Special Issues in Medicolegal Neurology 

End-of-life care

Neurologists are asked not only to diagnose brain death, but also to determine the extent and irreversibility of brain injury in various disease states. Even when other organ systems may limit survival, the brain remains essential to both the lay and medical concepts of personal integrity. For this reason, neurologists play a central role in evaluation and/or care of patients with a variety of terminal disease states.

Brain death

The concept of brain death means that "irreversible cessation of all functions of the entire brain, including the brain stem" is sufficient for determination of death. This concept is embodied in the Uniform Determination of Death Act of the United States and similar legislation that has been adopted as the standard in all 50 states and the District of Columbia.

Since clinical brain death criteria have been disseminated widely among neurologists, evaluation generally is fairly uniform, with few errors of technique or judgment. The American Academy of Neurology (AAN) practice parameter on determining brain death outlines the process in detail. Key features include the following:

1. Unresponsiveness, absent brainstem reflexes (eg, pupillary, caloric, corneal, pharyngeal), and absence of effective respiratory movements in the presence of adequate oxygenation and arterial pCO₂ of 60 mm Hg
2. Clinical or neuroradiologic evidence of an etiology adequate to explain the clinical findings
3. Adequate observation period to guarantee irreversibility
4. Exclusion of reversible factors that can confound assessment, such as drug intoxication or body core temperature less than 90°F
5. Use of serial exams or confirmatory tests (eg, EEG, blood flow studies) to assist in diagnosis in situations of clinical uncertainly, but these are not routinely required for diagnosis

The most common technical error is failure to perform an adequate apnea test, with documentation of apnea at pCO₂ of at least 60 mm Hg. Complicating factors such as severe facial trauma; toxic levels of sedatives, tricyclic agents, or neuromuscular blockers; or preexisting pulmonary disease with high CO₂ retention may create confusion requiring additional studies or neurologic opinions. Judgment and experience are required to distinguish between spinal and cerebral motor responses or effective and ineffective movements on apnea testing.

In most hospitals, organ donation is generally not broached with the family until after a secure diagnosis of brain death, so that the family generally is told clearly that the patient is now legally dead, and that any support such as pressors or ventilation will be maintained only until the family has had an opportunity to say goodbye or until organs are harvested for donation.

Rarely, families may reject the diagnosis of brain death for religious or other reasons. If relatives are uncertain or feel uncomfortable about brain death concepts on religious grounds, it is strongly recommended to bring the hospital chaplain into the picture to discuss issues with the family and also to offer to contact a specific clergyman for pastoral assistance. This helps to improve communication within the family and between the family and the medical personnel. Regulations in New York and New Jersey specifically accommodate religious or moral objections, but even in other states, hospital staff is unlikely to insist on suspending life support over vigorous family objections. Options at this point include additional consultations including an ethics committee, arranging a judicial hearing, or attempting to transfer the patient to a chronic-care facility that will honor the family's wishes.

Actual risk of malpractice suits is small in terminal care, as disagreements regarding brain death criteria or other care issues still admit for only a very limited lifespan, at least in the vast majority of cases. Since malpractice requires not only negligence but also resultant damage or loss to the patient, loss of a few more weeks in a terminal state would be considered trivial by most courts. Under extraordinary circumstances, behavior that gives the appearance of flagrant disrespect for the family or the dying patient could result in criminal penalties.

Care given prior to the patient becoming terminal remains a very different matter. For example, if a surgical complication results in cardiopulmonary arrest, the surgeon, anesthesiologist, and other parties involved prior to and during the arrest could well be sued. On the other hand, the neurologist or others brought in for evaluation or care following the irreversible injury will have negligible liability risk, possibly becoming the "white hats" who can communicate most effectively to the family, in some cases preventing needless litigation.

Vegetative state

Many medical conditions result in brain injury that fails to meet the criteria for brain death but nevertheless is clearly severe enough to leave the patient permanently incapable of satisfactory recovery. Despite good prognostic studies on traumatic and nontraumatic coma, the application of these studies to the individual case may be difficult, and families invariably desire a greater degree of actuarial precision than what is permitted by clinical information.

