The Utility of EEG in the Emergency Department
The Utility of EEG in the Emergency Department
After many years of development and use in daily practice, the EEG remains a dependable, inexpensive and useful diagnostic tool in a number of clearly-defined emergency situations. However, eEEG ordering policies at one hospital may not be applicable to others because of differences in staffing, payment difficulties or local problems such as an EEG laboratory working only during the day. These factors usually limit performing an EEG in EDs. In a multicentre study, EEG was performed in the ED in only 3% of patients. Another more recent study found that only in 51 of 3215 patients with a seizure was an EEG ordered as a diagnostic test. Many emergency physicians prefer to order a CT scan in the ED since it is a more widely available diagnostic tool. Positive CT findings in patients with a seizure varied from 34% to 56%. The rates of abnormal CT findings in patients with chronic seizures and a first seizure in the emergency setting were not different, and approximately 7–21% of patients with chronic seizures had abnormal imaging studies. We found similar results in our study with an abnormal CT ratio of 45%, whereas the abnormal EEG ratio in all patient groups was 62.6% and in patients without a neuroimaging study an abnormal EEG was seen in 59.2%.
The American Epilepsy Society concluded that EEG should be considered as part of the routine neurodiagnostic evaluation of adults presenting with an unprovoked first seizure as a level B recommendation. The timing of the study after the initial event could influence the value of the results. In the study by King et al an EEG performed within 24 h after a first seizure detected epileptiform abnormalities in 51% of patients compared with only 34% of those who had a later EEG. Abnormal EEG ratios of 70.7%, 45% and 69% have been reported in other studies in which EEG recordings were conducted within 48 h. There is also some evidence in children that an EEG performed within 24 h of a presenting seizure gives a higher yield of significant abnormalities. In our study, EEG recording was performed within 16 h and abnormal EEG findings were determined in 73.5% of patients with a first seizure. It is also well documented that epileptiform abnormalities in the EEG increase the risk of seizure recurrence and an early EEG after the initial event is more valuable for detecting the risk. We did not evaluate the recurrence of seizures from retrospective chart reviews in this study.
Other than a first seizure, it is well described that the EEG is valuable for establishing the nature of a seizure disorder and guiding optimal therapy in such patients or in patients with recurrent seizures or established epilepsy. In the present study, abnormal EEGs were found in 61.1% of patients with recurrent seizures.
Many providers disagree about the indications for EEG and there is no clear consistency between centres with regard to which clinical syndromes are appropriate for emergency study. Despite ongoing disagreements, the most common reasons for ordering an EEG are a change in mental status and coma for diagnosing non-convulsive status epilepticus and subtle convulsive status epilepticus and determination of cerebral death. Even though none of the papers mentioned above were solely searching ED patients, it is well documented that eEEG is indicated in cases of unexplained and prolonged altered consciousness because it is the only test for diagnosing non-convulsive status epilepticus and the American College of Emergency Physicians recommends considering EEG in these patients as a level C recommendation. The other indications still remain controversial. In the present study EEG provided a statistically significant diagnostic help in patients with trauma and those with an encephalomalasia or old trauma sign on the CT scan. These findings have similarities with the study of Legros et al. However, patient numbers in other clinical situations are too small to draw definite conclusions about the indications for eEEG.
It is therefore difficult to define the diagnostic yield of a procedure simply by the number of positive diagnoses. As in daily medical practice, EEG should be considered as part of the clinical approach and the utility of the test should be discussed with regard to how it influences patient management. Praline et al studied the clinical benefits by asking the physicians who ordered the EEG whether the results of the eEEG modified, confirmed or ruled out the suspected diagnosis and if they changed the management of the patient. In 77.5% of cases the clinician considered the EEG had contributed to the diagnosis by establishing (36%) or ruling out (64%) the clinically suspected diagnosis and, overall, the EEG resulted in a modification in treatment in 37.8% of cases. In the present study abnormal EEG findings resulted in significantly more inpatient treatment and, interestingly, abnormalities showing slow encephalopathic discharges rather than epileptic discharges resulted in more hospital admissions. Because of the retrospective nature of the study, the effects of an abnormal EEG on patient evaluation and treatment cannot be assessed. A prospective study to determine the effects of EEG on patient management and physician behaviour is currently ongoing and the results will be reported soon.
When considering all the results of the present study, the clinical situations for ordering an EEG from the ED as eEEG—specified as being undertaken within 1 h with results within 3 or 4 h—are not clear. EEG may be deferred to the outpatient as was suggested in previous studies. However, when considering this approach, physicians should determine their own local national and hospital operating systems. In a study conducted in London, 83% of patients were discharged from the ED after the first seizure with a letter to take to their general practitioner and only 20% were referred directly to a neurology clinic. Of the patients seen as outpatients, the average time to having an MRI scan and EEG performed were 12 and 15 weeks, respectively. In such a situation, arranging EEG recordings to be done within 24 h or 'as soon as possible' for the patient from the ED may be better management.
