Home > Browse Emergency Issues
<< Seizures and Status Epilepticus: Diagnosis and Management in the Emergency Department
Diagnostic Testing
Laboratory
A number of studies have shown that history and physical examination predict a majority of patients with laboratory abnormalities. 45, 46, 47 Patients in status epilepticus or who have persistent altered mental status, fever, or a new non-focal neurological deficit require additional observation and extensive evaluation and diagnostic testing. Tests include a determination of serum glucose, electrolytes, urea nitrogen, creatinine, magnesium, phosphate, calcium, complete blood count, pregnancy tests in women of childbearing age, anti-epileptic drug levels, liver function tests, and drugs of abuse screening.48
A seizure is a physiological stressor, so a thorough understanding of normal lab abnormalities is critical for accurate analysis. If an arterial blood gas analysis (ABG) is obtained in a convulsing patient (and this is not routinely indicated), it may show an anion gap metabolic acidosis, usually secondary to lactic acidosis.28 The anion gap acidosis should resolve in less than one hour after the seizure ends; persistence beyond this time suggests an underlying abnormality. If the anion gap does not resolve, one should test for ketosis (alcoholic or diabetic) and consider poisoning (methanol, iron, isoniazid, ethylene glycol, salicylates, carbon monoxide or cyanide).
Most of the controversy regarding diagnostic testing involves healthy patients presenting after first-time seizures who return to an alert, normal baseline without focal findings on neurologic exam. If a patient with a new-onset seizure has no significant co-morbid disease and a normal examination (including a normal mental status), the likelihood of an electrolyte disorder is very low. In one prospective study of 136 patients with new onset seizures, only two had electrolyte abnormalities not suspected on the basis of history and physical examination, (both had hypoglycemia). 47 Unexpected hyponatremia has been reported rarely, 1 in 98 in one study49 and 1 in 247 in another.50
Women of childbearing age require a pregnancy test. Pregnancy causes significant physiologic stress and thus can lower the seizure threshold in a patient with an underlying focus. Approximately 25% of patients with new onset seizures in pregnancy are diagnosed with gestational epilepsy, i.e., seizure disorder that occurs only during pregnancy.51
Identification of pregnancy in a patient with a first time seizure has significant impact on disposition, initiation of therapy, and further testing. Non-eclamptic pregnant patients with new onset seizures and no co-morbidities, such as drug use or HIV, require a neuroimaging study and an electroencephalogram (EEG). If these tests are normal, it is reasonable to observe the patient without initiating therapy.51
There are no good prospective studies in either children or adults at this time to support more indepth routine laboratory testing, such as serum calcium, magnesium, or phosphate levels, in otherwise healthy patients evaluated in the ED.51
Based on a systematic review of the literature, the American College of Emergency Physicians (ACEP) has published a Clinical Policy on the initial approach to patients presenting with a chief complaint of seizure: they do not recommend extensive metabolic testing in patients with a first time seizure who have returned to a normal baseline.48 The ACEP Clinical Policy recommends that only a serum glucose and sodium level as well as a pregnancy test in women of childbearing age be done on patients who are otherwise healthy adults with a new-onset seizure who return to baseline neurological status. This conclusion has also been reached in a practice parameter published by the American Academy of Neurology on the evaluation of first time seizures in children. 52
Toxicological Testing
A drug-of-abuse screen and alcohol level are considerations in patients with first time seizures, though there is no evidence that such testing changes outcome. 53, 54, 55 A positive drug of abuse screen does not prove a cause and effect and the patient would still require an EEG and neuroimaging study in order to direct management. Pesola et al reported four cases of cocaine related seizures in 120 patients studied, though not all patients received the same tests nor was a direct correlation demonstrated.55 Seizure due to alcohol intoxication or withdrawal is a diagnosis of exclusion, as alcoholics are at increased risk for electrolyte abnormalities and traumatic injuries.56 Testing for drugs of abuse and alcohol is of unknown benefit, but may suggest an etiology and help with future medical and psychiatric management.
Other Laboratory Testing
At times, the emergency physician is faced with a patient who suffered an unexplained loss of consciousness but the lack of witnesses make it hard to determine the etiololgy. Several laboratory tests have been studied, however, no reliable indicator has been found.
Creatine kinase (CK) has been suggested as a possibility based on a retrospective study.57 However, in another smaller but more rigorous trial (using video EEG recordings) the authors did not find serum CK levels to be useful in differentiating seizures from other causes of loss of consciousness.58 Elevated CK may indicate rhabdomyolysis in prolonged seizure, so the study may prove to be useful, but not in determining the cause of unexplained loss of consciousness or seizure-like activity.