When persistent brain injury precludes adequate neurologic recovery, a series of court rulings supports the withdrawal of futile therapy, which may include nasogastric feedings, intravenous hydration, artificial ventilation, or any other modalities. Under these circumstances, agreement between all family members and the medical providers is mandatory to withdraw care, unless an "advance directive" by the patient limits therapy or a durable power of attorney for health care (DPAHC) specifically designates one person to be the decision maker. In the absence of formal documents, any credible statement of the patient's prior wishes should be the controlling consideration.

Forceful persuasion to withdraw care is usually ill advised, as the underlying disagreement may arise from religious or cultural issues that are unlikely to be resolved by physician pressure, and if some family members feel "brow-beaten" they are likely to feel guilty later, and may become litigious as an outlet for their own feelings. Involvement of pastoral counseling can also be helpful if sought at this point. If indecision continues, the wiser course is to arrange for tracheostomy and transfer the patient to a long-term ventilator facility as soon as practical.

Unfortunately, organ donation is currently impermissible in the absence of full brain death. Despite the logic of expanding the donor pool by including patients in persistent vegetative states, no change on this issue is likely in the near future.

Terminal care

Neurologists may become involved in the end-stage care of many patients, including those with primary neurologic disease such as motor neuron disease or brain tumors, as well as primarily nonneurologic disease that involves neurogenic pain or other neurologic impairment.

From a legal perspective, the key issue is the fine line between adequate pain control and impermissible euthanasia. In the past, many clinicians feared giving large doses of opiates for chronic pain, even in terminal illness. Both the medical literature and state-level legislation offer increasing support for adequate pain control, even at some risk of depressing respiration. The line between adequate pain control and euthanasia is quite subjective. Euthanasia is deliberate action to end a patient's life, while physician-assisted suicide (PAS) is the physician prescribing a drug or other action to facilitate a patient taking his/her own life, with the committed action taken by the patient.

PAS but not euthanasia was legalized recently in Oregon, with the initial experience published in New England Journal of Medicine. Potential concerns include prescribing by physicians with insufficient familiarity with the patient's case and possible insufficient exploration of alternative management. A blurring of distinction between PAS and euthanasia is feared, although no evidence of a breach has been presented so far. In the rest of the United States, actions that give the appearance of deliberate euthanasia rather than pain control will, in most circumstances, lead to criminal prosecution as a homicide, as well as loss of medical license. Particular circumstances may reduce chances of conviction.

In the wake of the various recent and past controversy, some states have outlawed PAS specifically, whereas in others it would be pursued under negligent homicide or a variety of criminal charges. Local legal and political environment would determine likelihood of prosecution. Civil suits by dissident relatives would fall outside most malpractice coverage as soon as criminal acts are alleged.

The ethical and legal alternative to assisted suicide or euthanasia is to search creatively with the patient and family for ways to make life bearable. This may include antidepressants, better pain control, supportive counseling, or simply information. Often, patients' fears are based on misperception of their illness and its natural history, its burden on family members, and lack of familiarity with advance directives, durable power of attorney (DPOA), and other such provisions. Asking the patient, "What are your worst fears?" and beginning with strategies to deal with the patient's specific concerns may be useful. Hospice and other adjunctive services can be critically helpful, as can spiritual counselors and resources.

More commonly, the patient and family simply need guidance on resuscitation status and overall management in a terminal condition. American medicine performs badly in this area, and neurologists should take the lead in addressing these issues. For many neurodegenerative conditions, the neurologist may be the principal care provider; even if another physician is the primary care physician, the neurologist's experience in CNS disease may still make him/her the logical leader in decision making. Patients or their surrogates (ie, family members or designated decision makers) are permitted to refuse any or all therapies, including both medically futile and medically indicated therapies.

