Ethical Issues in Stroke Management

A variety of ethical issues may arise in the management of patients with stroke.¹ The adequacy of informed consent may be questioned in the incapacitated patient with acute stroke for whom IV alteplase administration is planned despite the absence of a lawful surrogate decision maker. The decision-making capacity of a patient with a recent dominant hemispheric infarction causing nonfluent aphasia may become an issue when he appears to understand but cannot communicate his treatment preference. Neurologists may be unsure whether to provide medical hydration and nutrition to an elderly patient admitted with a massive intracranial hemorrhage and incipient herniation whose advance directive states that life-sustaining treatment is to be withheld in the event of a hopeless prognosis. A patient with chronic locked-in syndrome from pontine infarction may order that further life-sustaining treatment be withdrawn, but her neurologist may be unsure of the ethical adequacy of her treatment refusal because of communication limitations.

Three articles in this issue raise additional ethical questions: (1) Are existing prognostic data sets of intracerebral hemorrhage (ICH) outcomes adequate to make life-death treatment decisions? (2) What is the ethical relevance of survey data showing how people regard chronically debilitating brain damage to be states worse than death? (3) What ethical issues are raised in the patient with factitious illness who appears to have a threatened stroke and is treated with alteplase? Here, we briefly comment on the ethical issues raised by these articles.

Shah et al² described a remarkable case of a young woman with chronic hydrocephalus who, after revision of a ventriculoperitoneal shunt, developed coma, absent pupillary light reflexes, extensor posturing, and raised intracranial pressure caused by a thalamic and midbrain hemorrhage with intraventricular rupture. With aggressive medical therapy, but no surgery other than external ventricular drainage, and despite a Glasgow Coma Scale = 4, FOUR Score = 4, and ICH Score = 4, she made an excellent functional recovery after a year to the modified Rankin Scale = 2. The authors point out that her poor ICH Score correlated with a 30-day mortality of 97%, so her striking functional recovery seems highly improbable. One possible explanation is that most patients with poor ICH Scores have supratentorial hemorrhages with brain stem dysfunction caused by irreversible transtentorial herniation. Whereas her poor ICH Score resulted from the direct effect of the midbrain hemorrhage and hydrocephalus, which may be more potentially reversible than brain stem dysfunction caused by herniation.

The authors also note the fallacy of the self-fulfilling prophecy, which may create a serious problem when using many published ICH outcome data to establish prognosis. This fallacy, which also occurs in outcome studies of coma after cardiac arrest and other causes of diffuse brain damage, results when prognostic scores are developed using outcome data sets in which many patients died as a direct result of withdrawal of life-sustaining treatment.³ Although a treatment withdrawal decision may have been justified in individual cases within the series, it is fallacious to rely on such a series to determine the natural history of the treated condition because it generates a bias toward poor prognosis. As a result of the recognition of this fallacy, many guidelines prohibit or caution against writing do-not-resuscitate orders or orders to withdraw life-sustaining therapy early in the course. Neurologists should use the latest and best validated, least biased summary of outcome studies to form their prognoses for patients with ICH.⁴

An irreducible clinical-ethical problem in massive ICH is that although most patients do not do well, a small number unexpectedly do make a good recovery, such as in the patient described by Shah et al.² There are poignant examples of patients undergoing aggressive life-saving procedures, such as hemicraniectomy, who survived to a functional state with which they were satisfied and happy to have been treated aggressively despite the low prior probability of a good outcome.⁵ The obvious issue is how neurologists should counsel patients and their surrogate decision makers to make such treatment decisions. Particularly vexing is the decision whether to pursue life-saving hemicraniectomy after malignant middle cerebral artery infarction with raised intracranial pressure and incipient transtentorial herniation because survival usually yields a severe, permanent neurologic deficit. This decision, which requires balancing the benefits of survival vs a subsequent life with a permanent major neurologic deficit, often must be made during an after-hours emergency telephone conversation with the surrogate and requires a snap decision. The decision is filled with uncertainty: the neurologist cannot be certain about the ultimate neurologic outcome with treatment, and the surrogate may not know whether the patient would want to live with a profound disability and resultant dependency.⁶