Electrocardiogram (ECG)
Patients who continue to seize or those suspected of overdose may benefit from cardiac monitoring. An ECG is also an early screen for drug toxicity.59 Tricyclic cardiotoxicity may manifest as a QRS complex greater than 0.10 seconds or a rightward shift of the terminal 40 ms of the frontal plane QRS complex (a prominent R wave in lead AVR).60 The ECG also may identify a prolonged QT, a delta wave, or heart block.
Neuroimaging
While there is agreement that neuroimaging in patients with a first-time non-febrile seizure is indicated, the timing of head computed tomograph (CT) is controversial. Three to 41% of patients with first time seizures will have an abnormal head CT.50 In one retrospective review, 22% of patients with a first-time seizure who had a normal neurological examination had an abnormal head CT scan result. 61 The question remains whether identifying such abnormalities in the ED has an impact on outcomes for patients with a normal neurologic exam and new onset seizures. This, of course, depends on the outcome measure used. Identifying an intracranial lesion may influence disposition and argues in favor of ED neuroimaging.
In a multidisciplinary collaboration between emergency medicine, neurology, and neuroradiology, an evidence-based clinical policy on neuroimaging of patients with a first time seizure was published in 1996, Table 3.62 It was recommended that a head CT be performed acutely whenever an acute intracranial process is suspected and in patients with a history of acute head trauma, 63 malignancy, immunocompromise, 55 fever, persistent headache, anticoagulation, or a new focal neurologic examination, age over 40,50 or focal onset with secondary generalization. For all other patients, acute neuroimaging is probably beneficial but not mandatory, and may be deferred if scanning is not immediately available.
Magnetic resonance imaging (MRI) is generally the preferred diagnostic test by neurologists in evaluating first time seizure since it is better than CT in identifying small lesions. MRI is not better than CT for detecting acute hemorrhage. There are no ED based studies that have evaluated MRI. Also, the joint practice guideline discussed previously deferred on making a recommendation regarding emergent MRI.
Lumbar Puncture
No prospective studies suggest the need for routine lumbar puncture in the evaluation of new-onset seizures in patients who are alert, oriented, asymptomatic, and not immunocompromised. Lumbar puncture should be considered in patients with fever, severe headache, or persistently altered mental status. 66, 50, 64 Of note, asymptomatic patients with a history or strong suspicion of immunocompromise should get a lumbar puncture at some point in their inpatient evaluation. In a prospective cohort, Sempere et al reported on eight HIV patients found to have a CNS infection as a cause of their seizure, two of whom were afebrile with no meningeal signs.49 No reports of bacterial central nervous system infection presenting as an isolated seizure without fever or an abnormal neurological exam were identified in a MEDLINE search. However, an exception may occur in cases of partially treated meningitis. In children, it has been demonstrated that those who have been taking antibiotics and present with a complaint of seizure may have meningitis even if afebrile; therefore, lumbar puncture should be considered.65 One retrospective study reported four cases of meningitis in seizure patients with normal physical exams, but none were bacterial: the majority of patients in this study did not receive a lumbar puncture and indications for lumbar puncture were not clear.66
A transient CSF pleocytosis of up to 20 WBC/mm3 has been reported in 2 to 23% of patients with seizures.27, 66 However, the emergency physician is obligated to assume that the presence of white cells in the CSF of a seizing patient represents meningitis until proven otherwise.
Electroencephalography (EEG)
While the EEG is the definitive test for diagnosing a seizure disorder, defining clear indications for obtaining an EEG in the ED has been problematic. It can certainly be helpful in cases where the diagnosis is in doubt, such as acute confusional states and coma.67, 68
In one series, nonconvulsive status epilepticus (NCSE) was detected in 8% of all patients who present with coma.69 In addition to making the diagnosis, the EEG can be helpful with monitoring medication effects and recurrence of seizures. The EEG has a role in critical care monitoring of patients with pharmacologically induced sedation, paralysis, coma, and refractory status epilepticus. An EEG should be used to monitor patients who initially had a motor seizure and have persistent altered mental status after the episode. One study found that continued electrical activity occurred in 14% of patients initially treated for GCSE 70 and NCSE was detected in 8% of all comatose patients.74 In the VA Cooperative trial, performance of early EEG found that continued electrical activity occurred in 25% of patients in whom the seizure was thought to be terminated by bedside observation.4 Delay in diagnosis of subtle status epilepticus was strongly associated with mortality in one ICU based study.71
A number of studies have shown that history and physical examination predict a majority of patients with laboratory abnormalities. 45, 46, 47 Patients in status epilepticus or who have persistent altered mental status, fever, or a new non-focal neurological deficit require additional observation and extensive evaluation and diagnostic testing. Tests include a determination of serum glucose, electrolytes, urea nitrogen, creatinine, magnesium, phosphate, calcium, complete blood count, pregnancy tests in women of childbearing age, anti-epileptic drug levels, liver function tests, and drugs of abuse screening.48
A seizure is a physiological stressor, so a thorough understanding of normal lab abnormalities is critical for accurate analysis. If an arterial blood gas analysis (ABG) is obtained in a convulsing patient (and this is not routinely indicated), it may show an anion gap metabolic acidosis, usually secondary to lactic acidosis.28 The anion gap acidosis should resolve in less than one hour after the seizure ends; persistence beyond this time suggests an underlying abnormality. If the anion gap does not resolve, one should test for ketosis (alcoholic or diabetic) and consider poisoning (methanol, iron, isoniazid, ethylene glycol, salicylates, carbon monoxide or cyanide).
Most of the controversy regarding diagnostic testing involves healthy patients presenting after first-time seizures who return to an alert, normal baseline without focal findings on neurologic exam. If a patient with a new-onset seizure has no significant co-morbid disease and a normal examination (including a normal mental status), the likelihood of an electrolyte disorder is very low. In one prospective study of 136 patients with new onset seizures, only two had electrolyte abnormalities not suspected on the basis of history and physical examination, (both had hypoglycemia). 47 Unexpected hyponatremia has been reported rarely, 1 in 98 in one study49 and 1 in 247 in another.50
Women of childbearing age require a pregnancy test. Pregnancy causes significant physiologic stress and thus can lower the seizure threshold in a patient with an underlying focus. Approximately 25% of patients with new onset seizures in pregnancy are diagnosed with gestational epilepsy, i.e., seizure disorder that occurs only during pregnancy.51
Identification of pregnancy in a patient with a first time seizure has significant impact on disposition, initiation of therapy, and further testing. Non-eclamptic pregnant patients with new onset seizures and no co-morbidities, such as drug use or HIV, require a neuroimaging study and an electroencephalogram (EEG). If these tests are normal, it is reasonable to observe the patient without initiating therapy.51
There are no good prospective studies in either children or adults at this time to support more indepth routine laboratory testing, such as serum calcium, magnesium, or phosphate levels, in otherwise healthy patients evaluated in the ED.51
Based on a systematic review of the literature, the American College of Emergency Physicians (ACEP) has published a Clinical Policy on the initial approach to patients presenting with a chief complaint of seizure: they do not recommend extensive metabolic testing in patients with a first time seizure who have returned to a normal baseline.48 The ACEP Clinical Policy recommends that only a serum glucose and sodium level as well as a pregnancy test in women of childbearing age be done on patients who are otherwise healthy adults with a new-onset seizure who return to baseline neurological status. This conclusion has also been reached in a practice parameter published by the American Academy of Neurology on the evaluation of first time seizures in children. 52
Toxicological Testing
A drug-of-abuse screen and alcohol level are considerations in patients with first time seizures, though there is no evidence that such testing changes outcome. 53, 54, 55 A positive drug of abuse screen does not prove a cause and effect and the patient would still require an EEG and neuroimaging study in order to direct management. Pesola et al reported four cases of cocaine related seizures in 120 patients studied, though not all patients received the same tests nor was a direct correlation demonstrated.55 Seizure due to alcohol intoxication or withdrawal is a diagnosis of exclusion, as alcoholics are at increased risk for electrolyte abnormalities and traumatic injuries.56 Testing for drugs of abuse and alcohol is of unknown benefit, but may suggest an etiology and help with future medical and psychiatric management.
Other Laboratory Testing
At times, the emergency physician is faced with a patient who suffered an unexplained loss of consciousness but the lack of witnesses make it hard to determine the etiololgy. Several laboratory tests have been studied, however, no reliable indicator has been found.
Creatine kinase (CK) has been suggested as a possibility based on a retrospective study.57 However, in another smaller but more rigorous trial (using video EEG recordings) the authors did not find serum CK levels to be useful in differentiating seizures from other causes of loss of consciousness.58 Elevated CK may indicate rhabdomyolysis in prolonged seizure, so the study may prove to be useful, but not in determining the cause of unexplained loss of consciousness or seizure-like activity.
