Registries have the potential to tackle some of the current limitations in determining the longterm impact of multiple sclerosis (MS). Online assessments using patient-reported outcomes
(PROs) can streamline follow-up enabling large-scale, long-term, cost-effective, home-based,
and patient-focused data collection. However, registry data are sparsely sampled and the
sensitivity of PROs relative to clinician-reported scales is unknown, making it hard to fully
leverage their unique scope and scale to derive insights.
This retrospective and prospective cohort study over 11 years involved 15,976 patients with
multiple sclerosis from the United Kingdom Multiples Sclerosis Register. Primary outcomes
were changes in two PROs: Multiple Sclerosis Impact Scale (MSIS-29) motor component, and
Multiple Sclerosis Walking Scale (MSWS-12). First, we investigated their validity in
measuring the impact of physical disability in multiple sclerosis, by looking at their sensitivity
to disease subtype and duration. We grouped the available records (91,351 for MSIS-29 and
68,092 for MSWS-12) by these two factors, and statistically compared the resulting groups
using a novel approach based on Monte Carlo permutation analysis that was designed to cope
with the intrinsic sparsity of registry data. Next, we used the PROs to draw novel insights into
the developmental time course of subtypes; in particular, the period preceding the transition
from relapsing to progressive forms.
We report a robust main effect of disease subtype on the PROs and interactions of disease
subtype with duration (all p<0.0001). Specifically, PROs worsen with disease duration for all
subtypes (all p<0.0001) apart from Benign MS (MSIS-29 motor: p=0.796; MSWS-12:
p=0.983). Furthermore, the PROs of each subtype are statistically different from those of the
other subtypes at all time bins (MSIS-29 motor: all p<0.05; MSWS-12: all p<0.01) except
when comparing Relapsing Remitting MS with Benign MS and Primary Progressive MS with
Secondary Progressive MS. Notably, there were statistically significant differences between relapsing and progressive subtypes at disease onset. Critically, the PROs are sensitive to future
transitions to progressive subtypes, with individuals who transition presenting with higher
PROs in their relapsing phase compared to individuals who don’t transition since onset (all
p<0.0001).
PROs capture different patterns of physical worsening over disease length and across subtypes;
therefore, they are a valid tool to measure the physical impact of multiple sclerosis over the
long-term and cost-effectively. Furthermore, more advanced physical disability manifests
years prior to clinical detection of progressive subtypes, adding evidence to the presence of a
multiple sclerosis prodrome.