Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Electroencephalography is not useful in the routine evaluation of headache

ACP J Club. 1996 Jan-Feb;124:18. doi:10.7326/ACPJC-1996-124-1-018

Source Citations

Gronseth GS, Greenberg MK. The utility of the electroencephalogram in the evaluation of patients presenting with headache: a review of the literature. Neurology. 1995 Jul; 45:1263-7. [PubMed ID: 7617180]

Report of the Quality Standards Subcommittee of the American Academy of Neurology. Practice parameter: the electroencephalogram in the evaluation of headache (Summary statement). Neurology. 1995 Jul;45:1411-3.



To review the evidence for the routine use of the electroencephalogram (EEG) in the evaluation of patients presenting with headache.

Data Sources

English-language studies were identified using MEDLINE (1966 to March 1994) using the exploded terms electroencephalography and headache and reviewing bibliographies of identified studies.

Study Selection

Studies were excluded if they were case reports, case series in which patients were selected because of abnormal EEGs, abstracts with insufficient information for analysis, and reviews without original data.

Data Extraction

Data pertaining to the study design, patient characteristics, and test results were extracted. Articles were compared against 6 criteria for the validity of evaluating a diagnostic test: comparison against established diagnostic criteria, the use of appropriately matched controls, a study setting that minimizes biases, a blinded interpretation of the EEGs, a sensible interpretation of the abnormalities observed, and the replicability of the techniques used.

Main Results

40 articles were reviewed in detail. 1 article met all 6 criteria for determining the validity of the data, 3 articles met 5 criteria, 11 met 4, 9 met 3, 7 met 2, 7 met 1, and 2 met none. Studies designed to determine whether patients with headache have an increased prevalence of EEG abnormalities reported conflicting results. Prominent driving in response to photic stimulation (the H-response) in patients with migraine was the most consistently reported difference between patients with and without headaches. The reported sensitivity of the H-response varied from 26% to 100%, and the specificity from 80% to 91%. The H-response was less effective than clinical criteria in distinguishing patients with headaches from persons without headaches, and in distinguishing migraine from other types of headache. Studies reviewed did not show that the EEG is an effective screen for structural causes of headache, nor does the EEG effectively identify headache subgroups with different prognoses.


The electroencephalogram is not useful in the routine evaluation of patients presenting with headache.

Source of funding: Not stated.

For article reprint: Lt. Col. G.S. Gronseth, Department of Neurology (PSMN), Wilford Hall Medical Center (AETC), 2200 Bergquist Drive STE 1, Lackland Air Force Base, TX 78236, USA. FAX 210-292-6953.


The holy grail of headache medicine would be an objective, noninvasive test that could make a reliable distinction between serious and benign causes of headache and could impart information useful in treatment. As pointed out in this carefully done literature review and practice parameter issued by the American Academy of Neurology, in the general headache population, the EEG is not such a test.

The link between migraine and epilepsy has long fascinated clinicians, and it is not surprising that many patients with headaches have had EEGs. It is difficult, however, to identify cases in which EEG information substantially affects the clinical outcome. For example, no evidence suggests that patients with headaches whose EEGs show the H-response are more likely to respond to prophylactic treatment with anticonvulsant agents. Further research clearly is needed to identify subgroups of patients who benefit from EEG testing. These patients might include those with prolonged confusional states, memory lapses, or loss of consciousness. In situations in which the diagnosis is difficult to determine, clinical experience suggests that referring these patients to an epileptologist and hospitalizing them for concomitant video and EEG monitoring is more helpful than a simple EEG.

A test-oriented approach to diagnosis of headache reinforces the incorrect notion that benign headache disorders, such as migraine, tension-type, and cluster headaches, are diagnoses of exclusion. In fact, the International Headache Society (IHS) has issued widely accepted diagnostic criteria for these disorders (1). Although imperfect and subject to future revision, these criteria were designed to minimize the possibility of overlooking serious underlying conditions. Patients whose headaches meet IHS criteria for 1 of the benign headache disorders should rarely need further testing. For the average patient with headaches, the most useful test in diagnosis and treatment is still a careful history and physical examination.

Elizabeth Loder, MD
The Spaulding HospitalBoston, Massachusetts, USA


1. Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalagia. 1988;8(Suppl 7):1-96.