Bayesian Health's FDA-cleared sepsis monitor got a Medicare New Technology Add-on Payment in the FY2027 inpatient rule — up to $61.84 per case, across an estimated 739 MS-DRGs, billable from October 1. Here is every MS-DRG in Medicare fee-for-service inpatient, drawn to scale, so you can see where that money can actually land.
The clearance came in April. The payment came in September. That gap — not the AUC — is the thing almost nobody building clinical AI plans for.
So look at the terrain. Each dot below is one MS-DRG. Horizontal: what Medicare already pays that DRG, on average. Vertical: how many Medicare discharges it had. Dot size is the add-on dollars available — discharges × the per-case amount. Drag the amount and the whole field breathes.
DRG 871 — septicemia or severe sepsis without prolonged ventilation, with MCC — is the number-one MS-DRG in Medicare fee-for-service inpatient by discharges: 561,795 in the 2023 file, across 2,678 hospitals. Add DRGs 870 and 872 and sepsis is 687,506 discharges, 13.9% of everything. The next biggest DRG, heart failure with MCC, is 319,702.
That is why $61.84 is not obviously trivial. It is 0.41% of what DRG 871 already pays — and 0.41% of the biggest denominator in the inpatient system.
Twenty-two MS-DRGs carry half of all Medicare inpatient discharges. Seventy-four carry 80%. An add-on approved across “739 MS-DRGs” sounds like near-universal coverage. In volume terms, almost all of it is decided by whether sepsis, heart failure, pneumonia and respiratory infection are in the list. If you are modelling revenue for an inpatient tool, model those twenty-two and stop.
Push the volume floor and watch what leaves. 124 of the 534 DRGs — 23% of the dots — have fewer than 100 discharges between them and account for 0.1% of the volume. They occupy a third of the visual field and none of the money. The bottom edge of that cloud is not a finding either: CMS suppresses any hospital-DRG cell under 11 discharges in this public file, so the floor of the chart is a publication rule, not a clinical fact.
Turn on the fit line and the same discipline applies. Across all 534 DRGs the log-log correlation between average payment and volume is r = −0.10; filter to DRGs with 1,000+ discharges and it is −0.22. Neither is a relationship. Expensive DRGs are not systematically rare and cheap ones are not systematically common — which is exactly why you cannot infer a market from a price.
| MS-DRG | Discharges | Avg Medicare pay | Add-on ceiling | % of DRG |
|---|
In the same FY2027 rule, CMS approved 19 new technologies for add-on payments — three through the traditional pathway, sixteen through the alternative one — and finalized the closing of that alternative pathway. One exception survives: a device holding Breakthrough designation as of September 30, 2026 can still use the old door through FY2029, if it reaches marketing authorization by May 1, 2028. That is four weeks from today. Bayesian's monitor and Reflow Medical's Spur stent both came in on the pathway now being repealed.