Identification of Migraine at Sutter Health: An Application of an EHR-Based Algorithm

Authors

Alice Jacobson, Sutter Health, Walnut Creek, CAFollow
Heather Law, Sutter Health, Walnut Creek, CA
Andrew Avins, Kaiser Permanente Division of Research, Oakland, CAFollow
Sylvia Sudat, Sutter Health, Walnut Creek, CA
Jessica Liu, Sutter Health, Walnut Creek, CA
Lisa Dean, Sutter Health, Walnut Creek, CAFollow
Alice Pressman, Sutter Health, Walnut Creek, CA

Publication Date

8-10-2017

Keywords

observational studies, primary care, chronic disease, epidemiology

Abstract

Background: Migraine is one of the most common and disabling disorders in the world. In the United States, 1 in 4 households is affected by it. With no diagnostic test, migraine is most commonly detected through patient interview and rule out of other causes of headache. Studies to identify patient populations for prevalence estimates or clinical research often involve costly and time-consuming population-based surveys or data collected from specialty headache clinics. To identify migraineurs in Sutter Health, we extracted data from electronic health records (EHR) and applied a migraine probability algorithm (MPA) that we previously developed in a separate large integrated health care system. Our goal was to identify a target population of migraineurs from which to recruit for future studies.

Methods: We identified all migraine diagnoses, migraine-specific prescriptions and migraine listed as a significant health problem from the EHR for the 5-year period November 2010 through October 2015. Applying the MPA, we calculated a migraine probability for all patients in the Sutter Health primary care population, defined as adults with one or more visits to primary care during the study period. We report prevalence estimates by age, sex and race.

Results: The prevalence of medically ascertained migraine among the Sutter Health primary care population was 7.3% overall. Migraine was more prevalent among women than men, 9.8% and 3.4% respectively, and peaked for women aged 35–54 years (12.2%). Migraine prevalence was highest among whites (8.4%) and lowest among Asians (4.4%).

Conclusion: Our study confirms that using EHR data along with the MPA is an inexpensive and easily applied method of identifying migraineurs for use in research. We have calculated prevalence estimates at Sutter Health and will use this methodology to identify potential participants for a concurrent National Institutes of Health trial of mindfulness for migraine. Although our calculated prevalence estimates are lower than those reported in the literature, we believe this is due to the restriction that patients utilize Sutter Health primary care services. To test this, we included specialty care patients and identified as many as 3% more migraineurs, thus warranting further study. Future plans include a patient survey to validate the MPA in the Sutter Health patient population.

Recommended Citation

Jacobson A, Law H, Avins A, Sudat S, Liu J, Dean L, Pressman A. Identification of migraine at Sutter Health: an application of an EHR-based algorithm. J Patient Cent Res Rev. 2017;4:199.