The role of a neurologist is to initiate discussion and provide medical information. The neurologist should carefully explain reflex movements and findings that could be mistakenly construed as purposeful to the untrained lay person. The issues of persistent vegetative state, minimally conscious state, and brain death can be explained and the prognosis outlined. Often, avoiding injecting some bias is impossible, so the most honest approach is to give both sides of an issue, then state "what I would do," while acknowledging that other people's culture or beliefs might lead to different conclusions. Explaining the concrete reality of different options often is most helpful.

Asking for guidance from the family as to what the patient would want if he or she were able to participate in the discussion can be very valuable and relieve the family of the burden of making a unilateral decision. Often explaining the issue of following a loved one's last request can be helpful in putting the decision into proper ethical context.

Although economic and limited resource arguments weigh heavily with many clinicians, families rarely respond well to this approach, less so with burgeoning distrust of managed care. On the other hand, most patients and families want information on economic consequences that they will have to bear, for example, critical care transport to go "back home" if a disaster occurs while traveling. Overall, discussions are best framed in terms of how to best support the comfort and/or dignity of the patient and the family's later memories.

For excellent patient education resources, visit eMedicine's Senior Health Center. Also, see eMedicine's patient education article End-of-Life Decision Making.

Stroke affects some 750,000 persons yearly in the United States and is the leading cause of serious long-term disability. Annually over 150,000 deaths are attributed to stroke and approximately 4 million Americans live with its sequelae. In 1998, the annual national cost was in excess of $60 billion, with an average direct cost per case of approximately $59,000. The devastating impact of stroke along with rising expectations for correct diagnosis and treatment create fertile ground for malpractice claims. Stroke-associated civil suits fall basically into 2 areas—informed consent and failure to diagnose, as illustrated by a discussion of treatment issues.

Acute thrombolysis

Tissue plasminogen activator (tPA) has been shown to be effective in reducing the morbidity rate in stroke by approximately 30% when given according to a strict protocol. Efficacy of tPA gradually diminishes over time, approaching that of placebo at 180 minutes after onset of stroke symptoms. Symptomatic cerebral hemorrhage occurred in approximately 6% of cases, and the rate increased with treatment beyond 180 minutes. tPA use ordinarily is reserved for moderately severe deficits that are not improving while under observation in the emergency department (ED). Other thrombolytic agents (Pro-urokinase, Ancrod) are also effective, but await US Food and Drug Administration (FDA) approval.

Patients with acute stroke may not be able to participate fully in the informed consent process because of lethargy, impaired communication, impaired recall, or lack of cognitive appreciation of benefits and risks inherent in the use of tPA. Some would consider the risks of systemic or cerebral hemorrhage potentially excessive. As a general protection for both patients and physicians, good informed consent is essential. If the patient is unable to give consent, then the consent should be obtained from a close family member if possible. While written consent is not required ordinarily, the chart should document the discussion and the family member who was present.

Appropriate elements to discuss would include the potential benefits of using tPA, along with its risks including asymptomatic or symptomatic cerebral hemorrhage and bleeding in other organs including the possibility of death. Factors raising or lowering the risk or benefit in that patient (eg, recent surgery or other invasive procedure) should be mentioned if especially significant.

Given widespread efforts to educate the physician and lay communities regarding acute stroke ("brain attack"), asserting that tPA is now the standard of care in the appropriate setting would be easy. This puts great burden on the hospital, the ED staff, and consulting neurologists to provide appropriate care or rapid transfer to another facility. As such, an effort is now in place across the United States to certify hospitals as primary stroke centers that meet basic standards of stroke care including the availability of CT scanning, neurological evaluations, and timely handling of acute stroke patients. Many hospitals have developed a "Code Stroke" approach in this fashion with standardized protocols that allow for rapid implementation of stroke evaluation and care. In addition, certain core indicators of quality are being developed to track quality in stroke care.

If a physician elects not to use tPA, the reasons should be documented clearly. Given the potential for long-term disability, refusal of tPA by patients with acute stroke or their families is in fact somewhat unusual, although not unheard of.

