Impact of Approaches for Clinical and Radiological Monitoring on Predicting of Short-Term and Long-Term Disability Outcomes in Multiple Sclerosis

• Activity Number: 338 - SPEED: Biostatistical Methods, Application, and Education, Part 1
• Type: Contributed
• Date/Time: Tuesday, July 30, 2019 : 10:30 AM to 12:20 PM
• Sponsor: ENAR
• Author(s): Brian Healy*
• Companies: Biostatistics Center/Massachusetts General Hospital
• Keywords: Multiple scleoris; Neurology; MRI; Prediction

Abstract

Multiple sclerosis (MS) is a neurologic disorder that is associated with progressive declines in neurologic function over a long disease course. A key challenge in managing MS patients is the heterogeneity of the disease, which leads to difficulty in predicting the longer-term status of individual patients. Both magnetic resonance imaging (MRI) and clinical measures can be used to monitor the disease. Given the heterogeneity of the disease course, there is uncertainty regarding the best approach for following MS patients both in terms of the length of time required to perform accurate prediction and the frequency of both clinical and radiological monitoring. Using subjects from the CLIMB study, we compared potential monitoring schemes in terms of predictive ability over the short term (2 years) and long term (7 years) using random effects models to account for multiple measurements per subject. Our results showed that multiple clinical visits are important for prediction of both the short and long term, and the six-month intervals for visits led to slightly improved prediction compared to yearly intervals. MRI information added limited additional value for any of the models.