A study to investigate the prevalence of headache disorders and migraine conducted using medical claims data and linked results from online surveys: post-hoc analysis of other headache disorders

Abstract

Abstract Background Surveys using questionnaires to collect epidemiologic data may be subject to misclassification. Here, we analyzed a headache questionnaire to evaluate which questions led to a classification other than migraine. Methods Anonymized surveys coupled with medical claims data from individuals 19–74 years old were obtained from DeSC Healthcare Inc. to examine proportions of patients with primary headache disorders (i.e.; migraine, tension-type headache, cluster headache, and “other headache disorders”). Six criteria that determined migraine were used to explore how people with other headache disorders responded to these questions. Results Among the 21480 respondents, 7331 (34.0%) reported having headaches. 691 (3.2%) respondents reported migraine, 1441 (6.7%) had tension-type headache, 21 (0.1%) had cluster headache, and 5208 (24.2%) reported other headache disorders. Responses of participants with other headache disorders were analyzed, and the top 3 criteria combined with “Symptoms associated with headache” were “Site of pain” (7.3%), “Headache changes in severity during daily activities” (6.4%), and the 3 criteria combined (8.8%). The symptoms associated with headache were “Stiff shoulders” (13.6%), “Stiff neck” (9.4%), or “Nausea or vomiting” (8.7%), Photophobia” (3.3%) and “Phonophobia” (2.5%). Conclusions Prevalence of migraine as diagnosed by questionnaire was much lower than expected while the prevalence of “other headache” was higher than expected. We believe the reason for this observation was due to misclassification, and resulted from the failure of the questionnaire to identify some features of migraine that would have been revealed by clinical history taking. Questionnaires should, therefore, be carefully designed, and doctors should be educated, on how to ask questions and record information when conducting semi-structured interviews with patients, to obtain more precise information about their symptoms, including photophobia and phonophobia.