How Artificial Intelligence is Unlocking Unstructured Healthcare Data to Help Transform Neurodegenerative Disease Treatment

For the roughly 1 million people in the U.S. living with multiple sclerosis (MS), the three-character ICD-
10-CM code ‘G35’ will often become a catch-all for a panoply of symptoms, various disease subtypes
and unique patient experiences that make the disease difficult to diagnose and effectively treat. The single code, which is used to identify the disease for billing and reimbursement purposes in an electronic health record (EHR), does not confer any detail about disease severity or progression, but rather confirms that a patient has MS. Unlike other conditions for which far more structured, systematized coding and measurement have been implemented over the years, MS, and many other neurological conditions, live in a world of nuance and clinician-specific interpretation that can make data-driven approaches to understanding the disease incredibly difficult.

The absence of structured, standardized information in EHRs is a phenomenon that is not often discussed in conversations about the revolution in real-world evidence (RWE) and patient-centric data. Over the last decade, this type of data-driven analysis has made it possible to transform clinical trial design, revolutionize health economics and outcomes research (HEOR), and even drive regulatory approvals with synthetic control arms informed by RWE. Across dozens of therapeutic areas, RWE has made it possible to understand detailed patient journeys at a scale and level of detail that was never before possible.