The rising prevalence and changing age distribution of multiple sclerosis in Manitoba

OBJECTIVE: Several studies suggest an increasing prevalence of multiple sclerosis (MS) in Canada. We aimed to validate a case definition for MS using administrative health insurance data, and to describe the incidence and prevalence of MS in Manitoba, Canada. METHODS: We used provincial administrative claims data to identify persons with demyelinating disease using International Classification of Diseases 9/10 codes and prescription claims. To validate the case definition, questionnaires were mailed to 2,000 randomly selected persons with an encounter for demyelinating disease, requesting permission for medical records review. We used diagnoses abstracted from medical records as the gold standard to evaluate candidate case definitions using administrative data. RESULTS: From 1984 to 1997, cases of MS using claims data were defined as persons with > or = 7 medical contacts for MS. From 1998 onward, cases were defined as persons with > or = 3 medical contacts. As compared to medical records, this definition had a positive predictive value of 80.5% and negative predictive value of 75.5%. From 1998 to 2006, the average age- and sex-adjusted annual incidence of MS per 100,000 population was 11.4 (95% confidence interval CI 10.7-12.0). The age-adjusted prevalence of MS per 100,000 population increased from 32.6 (95% CI 29.4-35.8) in 1984 to 226.7 (95% CI 218.1-235.3) in 2006, with the peak prevalence shifting to older age groups. CONCLUSION: The prevalence of multiple sclerosis (MS) in Manitoba is among the highest in the world. The rising prevalence with minimally changing incidence suggests improving survival. This study supports the use of administrative data to develop case definitions and further define the epidemiology of MS.