The calculator

What a year of preterm births costs in your state, how much of it a program could reach, and what the missing weeks cost.

Three dials: how many pregnancies are in care by the end of the first trimester, how many of those a program treats, and what the program achieves among the ones it treats. The arithmetic is shown underneath. There are no demographic inputs, on purpose.

Where

Births and preterm-birth rate from CDC/NCHS Births: Final Data for 2023 (Tables 6 and XV); Medicaid share from KFF State Health Facts, 2024 validated. Edit the births to model a plan or a county.

Births per year

The cost, as it stands

Full lifetime cost of one year of preterm births here—

Preterm birth rate—

Preterm births / year—

Births financed by Medicaid (2024)—

$64,815 average lifetime cost per preterm birth, 2016 dollars, blended across all severities (Waitzman et al. 2021) validated. Not annual, not inflated.

The three dials

Pregnancies in care by the end of the first trimester 76%

Default: 76.1% of U.S. mothers began prenatal care in the first trimester in 2023 (NCHS) validated · national. Your state's figure is in the NVSR tables; set it here. This is the one entry figure the national record gives. Some windows stay open past the first trimester, so this understates the share in time for those, and overstates what the missing weeks cost. The timeline shows what closes when.

Share of those pregnancies the program screens or treats 55%

Program reach among pregnancies that arrived in time. Research-consent rates in the trials ran 55–73% plausible; real-world clinical uptake is not yet published. The default sits at the bottom of that range.

Preterm births prevented among those treated 5%

A relative reduction, across all the pregnancies the program touches. Population-level effects are modest because each intervention works in a subgroup: vaginal progesterone cuts preterm birth before 34 weeks by ~22% in short-cervix pregnancies (EPPPIC) validated. The sources page lists what each intervention has shown, in whom, and at what grade. The default here is set from none of them; it is a dial. Start low and defend what you set.

Program cost per pregnancy screened or treated $60

What the program spends per pregnancy it screens or treats: screening, drug, counseling, navigation. For a screen-and-treat program, this is the screen's cost per pregnancy plus the total cost of treatment divided by the number of pregnancies screened. A proteomic screen is the costliest item on the chart; aspirin and generic progesterone cost pennies to dollars.

What the program does

Pregnancies reached in time

—

Preterm births prevented / year

—

Lifetime cost avoided / year

—

Program cost / year

—

Net value / year

—

Value returned per $1 spent

—

Every result here is only as good as its weakest input, which is the effect dial. Whatever you set it to, the answer is no better than your reason for setting it.

What the missing weeks cost

Lifetime cost the same program would not reach among pregnancies that arrived after the first trimester—

Pregnancies arriving after the first trimester—

Preterm births among them the program would have prevented—

This is the same program, at the same uptake and the same effect, applied to the pregnancies that arrived after the first trimester. It's what earlier entry would be worth if those pregnancies responded to the program as the others do. That's an assumption, not a finding: pregnancies that enter care late may carry different risks, and this page doesn't model them.

How this is computed

Every term is either the public record (births, rate, cost per preterm birth) or a dial you set. There is no hidden model: no risk stratification, no demographic weighting, no cost-by-week schedule. Those exist, and they change the answer, but they belong in a payer's actuarial review, not on a public page.

Corrected 30 October 2026: program cost now counts only the pregnancies the program screens or treats. It had counted every pregnancy in care by the end of the first trimester, which overstated the cost whenever the program reached fewer than all of them.

Reading the result honestly

The lifetime cost is a civic-scale number, not a payer's. Medicaid would bank the NICU and first-year medical savings; the rest would land in school districts, employers, and the child's own earnings over decades. A payer signing a contract uses the acute cost by gestational-age band, which is lower and sharper. The order of magnitude would likely survive either way: medical care is about two-thirds of the lifetime figure (Waitzman et al., 2021) validated.
The effect dial does the work, so defend it. At 5% it describes a program that prevents one preterm birth in twenty among the pregnancies it treats. That is ambitious for a whole population and modest for a well-targeted subgroup. The sources page lists what each intervention has shown, in whom, and at what grade.
Entry timing is already measured. From January 2027, every visit after it is coded on its own, with its own date. The missing-weeks figure is the cost of care starting after the first trimester instead of during it, before any screening or drug is paid for.