Background: Because multiple sclerosis (MS) is variable and unpredictable, if symptom worsening could be predicted, patients may feel better prepared to manage changes in function.
Objective: The objective of this paper is to study the prediction of walking impairment in MS.
Methods: We retrieved data for all MS patients at our center (2008-2009), including baseline and follow-up timed 25-foot walk (T25FW) times. We assessed the incidence of ≥20% worsening in T25FW by developing two survival models: (1) disease course and (2) Multiple Sclerosis Performance Scales (MSPS) score. The outcome was days until ≥20% worsening in T25FW. Covariates were disease subtype, years since diagnosis, Patient Health Questionnaire-9 (PHQ-9) score, and demographics. Data were interval censored; missing data were handled with multiple imputation.
Results: Of 1544 patients, 309 (20%) experienced ≥20% worsening T25FW. For disease course, time to worsening was significantly shorter for secondary progressive vs. relapsing-remitting disease (p < 0.001). For MSPS, patients with lower baseline MSPS scores progressed more slowly (p = 0.001). In both models, sex, baseline T25W, and time since diagnosis were significantly associated with worsening. In the disease course model, PHQ 9 score may be related to worsening (p = 0.07).
Conclusion: These findings suggest factors associated with worsening in T25FW and a potential approach to establishing indicators associated with clinically significant change.
Keywords: Multiple sclerosis; accelerated failure time; disease progression; electronic health record; survival analysis; timed 25-foot walk test.
© The Author(s), 2014.