GiveBetter x SF

Changent / Nurse-Family Partnership

National implementation support for Nurse-Family Partnership and Child First

Research time: ~10 min on GPT-6 Astra Lite + GPT-5.6 Sol
  • Source audit — checked claims, assumptions and calculations.
  • Modeling — estimated costs, QALYs and uncertainty.

Published: 10 September 2026.

Donate

Summary

What do they do? Changent is the national umbrella for two material programs. NFP partners employ nurses while Changent supplies education, infrastructure, monitoring, support, and network development from pregnancy through age two. Child First supports affiliate teams consisting of a licensed mental-health clinician and care coordinator for trauma-affected families with children birth through five, typically for 6–12 months. More

Why this approach interests us

  • NFP's Memphis trial reported fewer preventable-cause deaths through age 20. Child First's original 157-family trial reported improvements in child and caregiver outcomes. Changent operates both national replication systems at meaningful scale.

Our main reservations

  • A 5,670-person modern NFP trial found a null age-two primary composite, and a newer 224-family multisite Child First trial found null primary family outcomes. Both utility bridges and ordinary-gift additionality are analyst priors; no priced next-gift expansion or Bay funding gap was found.

What do you get for your dollar?

A $100,000 two-program lifetime model yields 0.09766 signed QALY, or $10.24 million donor and $15.49 million modeled gross resources per 10 QALYs. It credits all baseline-discounted survival through age 20 plus 50% of the actuarial post-age-20 tail; this persistence/catch-up factor is an explicit prior. The conservative follow-up-bound diagnostic is $14.03 million/$21.23 million. The lifetime estimate remains 84.46% favorable-tail-dependent; without that case, donor cost is $62.60 million. These are conditional priors, not a verified marginal offer. Inspect the model, assumptions and scenarios.

lifetime modeled gift
$10.24M — donor per 10 signed modeled QALYs; modeled gross-resource estimate $15.49M
follow-up-bound diagnostic
$14.03M — donor per 10 modeled QALYs; $21.23M modeled gross, no survival credit after age 20
favorable-tail dependence
84.46% — without the favorable case, lifetime donor cost is $62.60M per 10 QALYs
bay / sf impact share
2.370% / 0.1896% — primary lifetime model; NFP-only local priors and no verified local Child First pathway
funding room
Constrained / unverified — large total-net-asset scale, unknown liquidity, and no priced next-gift plan

1. What do they do?

Changent is the national umbrella for two material programs. NFP partners employ nurses while Changent supplies education, infrastructure, monitoring, support, and network development from pregnancy through age two. Child First supports affiliate teams consisting of a licensed mental-health clinician and care coordinator for trauma-affected families with children birth through five, typically for 6–12 months.

Enroll during pregnancy

Eligible first-time mothers are connected to a trained nurse through a local implementing partner; Changent is the national support organization rather than the direct SF employer.

Visit through age two

Nurses provide a roughly 2.5-year pregnancy-to-age-two course focused on maternal health, child health and development, and family economic self-sufficiency.

Deliver Child First

Affiliate clinician/coordinator teams provide intensive home-based psychotherapy and care coordination, typically for 6–12 months; 2024 reporting lists 2,564 families and 60,971 contacts.

Support both networks

Changent trains and supports local partners, monitors implementation, and operates shared infrastructure. An ordinary gift is not represented as restricted to one program or geography.

Scope of this review. The full $30,712,013 FY2024 expense is allocated across the two named programs in proportion to Form 990 direct program expense: $22.656M NFP and $8.056M Child First. One organization-scale fraction applies to 57,005 reported NFP active families, converted to 22,802 annual full-course equivalents using the 2.5-year span, and 2,564 reported Child First families, treated as approximate annual course equivalents. These are unit-alignment approximations, not starts/completions or marginal prices; household overlap across programs was not measured.

2. Monitoring and information sharing

Memphis pregnancy-plus-infancy nurse-visited and comparison families; 690 surviving children in the mortality analysis

Randomized trial, two-decade follow-up. Nine preventable-cause deaths occurred in the comparison group and none in the nurse-visited group; estimated preventable mortality was 1.6% versus 0 through age 20 (P=.04). All-cause mortality was 2.7% versus 0.9% (P=.11).

Our assessment. Mortality was not prespecified and events were sparse. The model uses a finite 0.144-QALY/full-course calibration only as a ceiling and credits −0.010 to 0.080 QALY/course across scenarios.

5,670 Medicaid-eligible pregnant people across nine implementing authorities; 4,932 live births in child analysis

Large modern randomized controlled trial. The age-two primary composite was 27.3% with NFP versus 26.8% with usual care, adjusted +0.4 percentage points (95% CI −2.3 to 3.0). ED use fell 2.9 points but lost significance after multiplicity adjustment.

Our assessment. Strong counterevidence for the contemporary age-two primary composite, but it does not resolve later outcomes. The model gives null plus harm 50% weight rather than pooling incompatible endpoints.