Everett et al⁷ studied a related problem: do stroke patients believe that certain neurologic disabilities represent a state worse than death? Implicit in such a belief is that it would be preferable to have died of the stroke than to linger in such a state. In their survey of 80 patients admitted to a stroke center in an urban hospital, using a Likert scale, they found that patients ranked the following states in the following percentages as equal to or worse than death: reliance on a breathing machine (66%), reliance on a feeding tube (66%), persistent confusion (62%), inability to communicate (58%), and bowel/bladder incontinence (50%). Principal component analysis clustered these responses into 2 separable factors: (1) reliance on the feeding tube, breathing machine, incontinence, chronic pain, and persistent confusion; and (2) wheelchair-bound or bed-bound living in a nursing home and requiring help for activities of daily living.⁷

These data are noteworthy because many of the states that their surveyed stroke patients feared most are common consequences of stroke. Yet, several factors limit drawing sweeping conclusions from this survey. First, using only a Likert scale to grade responses to life-death questions oversimplifies and distorts the process of introspective reflection and complex decision-making in a frightened, acutely ill patient. Although the Likert scale survey methodology yields crisp numerical data, it cannot adequately assess the psychological interiority of the patients in whom nuances might be captured better in a discussion exploring values.

Second is the well-known gulf between what surveyed people say how they will feel or what they will do in a hypothetical situation and what they will really feel or do in the actual circumstance. The reality of a situation alters hypothetical assumptions, and prior feelings may not accurately predict future attitudes.

Third, after adequate rehabilitation, many stroke patients become adjusted to and satisfied with their new functional status and rate their quality of life higher than healthy people might assume. The same phenomenon is seen in patients after spinal cord injury, amyotrophic lateral sclerosis, and even locked-in syndrome. Knowledge of this common phenomenon, called the “disability paradox,” should raise caution among clinicians when patients make rigid judgments in their advance directives and when lawful surrogates use a best interest standard to withdraw life-sustaining medical treatment from the incapacitated patients for whom they speak.⁸

Willenberg et al⁹ report a lesser known but real ethical concern in stroke medicine—the patient with Munchausen syndrome or other factitious illness who receives IV thrombolytic treatment for purportedly threatened stroke. They report a 30-year-old woman to whom they administered alteplase for her history of recent symptoms of threatened stroke. When factitious illness later was suspected, despite her denial, they were able to document that she had received the same treatment for the same symptoms in at least 3 other hospitals over the preceding 18 months, a typical story for a patient with Munchausen syndrome. On subsequent review of the 335 infusions of alteplase administered at their medical center over 29 months, 10 patients (3%) were diagnosed retrospectively as having Munchausen syndrome.⁹

The authors point out that the risk of postthrombolytic ICH in the stroke mimics subset, including Munchausen syndrome, is approximately 1%, which is much lower than the risk in the much larger subset of patients with cerebrovascular disease. Nevertheless, administering an expensive and potentially dangerous medication for what is a factitious illness raises obvious social and ethical issues. We agree that this is a more common situation faced by stroke neurologists than many appreciate. One of us (T.G.L.) tabulated similar data of 24 patients examined on the Stroke Service at Dartmouth-Hitchcock Medical Center over a 3-year period with what was ultimately diagnosed as “psychogenic pseudostroke.”¹⁰ Etiologies in this group include conversion reaction in addition to Munchausen syndrome.

The Internet may encourage more individuals to create illness (“Munchausen by Internet”) by providing easy access to detailed medical information and emotional support in the form of chat rooms and support groups, while also leaving an electronic trail that leads to the detection of their fraud.¹¹ The widespread use of electronic medical records may actually encourage this deception. Patients have been known to create false documents that get incorporated into their record without verification while blocking the sharing of valid records between incompatible electronic record systems that document previous episodes and normal test results.¹² Patients with factitious disorders create unique ethical and legal conflicts for their physicians who are tempted to invade a patient's privacy to uncover their deception, share the truth with family members, and warn other clinicians against a patient's wishes.¹³

Study Funding

No targeted funding reported.

Disclosure

The authors report no financial conflict disclosures. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.