← adambecker.org

Payer Pulse

Where the Medicare book and the Medicaid book are the same members.

Report 03: The duals seam  /  companion to Report 01: Medicare and Report 02: Medicaid

CMS SNP Comprehensive Reports, monthly  /  August 2023 – July 2026, all 36 months

The finding  ·  written 4 August 2026

Centene told investors that dual-eligible members are around 40 percent of its Medicare Advantage book. CMS's files say 39.9 percent. How it got there is the finding.

D-SNP members at July 2026, and each company's D-SNP share of its own Medicare Advantage book

Dual-eligible special needs plans, D-SNPs, are Medicare Advantage plans sold specifically to people who also have Medicaid. They are the seam where this site's first two reports meet: the same person is a Medicare member in Report 01 and a Medicaid member in Report 02. Nearly 6.5 million people are enrolled in one, and half the market belongs to UnitedHealth and Humana.

The verified 40 percent is not a growth story, and saying so plainly matters. Centene's D-SNP book actually shrank 10 percent over these three years. Its share climbed from 32.6 to 39.9 percent because the non-dual rest of its Medicare Advantage book fell 34 percent, more than three times faster, which is the company's stated plan, aligning its Medicare footprint to its Medicaid footprint, carried out in the files. The quiet growth story is elsewhere: CVS nearly doubled its D-SNP book over the same window. The notes take each in turn.

Scope: D-SNP enrollment only. Many dual-eligible people are in ordinary Medicare Advantage plans or original Medicare instead, so this undercounts every company's total dual membership; it measures the product built for duals, not the population. Same four companies as the other reports. One CMS data quirk is disclosed in Panel 01's caveat.

The D-SNP market

Panel 01

D-SNP members by parent company, monthly, from CMS's SNP Comprehensive Reports. Contract-to-parent attribution reuses the same per-month mapping as Report 01, so sub-brands roll up identically. The national market grew 16.6 percent across the window, and the four companies' combined share drifted down from 65 to 63 percent as the rest of the market grew faster.

One month reads low by design: October 2023's workbook is internally inconsistent in CMS's own publication, its plan-level sheet running about 2.8 percent under its own summary because it was generated from a stale data pull. Every other month reconciles to CMS's summary exactly. The dip that month is the file, not the market.

D-SNP share of each company's own MA book

Panel 02

The same D-SNP members, divided by each company's total Medicare Advantage enrollment from the CMS contract-level file, month by month. This is the strategy view: it asks not who is big in D-SNP but whose Medicare business depends on it. Humana and UnitedHealth hold steady, between 13 and 25 percent. The other two lines both climb, for opposite reasons, which is the point of reading this panel next to Panel 01.

A rising share does not mean a growing book. Centene's line climbs 7 points because its D-SNP enrollment fell 10 percent while the non-dual rest of its book fell 34 percent, and almost all of that climb lands in two January plan-year boundaries: 4.3 points in 2024 and 2.8 in 2026, when plan exits take effect. CVS's line climbs the other way, on a D-SNP book that nearly doubled while its overall Medicare book also grew. Same direction on the chart, opposite mechanics.

Integration depth

Panel 03

Not all D-SNPs are equally connected to Medicaid. CMS classifies each as coordination-only (CO), highly integrated (HIDE), or fully integrated (FIDE), by how much of the member's Medicaid benefit the same organization actually manages. The share of each company's D-SNP book in the two integrated tiers, at July 2026, tracks how deep its Medicaid machinery goes: the company with the largest Medicaid book in Report 02 also runs the most integrated D-SNPs here.

Integration status reflects plan design and state contracts, not quality. This panel is the latest month only, by choice rather than necessity: CMS has published these columns since 2022, so the same three years of history exist and a trend could be drawn from them.

Analyst notes

Read alongside

Observations about an industry, not investment advice.