Healthcare Cost & Utilization Intelligence Platform
Gray counties are outside the analysis cohort (fewer than 500 Original Medicare beneficiaries, CMS suppression, or no PLACES estimate; PLACES has no 2023 estimates for Kentucky or Pennsylvania). Shapes are 2017-vintage boundaries, so Connecticut's planning regions (used by CMS and CDC since 2022) and Alaska's 2019 Chugach / Copper River split have no outline; their values are in the table.
Counties grouped into fifths by age-adjusted adult diabetes prevalence (CDC PLACES). Each column is the beneficiary-weighted mean for Original Medicare beneficiaries in those counties.
Spearman rank correlation between each PLACES measure and standardized Medicare spending per beneficiary, across cohort counties. Lines show 95% bootstrap intervals. Correlating ranks within each state removes everything a state shares (Medicaid policy, regional practice patterns), and most of the association goes with it.
Original Medicare spending per beneficiary keeps rising while the share of beneficiaries in Medicare Advantage passes half, so the FFS population behind these numbers is shrinking.
Share of all Medicare beneficiaries enrolled in Medicare Advantage
This page only ever shows data that passed every quality gate. Each run lands the source files, validates them against data contracts, builds silver and gold with dbt into a candidate database, runs the tests, and publishes only if they all pass (write-audit-publish).