Neuroimaging for Migraine: The American Headache Society Systematic Review and Evidence-Based Guideline

Randolph W. Evans, MD; Rebecca C. Burch, MD; Benjamin M. Frishberg, MD; Michael J. Marmura, MD; Laszlo L. Mechtler, MD; Stephen D. Silberstein, MD; Dana P. Turner, MSPH, PhD


Headache. 2020;60(2):318-336. 

In This Article


Authorship Committee

The AHS Guideline Committee and the Executive Committee of the Board of Directors accepted the guideline proposal of the working group chair RE and approved the working group. The working group consists of AHS members with expertise in migraine imaging and guideline development. No patients were involved in the development of this guideline. The guideline development process adhered to the standards for systematic reviews of the Institute of Medicine of the National Academies.[15] The AHS provided meeting rooms during annual society meetings and arranged for conference calls for group discussions.

Systematic Review Eligibility Criteria

A systematic review was designed to gather evidence on which to base the guideline. To be eligible for inclusion in the review, articles must have included adult females and males ages 18 and over who were seeking outpatient treatment for episodic migraine, chronic migraine, progressive migraine, migraine with aura, migraine without aura and undergoing neuroimaging (MRI or CT). English language articles published from 1973, the time of first CT use, to the time of the search were included. Meeting abstracts and case reports were excluded from the search.

Search Strategy

Two medical research librarians (recruited by Thomas N. Ward, MD) performed a search on December 5, 2016 and used the following databases: Medline (PubMed), Web of Science, and Cochrane Central Register of Controlled Trials. The search terms included the following: migraine, cluster headache, computed tomography, CT, magnetic resonance imaging, MRI, neuroimaging, imaging, diagnosis, diagnostic imaging, delayed diagnosis, differential diagnosis, early diagnosis, pathophysiology, sensitivity, specificity, predictive value, PPV, NPV, likelihood ratio, or testing. Full search strategies for each database are included in Appendix 1. The search was updated on August 31, 2018 using the same search strategy and databases.

Study Selection

The study selection process is detailed in Figure 1. After the search was conducted, RE and BF jointly screened the titles and abstracts of the returned articles and indicated appropriate articles for exclusion or for further review to answer the guideline question. RB then further applied the diagnostic criteria and selected the final list of articles for review. Additional articles were included from the reference lists of review articles and guidelines. Studies of migraine and white matter lesions, case studies, and reviews were excluded.

Figure 1.

Flow diagram of study selection process.

Data Extraction and Rating the Evidence

RB assigned 2 members of the guidelines committee as data extractors for each article (RB, RE, BF, MM, and 3 headache medicine fellows independently reviewed 7 or 8 articles each). DT served as the arbiter of this final review. The committee concurred on the use of a standardized data extraction form that included the following: study name, date of extraction, person extracting, publications type (full article or abstract), funding/conflicts of interest, study design, study location, participants/population, sample size, age (mean [SD]), gender (n (%) female), how migraine diagnosis was made, inclusion/exclusion criteria, recruitment methods, aim, dates of study, primary outcome measure, secondary outcome measure, missing data, analysis methods, primary outcome results, secondary outcome results, key conclusions. The extraction form was piloted by RB prior to use by the group. The Newcastle-Ottawa Scale adapted for cross-sectional studies was used for rating risk of bias and quality assessment.[16]

Synthesis of Results

Results were qualitatively synthesized by MM who summarized the findings of the data extraction. After extraction, we evaluated whether the results were appropriate for pooling and meta-analysis. Due to methodological and statistical heterogeneity among studies, quantitative synthesis was not appropriate.