Population Health Toolkit

Population Health Intervention Economics

Turn a promising intervention into a clearly labeled financial scenario.

1Know what this number is, and what it is not.

This is a scenario calculator. It estimates what follows from your assumptions; it does not prove the intervention works.

Every input has an : what to enter, an example and where to find it. All prefilled values are illustrative.

2Path A · Quick intervention scenario

3Path B · What changed beyond the comparison trend?

All four inputs must measure the same outcome, unit and aligned periods.

A comparator is necessary, but not sufficient. Parallel trends, selection, contemporaneous changes and statistical uncertainty still need evaluation. This arithmetic is not causal certification.

4MSSP truncation: when medical savings shrink before the ACO gets paid

A model that budgets every raw dollar saved can catastrophically overstate ACO revenue.
Start here · Published real-world example · Iris / Aledade, 2022

A large medical-cost difference became a much smaller ACO-level figure.

Headline medical-cost difference$994 PMPM$11,928 annualized per person
Reported after MSSP financial stop-loss$292 PMPM$3,504 annualized per person
Only 29.4% remained. The reported amount was 70.6% lower.

The authors reported an ACO-level adjustment, not an actual CMS payment. Annual figures above are derived ×12 illustrations; the report followed participants for an average of 14 months.

Read the original report and its limits

The 2022 author report describes a comparison group, 335 treated patients, and a financial stop-loss adjustment. It does not provide enough detail to independently reconstruct that adjustment or establish randomized allocation. Authors disclosed employment and financial interests in Aledade. Do not apply 29.4% as a universal factor or subtract costs already included in a total-cost result.

Truncation means CMS caps extreme annualized spending in its ACO calculations. The cap can limit the financial value of preventing dollars that were above it.

Show me the simple $150,000 → $35,000 story
Now make the mechanism visible · Hypothetical, not the published study

They budgeted around $150,000 of medical savings. Only $35,000 survived the cap.

Imagine one beneficiary would cost $400,000 without the program, but $250,000 with it. Use an illustrative $285,000 CMS spending cap.

What the medical bills suggest$150,000$400,000 − $250,000Medical spending avoided
What the capped comparison recognizes$35,000$285,000 − $250,000Spending reduction after truncation

The missing $115,000 was above the cap. It can be real medical spending avoided, but it was not fully counted in this ACO spending comparison to begin with.

Same full-year beneficiary, one enrollment type, nonnegative included expenditures, common cap. The $285,000 cap is invented for teaching. The $35,000 is still not the shared-savings check.

Try the same calculation with your numbers.

Enter total included A/B spending for one full-year beneficiary, not a whole population and not just hospital costs.

One more step: illustrate a sharing rate, not a final CMS payment
A cap-adjusted spending reduction enters the whole ACO financial calculation. A rate alone cannot reproduce settlement.
Calculation receipt and assumptions
Three different numbers: medical spending avoided → spending reduction after truncation → any shared-savings payment. They are not interchangeable. Final payment depends on the full ACO reconciliation, including benchmark, quality, minimum savings/loss rate, caps and other adjustments.

For partial years, changing enrollment type, completion adjustments, negative expenditures or a patient cohort, use the full CMS method. Apply truncation at the beneficiary/enrollment-type level before aggregation, not once to an ACO total.

5Program support after the FFS revenue credit

6Net change across the included spending categories

Positive = spending avoidedNegative = spending addedΔ means change

Use total dollars for the same cohort and period. Category totals must not overlap.

7Where the inputs usually live

Need this input?Start hereDo not assume
Assigned people / eligible timeACO-MS assignment package; current-year report instructions; program roster for your subset.Headcount equals person-years, or everyone with CCLF claims is assigned.
Actual expenditure and utilizationQuarterly expenditure/utilization report, often QEXPU; final reconciliation for finalized values; claims extract for cohort detail.Quarterly estimates are settlement, or a whole-ACO rate describes your selected group.
Thresholds and methodApplicable report parameters and current-year methodology. PY2026 v14 §3.1.3 describes truncation.A hypothetical cap or a different enrollment type’s cap is current and applicable.
Engagement / program fee / FFS collectionsProgram roster, contract, operating budget and RCM collections ledger.CMS supplies these program-specific assumptions.
Expected intervention effectA relevant controlled evaluation or a clearly marked scenario assumption.A pre/post decline proves causality.

Report structures change by performance year. No universal CMS worksheet cell is asserted here; an exact year-specific cell has not been independently verified. Source directions are intentionally general where an exact locator is not established. Older CMS tie-out examples are background, not current calculation instructions.