Neurologists may feel pressured to use tPA by the family, by the emergency staff, or by their own desire to reduce disability. A definite onset time is critical; in its absence, the last known neurologically normal period must be used (eg, if the attack begins during sleep, onset must be dated from the night before, unless the patient got up during the night with no deficit). While full consideration of risks and benefits occasionally may mandate using tPA outside the original strict criteria, deviations should be discussed clearly with the patient and/or family and the discussion documented in the chart.

Failure by the hospital to provide essential facilities (24 hour CT technician availability, streamlined care paths in the ED) may create liability for the hospital and/or emergency physicians. In the best interest of patient care and malpractice avoidance, the neurologist should collaborate actively with both emergency staff and hospital administration to guarantee appropriate care or transfer of patients who potentially meet tPA criteria.

Whether tPA should be given by a neurologist or an emergency physician is a continuing evolving issue, with strong partisans espousing each viewpoint. The standard of care continues to evolve and probably varies with locality. Implicit in the therapeutic potential of tPA is the failure to diagnose acute stroke; it carries liability risk but only if the patient met strict criteria for tPA at the time of assessment. If additional new therapies broaden the therapeutic window, liability risk for failure to diagnose will broaden correspondingly.

Stroke prevention

Improved management of preventable risk factors for stroke and cardiovascular disease is a major victory for public health. Certain risk factors deserve special consideration.

Atrial fibrillation

Atrial fibrillation increases stroke risk by 4- to 18-fold, resulting in recommendations that all patients with atrial fibrillation receive warfarin. This aggressive approach may be tempered by a consideration of risks and benefits in the individual patient, including prior history of embolization or associated cardiac pathology that increases benefit of anticoagulation, as well as bleeding history, gait instability, or other factors that increase risk.

Standard of care for warfarin use includes documented patient education, caution regarding concomitant medication, dietary counseling, and at least monthly INR monitoring, with appropriate dosage adjustments. Even under optimum conditions, INR may deviate from intended parameters, leading to continued risk of bleeding complications. If the physician elects not to use warfarin, brief documentation of the reasons may protect against litigation should a massive embolus occur. Similarly, when warfarin is given, careful documentation of lab and clinical follow-up is essential, as well as clear documentation of patient and family education on the management and hazards of warfarin.

If a decision is made to cardiovert a patient in atrial fibrillation or other chronic or subacute arrhythmia, appropriate use of anticoagulation reduces the risk of emboli at the time of successful cardioversion. Although the effectiveness of anticoagulation for long-term prophylaxis is undisputed, controversy is increasing regarding the risk/benefit ratio of acute heparinization to prevent recurrence immediately following acute, nonseptic cardiac embolus.

Carotid stenosis

Carotid stenosis increases the relative risk of stroke; the risk increases further if symptoms are associated with the stenosis. The asymptomatic carotid stenosis study (ACSS) demonstrated a 17-21% risk of stroke in patients with symptomatic carotid artery stenosis over 3 years and 21-29% risk over 8 years. By contrast, patients with asymptomatic carotid stenosis demonstrated a 5-11% risk at 3 years and 14-26% increase at 8 years. In the North American Symptomatic Endarterectomy Trial (NASCET), carotid endarterectomy significantly reduced stroke risk compared to best medical treatment for symptomatic carotid stenosis of 70-99%. Risk reduction was less positive but still significant for patients with 50-69% stenosis.

Any patient who has had a stroke or transient ischemic attack (TIA) should have carotid Doppler or MR angiography, except patients with clear posterior circulation events or those who are not candidates for surgery. Auscultation alone cannot exclude significant disease. Patients with symptomatic carotid artery stenosis greater than 70% typically are offered carotid endarterectomy, although carotid stenting may be offered at some centers. Patients with 50-69% stenosis also may be offered surgical intervention, but comorbid features and general health must be considered with special care.

Carotid endarterectomy should be offered as soon as possible following TIA, but once a definite stroke occurs, surgery could be delayed for about 4-6 weeks to avoid stroke extension or hemorrhagic transformation in the ischemic bed depending upon the situation. Although not proven, anecdotal experience suggests that stenting may be performed safely immediately following stroke, as no anesthesia (with its attendant BP swings) is needed.

