RWD and machine learning methods employed to generate an Estimated EDSS (eEDSS) from OM1’s PremiOM™ Multiple Sclerosis Dataset
OM1, a leading real-world data, outcomes and technology company with a focus on chronic diseases, today announced the publication of their validation of a machinelearning approach to estimate expanded disability status scale scores for multiple sclerosis in the Multiple Sclerosis Journal by Sage Publications.
Multiple sclerosis (MS) is an auto-immune, inflammatory disease that attacks the central nervous system (CNS) and causes a variety of neurological symptoms including muscle weakness, spasticity, fatigue, numbness, vision problems, dizziness, bowel and bladder dysfunction, depression and more. Nearly 1 million Americans suffer from MS. With no cure, delay in diagnosis and treatment can result in permanent disability. Existing medications aim to modify the disease course, treat relapses, and manage symptoms.
Once diagnosed, clinicians use the Expanded Disability Status Score (EDSS) to evaluate and measure disability levels in MS patients. Scores can help determine the course of illness, need for different levels of care, and help guide treatment decisions. The EDSS considers functional impairment in seven functional systems including pyramidal, cerebellar, brainstem, sensory, bladder and bowel, visual and cognitive. Although widely used in clinical trials, its use in routine clinical practice is limited due to the time required for clinicians to complete the scale and the complexity of scoring.
To address missing functional disability assessments, RWD and machine learning methods were used to generate estimated EDSS (eEDSS) scores. For this study, OM1 extrapolated data from the OM1 PremiOM MS Dataset to amplify and expand existing clinician rated EDSS scores. Nearly 14,000 MS patients with clinical notes were screened for a clinician EDSS score that was extracted from the notes using medical language processing (MLP) methods and a test set of 684 patients with 3,489 scores was further divided into a model training cohort (75%) and a model validation cohort (25%). The results showed an area under the curve of 0.91 with positive and negative predictive values of +/-0.85 (see the full study here).
“It is exciting to see a machine learning model that performs at this level based on real-world data,” said Dr. Carl Marci, Chief Psychiatrist and Managing Director of Mental Health & Neuroscience at OM1 and one of the authors on the paper. “Importantly, it can be easily applied to a neurologist’s clinical note – saving time for clinicians and adding valuable tracking information for patients.”
The PremiOM MS Dataset is a continually updating database of more than 19,400 MS patients prospectively followed with deep clinical, laboratory and other data, such as longitudinal outcomes, Expanded Disability Status Scale (EDSS) scores, relapses, subtypes, and treatment response. Additionally, researchers can tap into data from another 485,000 MS patients in the OM1 Real-World Data Cloud™, which can be used for modeling, analytics, and other research purposes.
To learn more about the OM1 eEDSS model or the PremiOM MS Dataset, please email firstname.lastname@example.org.
This article, “Validation of a machine learning approach to estimate expanded disability status scale scores for multiple sclerosis,” by Pedro Alves, Eric Green, Michelle Leavy, Haley Friedler, Gary Curhan, Carl Marci, and Costas Boussios, and published in Multiple Sclerosis Journal by Sage Publications will be free to access for a limited time and can be read here.
Visit AITechPark for cutting-edge Tech Trends around AI, ML, Cybersecurity, along with AITech News, and timely updates from industry professionals!