A Headache Diagnosis Project


William Pryse-Phillips, Health Sciences Centre, Faculty of Medicine, Memorial University of Newfoundland, St. John's, Newfoundland, Canada A1B 3V6.


Background.—Despite the availability of objective criteria, the diagnosis of migraine is thought to be missed frequently in primary practice.

Objective.—To determine the most important questions assisting in the clinical diagnosis of migraine headache.

Methods.—A cohort of 461 patients referred to headache specialists in Canada was assessed using a pro-forma questionnaire that was completed by the patients alone or administered by the physicians themselves. A final clinical diagnosis was recorded after a complete clinical evaluation. In a subsequent validation study, three questions derived from the results of the first phase of the study were administered to a new cohort of 128 patients, and diagnoses of “migraine” or “not migraine” were recorded according to the decision generated in the first part of the study. The final clinical diagnosis was taken as the “gold standard” for diagnosis, and the results from the two independently derived diagnostic methods were compared.

Results.—Statistical analysis of the responses from part 1 of the study yielded three questions (related to daily occurrence, unilaterally, and functional impairment) that distinguished between pure migraine and other headache diagnoses with high reliability and validity. The sensitivity and selectivity of the three-question protocol exceeded 91%.

Conclusions.—The use of three questions related to headache frequency, laterality, and impact on functioning may represent an attractive screening instrument in primary care practice, alerting physicians to the diagnosis of migraine in patients or to the possibility of a second or alternative headache diagnosis in patients in whom their diagnosis of migraine previously has been made. The presence of multiple headache syndromes in individual patients, as is common in tertiary referral practice, may reduce the discriminating power of the three-question protocol.