The asymptomatic carotid artery stenosis (ACAS) study demonstrated that carotid endarterectomy reduced stroke risk by approximately 53% (11% vs 5%) for arteries with greater than 60% stenosis. Carotid endarterectomy may be offered to patients with asymptomatic stenosis greater than 60% while factoring in comorbid features. A surgical team with demonstrated low morbidity/mortality rates is critical to support a decision in favor of surgery rather than medical therapy.

Malpractice risk may arise from failure to diagnose TIA or minor stroke or to search for carotid stenosis, allowing a patient to progress to a complete major stroke. Particularly for TIAs in the outpatient setting, rapid referral may prevent progress to stroke and delay may fall outside the standard of care. An early surgical intervention following stroke also may create liability risks. Failure to search for cardiac or even unusual etiologies of stroke such as arterial dissection may lead to liability but depends on the rarity of the ultimate diagnosis and the facts of the individual case—documentation of clinical reasoning and patient education is important.

Antiplatelet therapy is well established to reduce but not eliminate the risk of atherosclerotic and lacunar stroke. Failure to prescribe aspirin or one of the newer antiplatelet therapies (clopidogrel, ticlopidine, aspirin plus sustained release dipyridamole) occasionally results in liability. Antiplatelet therapy has never been proven to prevent stroke due to migraine, vasculitis, and other more "exotic" causes. While used by some clinicians for these purposes, it should not be considered standard of care. Recent data suggest that the use of combined aspirin and clopidogrel carries no additional benefit in patients with TIA or stroke and may add unnecessary morbidity.

Conclusion

Civil suits have begun to arise across the United States for both the use and nonuse of tPA as well as for the failure to diagnose carotid stenosis. Suits have also been brought for failure to give medical prophylaxis of the appropriate type in the setting of TIA. These issues typically can be avoided by fully discussing the pros and cons of various treatment options with the patient and family and obtaining informed consent for treatment. Careful documentation of reasons for either extending or withholding procedures and medical therapies also can help avert these lawsuits.

Legal issues in epilepsy

Among people with epilepsy, economic and quality-of-life studies reveal a bimodal population—a group with minimal or no disability and well-controlled seizures, and another group (as many as 50% of patients) with poorly controlled seizures and a variable combination of problems related to associated neurologic disease, medication effects, and psychological issues. Ironically, legal and advocacy issues arise for both groups, pushing back barriers for people who wish to function more normally while securing adequate protection for those with limitations. Criminal defense, general liability, and malpractice issues also may arise.

Driving

To most Americans, driving is symbolic of adult independence, as well as a practical necessity for work and family life in many geographic areas. Thus, loss of driving privileges becomes hotly contested on both practical and emotional grounds.

Procedures vary from state to state, but all states have some mechanism for balancing public and patient safety versus the liberty of the driver with medical conditions, including epilepsy, dementia, stroke, syncope, and sleep disorders. Patients with well-controlled seizures are safer drivers than the general public. Based on actuarial analysis, a 3-month seizure-free period is accepted increasingly as adequate, although some states have longer waiting periods written into law or regulations.

Clinical features such as exclusively nocturnal seizures, adequate warning auras, and specific triggers can permit safe driving even in selected patients whose seizures are controlled incompletely. Breakthrough seizures due to unusual circumstances, such as intercurrent illness, also do not impact overall driving safety. Many states permit consideration of these factors. The neurologist also should consider cognitive and motor impairments to driving associated with any underlying neurologic disorder, as well as medication side effects already demonstrated in that patient (not the theoretical list of all possible side effects as listed in the Physicians Desk Reference [PDR] or medical texts).

Generally, good-faith endorsement of a patient's driving safety is protected legally in the event of a subsequent accident with injury to the patient or others. Extent of immunity varies from state to state. Deliberate misreporting could result in liability to state sanction or civil suit by injured parties. The physician is not liable for false or misleading information provided by the patient.

