently, with a friend agreeing to visit him to ensure his safety. His discharge diagnoses were psychotic disorder not otherwise specified and cognitive disorder not otherwise specified.
Mr. A initially refused nursing home placement, but his refusal probably emanated from seriously impaired decisional autonomy. The ethics of informed consent require that patients who refuse recommended management be given information about the risks and benefits of the alternative that they prefer. This obligation is based on an assumption underscored by our ethical analysis of impairment in executive control functions: that the patient's understanding and, crucially for the purposes at hand, intentionality and voluntariness are substantially intact, so that the patient can make and carry out an alternative plan. Mr. A could not, and informed refusal would therefore not be ethically justified as a basis of clinical decision making. Surrogate decision making was therefore required in this case.
Mr. B accepted a discharge plan of going home with supervision, thus reducing the health-related risks of this setting. In addition, the environmental stressor that appeared to precipitate steep reduction of decisional autonomy—the demand of caring for his wife at home—had been removed. Nonetheless, Mr. B would not have 24-hour supervision, and his ability to carry out and, when necessary, adapt his self-care-with-supervision plan was significantly impaired. The team, therefore, was at risk for overestimating the weight that should ethically be given to the presumption of autonomy in this case.
We conclude that the plan in Mr. A's case was the correct one, and we can buttress that conclusion with ethical justification. In Mr. B's case, an ethically sound alternative would have been a time-limited trial at home with close monitoring, such as regular home health visits and periodic interviews with neighbors and other social supports. The trial would be stopped if evidence emerged that the presumption of autonomy had become incorrect. Admission to a nursing home could then be arranged as the only remaining viable option.
The power to control discharge plans for older patients with executive dysfunction may shift with ethical justification from patients to family members or others acting as surrogates, to clinicians, and to institutions. Unlike other forms of cognitive impairment, such as Alzheimer's disease, that primarily affect understanding, impaired executive control functions mainly affect intentionality and voluntariness. Impairments of these aspects of autonomy may often be overlooked when only traditional mental status exams are employed, with adverse consequences for the health of patients wrongly thought to possess intact ability to make and carry out plans of care.
Detailed evaluation of executive functions may help guide clinical recommendations by addressing these prime determinants of competent decision making. One promising bedside instrument, the Executive Interview (EXIT) (3), has strong correlation with neuropsychological tests tapping executive control functions and nursing assessments of problem behaviors. A 25-item interview, EXIT has been validated and found to discriminate among older adults in various settings and levels of care. It has high internal consistency and interrater reliability.
Unfortunately, perfect correspondence between neuropsychological testing of executive control functions and real-world decision-making capacity does not exist. Qualitative observations from daily interactions must supplement quantitative data. Evaluation of autonomous decision-making capacity is an ongoing process that requires integration of data from multiple sources, including neuropsychiatric examination, neuropsychological testing, family interviews, and detailed questioning over time by an interdisciplinary team.
The authors acknowledge the thoughtful reviews of J. D. Hamilton, M.D., Jennifer Moye, Ph.D., J.D., and Donald Royall, M.D.
Dr. Workman, Dr. Kunik, and Dr. Orengo are assistant professors, Dr. Molinari is associate professor, and Ms. Rezabek is clinical instructor in the department of psychiatry and behavioral sciences at the Houston Department of Veterans Affairs Medical Center and Baylor College of Medicine, 2002 Holcombe Boulevard (580/116A), Houston, Texas 77030 (e-mail, firstname.lastname@example.org). Dr. McCullough is professor at the Center for Medical Ethics and Health Policy and faculty associate at the Huffington Center on Aging at Baylor College of Medicine. Ms. Khalsa is a physician's assistant at the Department of Veterans Affairs Medical Center in Houston.