Abstract
Background: Progressive supranuclear palsy is a rare neurodegenerative movement disorder and little is known about its epidemiology. Objective: Estimate age-adjusted prevalence of progressive supranuclear palsy and describe antecedent diagnoses and progressive supranuclear palsy patient features in the 5 years before first diagnostic code. Methods: In a nested case-control study in the IBM MarketScan Commercial and Medicare Supplemental Databases, a large set of US insurance databases containing medical service and prescription drug claims from employer-based commercial and Medicare supplemental health insurance plans, progressive supranuclear palsy cases (identified via International Statistical Classification of Diseases 9th/10th revision codes) and controls were included if enrollment was ≥1 month in the study period (October 1, 2015-October 31, 2017). Two controls with no diagnosis codes for PSP were matched to cases on birth year, sex, enrollment time in the database, and pharmacy benefit eligibility. Controls were assigned a randomly selected index date from their eligibility period. Prevalence of progressive supranuclear palsy was estimated in 2016 among patients with ≥1 month of continuous enrollment in that year. Prevalence ratios for comorbidities (claim/diagnosis codes) were examined in the ≤ 5 years before index date (first progressive supranuclear palsy claim date). Results: Age-adjusted progressive supranuclear palsy prevalence was 2.95/100,000 in 2016. The most common diagnosis codes in cases vs. controls in the 5 years pre-index were gait abnormalities (79.3 vs. 21.8%), pain in joint (54.9 vs. 36.0%), Parkinson's disease (54.6 vs. 1.0%), fatigue (49.8 vs. 21.6%), and cerebrovascular disease (45.6 vs. 16.4%). Conclusions: In this large database analysis, based on preliminary analyses, the prevalence of diagnosed progressive supranuclear palsy was 2.95/100,000, which is lower than many prior studies. Typical symptoms suggestive of progressive supranuclear palsy were present before index date, indicating a potential delay in time to diagnosis. The identification of diagnostic codes for clinical features of progressive supranuclear palsy that occurred before index date may be used to develop predictive models to identify potential progressive supranuclear palsy patients earlier in their disease course.