Six states (California, Delaware, Nevada, New Jersey, Oregon, and Pennsylvania) require physicians to report all patients with seizure disorders, whereas the majority of states require the patients to report themselves, permit a report by patient or physician, or make other provisions. Mandatory physician reporting theoretically increases the "catch" of dangerous motorists. No statistics exists to support that mandatory physician reporting improves overall road safety, while the wedge driven into the doctor-patient relationship by such mandates is unfortunate and real. Updated information on state by state driving requirements is available on the Epilepsy Foundation of America web site at www.efa.org.

Under mandatory physician reporting, patients with seizures often feel forced to lie to their doctors regarding breakthrough seizures or to avoid contact with health care altogether for fear of administrative hassles or possible loss of driver's license or job. For the neurologist in a mandatory reporting state, the best way to maintain the doctor-patient relationship is to explain honestly the legal compulsion to report, then clearly identify all medical factors favorable to the patient (eg, good compliance, adequate warning aura) with an offer to make these part of the bureaucratic decision-making process. Depending on local procedures, the neurologist may be asked to fill out a state-provided form or to produce a short letter outlining relevant information and driving risk.

In most states, voluntary reporting is permitted by physicians of patients they consider dangerous drivers. Physician reporting usually is protected by statute from legal action by the patient, but the effectiveness of protection varies from state to state, so that advance consultation with one's state or local medical society may be desirable. Even if the irate patient is unable to sue for "wrongful reporting," hurt feelings might encourage the patient to explore other theories of litigation with an attorney; however, actual risk of suit is remote.

If a patient with epilepsy continues driving because of lack of instruction by the treating neurologist or physician failure to report where reporting is mandatory, an accident resulting from a seizure could trigger a malpractice suit by the patient or his or her estate. Best protection would be to document in the chart at the time the patient was told not to drive, and to keep a chart copy of any written notification to the state. The neurologist also should document any factors that influenced him or her (rightly or wrongly) not to report, such as "patient says he already gave up his license, sold his car." The physician is not required to pursue an independent investigation of such claims by the patient and such information mitigates (but does not necessarily eliminate) liability for nonreporting.

An emerging area of liability is third-party suit for failure to report. For example, a seizure patient causes an accident that injures or kills another driver, who then sues the patient's physician for failing to report the patient. Usually, these suits are quashed under the theory that the physician has no doctor-patient relationship with the third party, but occasionally litigation has been allowed to proceed.

Employment and epilepsy

The Americans with Disabilities Act (ADA) of 1990 protects the rights of disabled people to work. This federal legislation covers people with a substantial impairment of a major life activity, who have a history of such a disability, or who are regarded as having such a disability, if the person can perform the essential functions of a job, with or without a reasonable accommodation to the disability. Most states also have antidiscrimination laws covering medical disabilities including epilepsy.

Any employer with 15 or more employees is required to make reasonable accommodation to the disabled worker, unless the employer can demonstrate that the accommodation would cause undue hardship to the business or that the disabled worker would constitute a "direct threat" in the workplace. The risk of self-injury to the worker (eg, falling into dangerous machinery during a seizure), if accepted by the worker, does not constitute a "direct threat."

Since this legislation is recent, case law is still developing. Although legislative intent clearly covers people who are perceived as disabled, including people with epilepsy who often are stigmatized incorrectly as mentally defective, insane, or otherwise frightening, many courts are attempting to narrow coverage under the law. Patients with epilepsy may have trouble obtaining a job or may be threatened with job loss once a seizure occurs at work. If the job applicant's seizure disorder will not impede the essential functions of a job, advise the patient not to disclose any medical information until a firm job offer is received, preferably in writing. The patient should say only that he or she is able to perform the job, leaving any discussion of what is a "reasonable accommodation" until proof of the job offer is obtained.

A routine medical clearance may be required, either from a company doctor or from the patient's treating physician. If a disagreement arises about the patient's fitness for duty, further supporting medical reports may be needed. A firm job offer permits the patient to prove discrimination if the offer is withdrawn after medical evaluation, putting ADA sanctions into play. To avoid job loss or protracted litigation, the neurologist always should try to mediate between patient and employer needs. Patients with seizures may need simple accommodations regarding use of certain machinery or chemicals, or fixed rather than rotating shifts to avoid sleep loss. Patients should be allowed to accept small personal risks if they wish to do so, with written support from the neurologist to the employer.

Employers may need education that the patient is allowed to assume some risk. The neurologist's chart should reflect that the patient understands and accepts the small risk of self-injury attendant on the work prescription, protecting the neurologist against liability if the patient is injured on the job. When attainable, gainful employment tremendously boosts patient morale. Jobs for disabled people are always easiest to find during economic boom periods when unemployment is low; increasing computer skills or other skills may help employability in an appropriate job. Community colleges, state vocational rehabilitation programs, welfare-to-work programs, and local Epilepsy Foundation of America (EFA) resources all may help with skills training or job placement.

Social security and other disability benefits

When gainful work is not an attainable goal, the patient will need documentation to get a fair evaluation for benefits. Chart notes are sometimes sufficient, but often a brief summary report is more helpful. For "social security," most report charges are borne by the patient, but private disability policies may pay for the reports. Social security disability (SSD) criteria for epilepsy are under review, but if any changes are made, the current practice of counting generalized and/or partial seizures per month is likely to continue. Most accurate is to give both maximum frequency and average frequency, as the job will be lost for the worst month, not the average.

Fatigue due to medication or prolonged postictal effects should be mentioned, as should any associated depression. If the underlying brain process (eg, stroke, trauma, meningitis) leaves other nonepileptic cognitive or motor sequelae, these limitations should be mentioned explicitly too. Often, Social Security Administration staff require an independent evaluation of the patient's disability. If a patient is sent to the physician for such an evaluation, some information will be provided on relevant disability criteria. Many physicians, including neurologists, make such evaluations a significant part of their practice.

Criminal defense

Criminal defendants often claim amnesia during an assault, theft, or other criminal act. Overall, malingering and psychogenic blackout states are most likely, but complex partial seizures, postictal confusion, narcoleptic-related automatisms or other parasomnias, alcohol or drug-related blackouts, and other organic or quasi-organic conditions enter the differential diagnosis. Old brain injuries, developmental disability, and dementing conditions all may predispose to seizures, but more frequently they predispose to disinhibition, emotional lability, and other behavioral causes of violent or criminal acts.

Confused, nonpurposeful aggression is common during the postictal period, rarely during the seizure itself. Purposeful aggression, if it occurs at all during a seizure, must be extremely rare as an ictal manifestation. Many authorities feel that goal-directed, purposeful aggression is impossible during seizures. The controversies turn partly on interpretation of old, artifact-prone depth electrode studies, which probably cannot be repeated in today's environment.

Although true ictal violence is rare, in a small but significant fraction of patients with seizures, associated damage to the frontal and/or temporal limbic structures clearly disinhibits impulsive aggression as a nonepileptic, interictal phenomenon. Amnesia may or may not be associated with violent acts of this type.

The neurologist called to testify for defense or prosecution regarding these issues must consider the complexity of the violent act (ie, spontaneous, unprovoked assault vs prolonged, apparently premeditated stalking, bringing a weapon to the scene) and associated frontotemporal injury with nonepileptic impulsive violence. Remember that the standard in a criminal trial is "beyond a reasonable doubt," rather than "more likely than not" which prevails in civil cases.

Assorted liability issues

Activity restrictions other than driving may give rise to malpractice suit, for example, the patient who nearly drowns during a seizure in a bathtub or Jacuzzi. Obviously, all combinations of dangerous circumstances cannot be foreseen, but giving the patient and family some common-sense safety guidelines (written or verbal) or a referral to the local EFA office and documenting what was done (ie, "patient education was given regarding safety, referral to EFA for further information") in chart notes is prudent.

Teratogenicity: Of all complications of epilepsy, fetal damage carries the most potential for large malpractice judgments because of loss of lifetime earning capacity, lifetime medical costs, delayed statute of limitations, and other factors. All young female patients with epilepsy should be counseled regarding pregnancy issues on first or second consultation, with a brief chart note that this was discussed. EFA, AAN, and other organizations have drafted joint practice guidelines on this issue, but evolving knowledge will change the approach periodically.