Electrocardiogram (ECG)
Patients who continue to seize or those suspected of overdose may benefit from cardiac monitoring. An ECG is also an early screen for drug toxicity.59 Tricyclic cardiotoxicity may manifest as a QRS complex greater than 0.10 seconds or a rightward shift of the terminal 40 ms of the frontal plane QRS complex (a prominent R wave in lead AVR).60 The ECG also may identify a prolonged QT, a delta wave, or heart block.
Neuroimaging
While there is agreement that neuroimaging in patients with a first-time non-febrile seizure is indicated, the timing of head computed tomograph (CT) is controversial. Three to 41% of patients with first time seizures will have an abnormal head CT.50 In one retrospective review, 22% of patients with a first-time seizure who had a normal neurological examination had an abnormal head CT scan result. 61 The question remains whether identifying such abnormalities in the ED has an impact on outcomes for patients with a normal neurologic exam and new onset seizures. This, of course, depends on the outcome measure used. Identifying an intracranial lesion may influence disposition and argues in favor of ED neuroimaging.
In a multidisciplinary collaboration between emergency medicine, neurology, and neuroradiology, an evidence-based clinical policy on neuroimaging of patients with a first time seizure was published in 1996, Table 3.62 It was recommended that a head CT be performed acutely whenever an acute intracranial process is suspected and in patients with a history of acute head trauma, 63 malignancy, immunocompromise, 55 fever, persistent headache, anticoagulation, or a new focal neurologic examination, age over 40,50 or focal onset with secondary generalization. For all other patients, acute neuroimaging is probably beneficial but not mandatory, and may be deferred if scanning is not immediately available.
Magnetic resonance imaging (MRI) is generally the preferred diagnostic test by neurologists in evaluating first time seizure since it is better than CT in identifying small lesions. MRI is not better than CT for detecting acute hemorrhage. There are no ED based studies that have evaluated MRI. Also, the joint practice guideline discussed previously deferred on making a recommendation regarding emergent MRI.
Lumbar Puncture
No prospective studies suggest the need for routine lumbar puncture in the evaluation of new-onset seizures in patients who are alert, oriented, asymptomatic, and not immunocompromised. Lumbar puncture should be considered in patients with fever, severe headache, or persistently altered mental status. 66, 50, 64 Of note, asymptomatic patients with a history or strong suspicion of immunocompromise should get a lumbar puncture at some point in their inpatient evaluation. In a prospective cohort, Sempere et al reported on eight HIV patients found to have a CNS infection as a cause of their seizure, two of whom were afebrile with no meningeal signs.49 No reports of bacterial central nervous system infection presenting as an isolated seizure without fever or an abnormal neurological exam were identified in a MEDLINE search. However, an exception may occur in cases of partially treated meningitis. In children, it has been demonstrated that those who have been taking antibiotics and present with a complaint of seizure may have meningitis even if afebrile; therefore, lumbar puncture should be considered.65 One retrospective study reported four cases of meningitis in seizure patients with normal physical exams, but none were bacterial: the majority of patients in this study did not receive a lumbar puncture and indications for lumbar puncture were not clear.66
A transient CSF pleocytosis of up to 20 WBC/mm3 has been reported in 2 to 23% of patients with seizures.27, 66 However, the emergency physician is obligated to assume that the presence of white cells in the CSF of a seizing patient represents meningitis until proven otherwise.
Electroencephalography (EEG)
While the EEG is the definitive test for diagnosing a seizure disorder, defining clear indications for obtaining an EEG in the ED has been problematic. It can certainly be helpful in cases where the diagnosis is in doubt, such as acute confusional states and coma.67, 68
In one series, nonconvulsive status epilepticus (NCSE) was detected in 8% of all patients who present with coma.69 In addition to making the diagnosis, the EEG can be helpful with monitoring medication effects and recurrence of seizures. The EEG has a role in critical care monitoring of patients with pharmacologically induced sedation, paralysis, coma, and refractory status epilepticus. An EEG should be used to monitor patients who initially had a motor seizure and have persistent altered mental status after the episode. One study found that continued electrical activity occurred in 14% of patients initially treated for GCSE 70 and NCSE was detected in 8% of all comatose patients.74 In the VA Cooperative trial, performance of early EEG found that continued electrical activity occurred in 25% of patients in whom the seizure was thought to be terminated by bedside observation.4 Delay in diagnosis of subtle status epilepticus was strongly associated with mortality in one ICU based study.71
Sign up for free access