Towards a validated definition of the clinical transition to secondary progressive multiple sclerosis: A study from the Italian MS Register.
Multiple sclerosis
big data
data-driven algorithm
disease registry
prognosis
secondary progressive
Journal
Multiple sclerosis (Houndmills, Basingstoke, England)
ISSN: 1477-0970
Titre abrégé: Mult Scler
Pays: England
ID NLM: 9509185
Informations de publication
Date de publication:
12 2022
12 2022
Historique:
pubmed:
17
8
2022
medline:
23
11
2022
entrez:
16
8
2022
Statut:
ppublish
Résumé
Definitions for reliable identification of transition from relapsing-remitting multiple sclerosis (MS) to secondary progressive (SP)MS in clinical cohorts are not available. To compare diagnostic performances of two different data-driven SPMS definitions. Data-driven SPMS definitions based on a version of Lorscheider's algorithm (DDA) and on the EXPAND trial inclusion criteria were compared, using the neurologist's definition (ND) as gold standard, in terms of sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), Akaike information criterion (AIC) and area under the curve (AUC). A cohort of 10,240 MS patients with ⩾5 years of follow-up was extracted from the Italian MS Registry; 880 (8.5%) patients were classified as SPMS according to the neurologist definition, 1806 (17.6%) applying the DDA and 1134 (11.0%) with the EXPAND definition. The DDA showed greater discrimination power (AUC: 0.8 vs 0.6) and a higher sensitivity (77.1% vs 38.0%) than the EXPAND definition, with similar specificity (88.0% vs 91.5%). PPV and NPV were higher using the DDA than considering EXPAND definition (37.5% vs 29.5%; 97.6% vs 94.0%). Data-driven definitions demonstrated greater ability to capture SP transition than neurologist's definition and the global accuracy of DDA seems to be higher than the EXPAND definition.
Sections du résumé
BACKGROUND
Definitions for reliable identification of transition from relapsing-remitting multiple sclerosis (MS) to secondary progressive (SP)MS in clinical cohorts are not available.
OBJECTIVES
To compare diagnostic performances of two different data-driven SPMS definitions.
METHODS
Data-driven SPMS definitions based on a version of Lorscheider's algorithm (DDA) and on the EXPAND trial inclusion criteria were compared, using the neurologist's definition (ND) as gold standard, in terms of sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), Akaike information criterion (AIC) and area under the curve (AUC).
RESULTS
A cohort of 10,240 MS patients with ⩾5 years of follow-up was extracted from the Italian MS Registry; 880 (8.5%) patients were classified as SPMS according to the neurologist definition, 1806 (17.6%) applying the DDA and 1134 (11.0%) with the EXPAND definition. The DDA showed greater discrimination power (AUC: 0.8 vs 0.6) and a higher sensitivity (77.1% vs 38.0%) than the EXPAND definition, with similar specificity (88.0% vs 91.5%). PPV and NPV were higher using the DDA than considering EXPAND definition (37.5% vs 29.5%; 97.6% vs 94.0%).
CONCLUSION
Data-driven definitions demonstrated greater ability to capture SP transition than neurologist's definition and the global accuracy of DDA seems to be higher than the EXPAND definition.
Identifiants
pubmed: 35971322
doi: 10.1177/13524585221114007
doi:
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM