Estimates of incidence and prevalence are needed to determine disease risk and to plan for health service needs. Although the province of Nova Scotia, Canada is located in a region considered to have a high prevalence of multiple sclerosis (MS), epidemiologic data are limited.
We aimed to validate an administrative case definition for MS and to use this to estimate the incidence and prevalence of MS in Nova Scotia.
We used provincial administrative claims data to identify persons with MS. We validated administrative case definitions using the clinical database of the province's only MS Clinic; agreement between data sources was expressed using a kappa statistic. We then applied these definitions to estimate the incidence and prevalence of MS from 1990 to 2010.
We selected the case definition using ≥7 hospital or physician claims when >3 years of data were available, and ≥3 claims where less data were available. Agreement between data sources was moderate (kappa = 0.56), while the positive predictive value was high (89%). In 2010, the age-standardized prevalence of MS per 100,000 population was 266.9 (95% CI: 257.1- 277.1) and incidence was 5.17 (95% CI: 3.78-6.56) per 100,000 persons/year. From 1990-2010 the prevalence of MS rose steadily but incidence remained stable.
Administrative data provide a valid and readily available means of estimating MS incidence and prevalence. MS prevalence in Nova Scotia is among the highest in the world, similar to recent prevalence estimates elsewhere in Canada.