Real-World Evidence (RWE) on Migraine Treatment in Routine Clinical Practice: Insights From the Turkiye Migraine Registry Study (TMRS) Using AI Algorithms

OBJECTIVES: Migraine is a disease in which the subjective perception by the patient plays a major role in the assessment of the condition and treatment success. Therefore, outcomes will often be stated in electronic health records (EHR) as unstructured text. To make these data usable for analysis on a population level, they must be transformed into a structured format. This study leverages AI algorithms to extract and analyze unstructured data from EHRs and aims to analyze real-world treatment patterns and patient characteristics among migraine patients in Turkey.

METHODS: We used our application to obtain data from the medical records of 11,023 adult patients that were treated for migraine or suspected migraine at 21 hospitals in Türkiye between January 1^st, 2021, and November 1^st, 2023. The study involved automated and manual verification to ensure data validity and usability. Consistency checks were conducted on the extracted data to ensure reliability. The conversion to the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) facilitated comprehensive analysis and outlier identification, supporting the study's non-interventional and AI-assisted approach.

RESULTS: We found the extracted data to be well-aligned with our expectations. An independent expert confirmed validity of the data. Migraine prevalence among neurological patients from participating hospitals was 12.2%. Of the analyzed patients, 73% were female. One third of patients received acute treatment only, another third received both acute and preventive treatment, and one third received preventive treatment. 80.1% of the patients were outpatients, 10.3% seen at emergency department only. 36% patients seek for improved acute and 27% for improved preventive therapy.

CONCLUSIONS: The TMRS provides valuable RWE that can inform stakeholders about the real-world outcomes of migraine treatments. The study's innovative use of AI underscores the potential of technology in enhancing RWE research and improving patient care.

Funding: This study was sponsored by Pfizer.