Methodology

How Upheld calculates scores and dollars

Transparent formulas for agency QA and revenue leaders. All dollar figures are advisory — not a certified CMS HIPPS grouper or remittance. Full technical detail also lives in the repo at docs/calculations.md.

1 · What each number means

  • Submission readiness (0–100) — how clean the packet looks for submission (integrity, not payment).
  • Expected period total — advisory full 30-day PDGM period pay for this chart (national base × case-mix weight × wage index).
  • Capture — upside if documentation/coding is fixed.
  • Protect — exposure if submitted as-is.
  • Path to 80+ — ordered open findings that drag readiness under the gate.

2 · CMS rate anchors (CY 2026)

30-day national base

$2,038

SN per-visit

$177

Illustrative LUPA gap

$1,153

A certification period is often 60 days and may include two 30-day payments. Upheld models one PDGM payment period at a time.

3 · Expected period payment

Expected $ = National base × Case-mix weight × Wage index

National base = $2,038 (CMS CY 2026).

Case-mix weight is inferred from the packet (clinical group family, comorbidity, function, timing, admission). If a known HIPPS code is present, a sparse weight table is preferred.

Wage index defaults to 1.0 (national). Agencies can set a local wage index in Settings.

Clinical group familyRelative base weight
MMTA1.02
NEURO1.14
WOUND1.18
COMPLEX1.32
MS_REHAB1.10
BEHAVIORAL0.96
DEFAULT1.00

Then multiplied by comorbidity (none/low/high), functional band, early/late timing, and community vs institutional admission factors. Weight is clamped (~0.55–2.8).

4 · Capture and protect

Each finding has an advisory $ (from CMS-scaled rules) and type RECOVERY (capture) or EXPOSURE (protect).

Finding $ is scaled to the chart’s expected period: nationalFinding$ × (E / $2,038.22), then capped at E.

Protect: denial-class findings (F2F, homebound, cert, orders) take the maximum once; other exposure is summed; result ≤ E.

Capture: recovery findings are summed and capped at E. Capture and protect are separate paths — do not add them as a new claim total.

5 · Submission readiness

Severity penalties: Critical 22 · High 12 · Medium 6 · Low 2. Each module score = 100 − sum(penalties), or 92 if no findings.

Readiness = 0.35×Clinical + 0.40×Compliance + 0.25×Revenue

Gate for the fix path: 80. Live readiness uses OPEN findings only (resolved/dismissed no longer penalize).

6 · What this is not

  • Not a certified CMS HIPPS grouper
  • Not a remittance or payment guarantee
  • Not full CMS weight files (sparse HIPPS table + signal model today)
  • Not a substitute for human clinical/compliance judgment