Current recommendations include folic acid supplementation before conception (precise dose controversial), monotherapy when possible, and maintenance of least therapeutic level required to control generalized seizures during pregnancy. No presently available anticonvulsant is convincingly more or less teratogenic, especially when late effects on school performance are considered, but this may change with further data.

Anticonvulsant drug adverse effects: Other adverse anticonvulsant effects may provoke lawsuits, with liability often depending on local standards regarding informed consent. Stevens-Johnson syndrome, aplastic anemia, and liver failure are some of the most feared complications; fortunately all are quite rare. With aging of the population and increased awareness of osteoporosis, anticonvulsant-induced osteoporosis and pathologic fractures will become emerging areas of liability.

Imaging studies: Imaging studies are standard of care for any adult patient with new-onset epilepsy and for children with new-onset partial epilepsy. Many authorities do not recommend routine neurodiagnostic imaging of children who have clinical and EEG evidence of primary generalized epilepsy and normal findings on neurologic exam. In view of rare reports of tiny neoplasms being missed on initial CT scan, imaging probably should include MRI in all young patients with new-onset epilepsy, especially those of temporal lobe origin. Patients with long-standing seizures whose physical examination findings are stable and who have no other sign of an advancing structural lesion do not require periodic re-imaging or an MRI, if initial CT scan findings were normal.

Normal EEG findings: Recognizing that EEG results are normal in some cases of adult-onset epilepsy, most neurologists treat many patients with seizures empirically with anticonvulsants. Just as seizures may occur sporadically, abnormal EEG findings indicative of seizure risk may appear sporadically. Serial EEGs may eventually show needed findings but occasionally activation procedures like 24-hour sleep deprivation or prolonged EEG monitoring may be necessary in capturing positive abnormalities.

Rarely, cardiac arrhythmia or other serious conditions can present as apparent epilepsy, even when attacks are well described by witnesses or personally viewed by the neurologist. If such conditions lead to a medical disaster, lawsuit for failure to diagnose could ensue. Pursuing EEG evidence of seizures originating in the brain is helpful in this regard. Additional EEG studies can go forward with anticonvulsant use. The standard of care in terms of extent of appropriate workup depends on details of the individual case, but reviewing the differential diagnosis is always wise in any patient whose attacks remain refractory to adequate trials of several anticonvulsants and the absence of EEG documentation of cortical seizures stemming from brain discharges.

Seizures in the physician's office: When patients have seizures in the physician's office, any traumatic or anoxic-ischemic sequelae would be unlikely to provoke suit, but generally they would be handled by malpractice coverage. A family member or other nonpatient visitor with epilepsy is not considered to have a physician-patient relationship; however, a lawsuit against the physician's general liability policy can be possible.

Employment of persons with epilepsy in the medical office: People with epilepsy make excellent employees in the medical office. Employee injury due to on-the-job seizures is very rare. Although workers compensation law varies from state to state, employer liability generally would be restricted (at most) to treatment for associated injuries such as twisted ankles and thoracic muscle spasm.

Sudden death: Sudden death in epilepsy is rare but shocking when it occurs, both to the physician and the patient's family. Its etiology and prevention are unknown, which should eliminate direct malpractice liability. However, an irate family and an ill-informed plaintiff's attorney might get surprisingly far with a baseless suit, particularly with a young, photogenic decedent to show the jury. Most authorities recommend against including sudden death in the routine patient/family education discussions, as its extreme rarity and impossibility of prevention do not justify the induced fear. However, if the patient or family inquires directly, perhaps after the death of some prominent person with epilepsy, giving a brief discussion is best, emphasizing its rarity and its distinction from the brief respiratory cessation seen with many generalized seizures.

Patient education

For excellent patient education resources, visit eMedicine's Public Health Center. Also, see eMedicine's patient education article Patient Rights.