Analyzing recurrent events in multiple sclerosis: a review of statistical models with application to the MSOAC database.

Abstract

Patients with multiple sclerosis (MS) are susceptible to experience recurrent events of disability progression and relapses. Many studies still focus on analyzing MS events with traditional methods such as Cox proportional hazards, Poisson, and logistic regression that either ignore subsequent events or fail to account for overdispersion and dependency between events. The aim of this study was to conduct a literature review to identify the main recurrent event models, with subsequent application of these models to the Multiple Sclerosis Outcome Assessments Consortium (MSOAC) placebo database. A total of nine main recurrent event models were identified, compared and applied to the MSOAC database to evaluate the effect of the disease course on the number of changes in the Expanded Disability Status Scale (EDSS) and relapse rate. Recurrent events methods have provided more precise estimates than traditional methods. Despite the similarities in common and event-specific estimates for clinical MS outcomes, the interpretations of the parameter estimates resulting from the models are different. Medical researchers should prioritize recurrent event methods in their statistical plans to avoid information loss and improve the precision of estimated effects.