Discussion
After many years of development and use in daily practice, the EEG remains a dependable, inexpensive and useful diagnostic tool in a number of clearly-defined emergency situations. However, eEEG ordering policies at one hospital may not be applicable to others because of differences in staffing, payment difficulties or local problems such as an EEG laboratory working only during the day. These factors usually limit performing an EEG in EDs. In a multicentre study, EEG was performed in the ED in only 3% of patients. Another more recent study found that only in 51 of 3215 patients with a seizure was an EEG ordered as a diagnostic test. Many emergency physicians prefer to order a CT scan in the ED since it is a more widely available diagnostic tool. Positive CT findings in patients with a seizure varied from 34% to 56%. The rates of abnormal CT findings in patients with chronic seizures and a first seizure in the emergency setting were not different, and approximately 7–21% of patients with chronic seizures had abnormal imaging studies. We found similar results in our study with an abnormal CT ratio of 45%, whereas the abnormal EEG ratio in all patient groups was 62.6% and in patients without a neuroimaging study an abnormal EEG was seen in 59.2%.
The American Epilepsy Society concluded that EEG should be considered as part of the routine neurodiagnostic evaluation of adults presenting with an unprovoked first seizure as a level B recommendation. The timing of the study after the initial event could influence the value of the results. In the study by King et al an EEG performed within 24 h after a first seizure detected epileptiform abnormalities in 51% of patients compared with only 34% of those who had a later EEG. Abnormal EEG ratios of 70.7%, 45% and 69% have been reported in other studies in which EEG recordings were conducted within 48 h. There is also some evidence in children that an EEG performed within 24 h of a presenting seizure gives a higher yield of significant abnormalities. In our study, EEG recording was performed within 16 h and abnormal EEG findings were determined in 73.5% of patients with a first seizure. It is also well documented that epileptiform abnormalities in the EEG increase the risk of seizure recurrence and an early EEG after the initial event is more valuable for detecting the risk. We did not evaluate the recurrence of seizures from retrospective chart reviews in this study.
Other than a first seizure, it is well described that the EEG is valuable for establishing the nature of a seizure disorder and guiding optimal therapy in such patients or in patients with recurrent seizures or established epilepsy. In the present study, abnormal EEGs were found in 61.1% of patients with recurrent seizures.
Many providers disagree about the indications for EEG and there is no clear consistency between centres with regard to which clinical syndromes are appropriate for emergency study. Despite ongoing disagreements, the most common reasons for ordering an EEG are a change in mental status and coma for diagnosing non-convulsive status epilepticus and subtle convulsive status epilepticus and determination of cerebral death. Even though none of the papers mentioned above were solely searching ED patients, it is well documented that eEEG is indicated in cases of unexplained and prolonged altered consciousness because it is the only test for diagnosing non-convulsive status epilepticus and the American College of Emergency Physicians recommends considering EEG in these patients as a level C recommendation. The other indications still remain controversial. In the present study EEG provided a statistically significant diagnostic help in patients with trauma and those with an encephalomalasia or old trauma sign on the CT scan. These findings have similarities with the study of Legros et al. However, patient numbers in other clinical situations are too small to draw definite conclusions about the indications for eEEG.
It is therefore difficult to define the diagnostic yield of a procedure simply by the number of positive diagnoses. As in daily medical practice, EEG should be considered as part of the clinical approach and the utility of the test should be discussed with regard to how it influences patient management. Praline et al studied the clinical benefits by asking the physicians who ordered the EEG whether the results of the eEEG modified, confirmed or ruled out the suspected diagnosis and if they changed the management of the patient. In 77.5% of cases the clinician considered the EEG had contributed to the diagnosis by establishing (36%) or ruling out (64%) the clinically suspected diagnosis and, overall, the EEG resulted in a modification in treatment in 37.8% of cases. In the present study abnormal EEG findings resulted in significantly more inpatient treatment and, interestingly, abnormalities showing slow encephalopathic discharges rather than epileptic discharges resulted in more hospital admissions. Because of the retrospective nature of the study, the effects of an abnormal EEG on patient evaluation and treatment cannot be assessed. A prospective study to determine the effects of EEG on patient management and physician behaviour is currently ongoing and the results will be reported soon.
When considering all the results of the present study, the clinical situations for ordering an EEG from the ED as eEEG—specified as being undertaken within 1 h with results within 3 or 4 h—are not clear. EEG may be deferred to the outpatient as was suggested in previous studies. However, when considering this approach, physicians should determine their own local national and hospital operating systems. In a study conducted in London, 83% of patients were discharged from the ED after the first seizure with a letter to take to their general practitioner and only 20% were referred directly to a neurology clinic. Of the patients seen as outpatients, the average time to having an MRI scan and EEG performed were 12 and 15 weeks, respectively. In such a situation, arranging EEG recordings to be done within 24 h or 'as soon as possible' for the patient from the ED may be better management.
Source...