The incidence and prevalence of thyroid disease do not differ in the multiple sclerosis and general populations: a validation study using administrative data.

BACKGROUND Prior studies of a possible increased risk of autoimmune thyroid disease (AIT) in multiple sclerosis (MS) are inconsistent. We aimed to validate and apply administrative case definitions for the surveillance of AIT in MS. METHODS We used administrative health data to identify 4,192 persons with MS and an age-, sex- and geographically matched general population cohort (n = 20,940). We developed case definitions for AIT using International Classification of Disease-9/10 codes and prescription claims, compared them to medical records and applied them to estimate the incidence and prevalence of AIT. RESULTS When compared to medical records, the administrative case definition using ≥1 hospital or ≥2 physician or ≥2 prescription claims had a sensitivity of 73.5% and specificity of 98.4%. In 2005, the age-adjusted prevalence of AIT was 9.51% [95% confidence interval (CI) 8.46-10.6] in the MS population and 8.56% (95% CI 8.11-9.02) in the general population. The age-adjusted incidence of AIT per 100,000 persons per year was 422.8 (95% CI 204.4-641.3) in the MS population and 407.7 (95% CI 308.5-506.9) in the general population. From 1996 to 2005, the prevalence of AIT rose in both populations. CONCLUSION Administrative data can be used for surveillance of AIT in MS. The incidence and prevalence of thyroid disease are similar in the MS and general populations.