Verantos' AI Advancement in Neurology Bolstered by Groundbreaking Migraine Real-World Evidence

Image Source : Public Domain

Verantos  announced two peer-reviewed publications on high-quality real-world data for migraine. The manuscripts describe the use of artificial intelligence (AI) on medical records to enable access to valuable clinical information such as headache severity, comorbid depression, associated symptoms, and a composite migraine impact score.

These soft outcomes — subjective aspects of a medical condition or its management — are poorly and inconsistently captured during the course of routine care, but generation of useful real-world evidence often requires understanding them.

Migraine impacts more than a billion people globally every year, particularly women and young adults. Migraine is linked to various comorbidities, including stress, sleep disorders, and suicidal ideation.

"Breakthroughs in AI applied to rich clinical data enable life sciences organizations to understand real-world migraine care at a level that has not previously been possible," said Dan Riskin, M.D., CEO of Verantos. "High-quality data and high-validity evidence can enable better collaboration with providers, payers, and regulators to refine the standard of care and benefit patients."

By applying AI to medical records including clinical narrative data, it is possible to build and validate models of soft outcomes. In its highest form, a system can accurately develop an outcome score quantifying migraine severity at a point in time.

"Using artificial intelligence to identify patients with migraine and associated symptoms and conditions within electronic health records" was published in BMC Medical Informatics and Decision Making. This manuscript demonstrates how detailed patient characteristics (phenotype) can be identified using unstructured EHR data and AI. "Development and validation of a novel model for characterizing migraine outcomes within real-world data" was published in The Journal of Headache and Pain. This groundbreaking work builds on phenotyping to develop and validate a model of migraine outcomes that uses information typically available in routinely collected data.