NFP implementation across scale-up sites

National replication cost analysis. Estimated $8,580 present-value full cost per family in 2012 prices.

Our assessment. Historical full-delivery proxy, not a current SF marginal quote; used only for the favorable gross-resource case.

California NFP implementation

California NFP impact and cost profile. Reported $14,728 cost per family in 2019.

Our assessment. Historical California proxy used in harm/null/cautious gross-resource cases; does not establish current local resources or marginal capacity.

157 multirisk urban families with children ages 6–36 months

One-site randomized trial. Reported improvements in child language/externalizing symptoms, caregiver psychopathology and service access at 12 months, with less child-protective-services involvement at three years.

Our assessment. Small developer-associated trial; no QALYs. Supports a positive tail but not direct utility conversion.

224 randomized families, 182 followed across 10 sites in two states

Multisite pandemic-era randomized trial. No significant effects on caregiver psychological well-being, child behavior, child welfare or most economic outcomes.

Our assessment. Supports high null weight and low causal-transfer priors; pandemic hybrid delivery may limit transfer to normal implementation.

National 2024 NFP and Child First activity

Changent organization reporting. 57,005 NFP families/586,081 visits and 2,564 Child First families/60,971 contacts.

Our assessment. No counterfactual and no observed start/completion cohort. NFP active-family units are divided by 2.5; Child First annual families approximate 6–12-month course equivalents.

Whole Changent legal organization, FY2024

Official IRS Form 990 accounting data. Revenue $42.540M, expense $30.712M, surplus $11.828M and total net assets $111.212M; Part III reports $15.971M NFP and $5.679M Child First direct program expense.

Our assessment. Full expense remains in the denominator and support costs are allocated proportionally. Net assets are not proven liquid/unrestricted and do not establish a marginal gift opportunity.

3. Qualitative assessment

NFP's Memphis trial reported fewer preventable-cause deaths through age 20. Child First's original 157-family trial reported improvements in child and caregiver outcomes. Changent operates both national replication systems at meaningful scale.

Key reservations

  • Child First's utility bridge is analyst-created because neither randomized trial reports QALYs; the positive original trial and newer multisite null trial do not yield a stable pooled health effect.
  • Reported families served are annual active units, not starts, exposure-years, retention, or completed courses; dividing by 2.5 is an analyst unit-alignment approximation.
  • The Memphis preventable-mortality result was sparse, not prespecified, and sensitive to small event-count changes; all-cause mortality was not significant.
  • The modern South Carolina age-two primary composite was null, and the ED finding did not survive multiplicity adjustment.
  • The 9-QALY mortality bridge is an undiscounted analyst cap inside the age-20 window, not a trial QALY estimate.
  • The primary 12.523-QALY mortality judgment fully credits baseline-discounted survival through age 20 and 50% of the actuarial post-20 tail; the persistence/catch-up factor is unobserved.
  • Historical $8,580 and $14,728 delivery costs are not current San Francisco marginal prices.
  • Total net assets are not demonstrated liquid or unrestricted; restrictions, pledge timing, and liquidity are unavailable.
  • Bay 2.5% and SF 0.2% are unsourced round allocation priors built from mismatched capacity and served units, not observed local spending or reach.
  • Child First uniqueness factors reduce but cannot eliminate unmeasured cross-program household or utility overlap.
  • Gross-resource subtraction assumes historical delivery-cost proxies contain allocated Changent internal costs; this is not source-verified.
  • The weighted result is favorable-tail-sensitive, and marginal health-producing funding room is unpriced and unverified.

Benefits not included in our estimate

  • Maternal employment and family income
  • Avoided Medicaid, emergency, child-welfare, education, and justice-system spending
  • Sibling, caregiver, and intergenerational spillovers beyond modeled child QALYs
  • Non-mortality maternal health, child development, school-readiness, and parenting benefits not captured by the per-course QALY prior
  • National policy, training, research, and implementation-learning spillovers

4. What do you get for your dollar?

Explicit lifetime portfolio best guess: $10.24M donor per 10 QALYs; follow-up-bound diagnostic: $14.03M.

One ordinary-gift organization fraction scales both programs and the gift is counted once. Child First receives a separate finite utility/causal-transfer bridge. The lifetime headline scales positive NFP credits using an actuarial age-10 horizon discounted to pregnancy/birth, fully credits the observed window, and gives 50% weight to post-age-20 persistence. The accepted-v3 9-QALY bound remains a separate diagnostic. No Child First Bay/SF pathway was verified.

A better life is our comparison unit of 10 additional quality-adjusted life years (QALYs), potentially spread across people. These are uncertain estimates, not measured returns or verified donation offers.

How we calculate the estimate

WEIGHTED LIFETIME PORTFOLIO: 10 × $100,000 ÷ weighted[(NFP realized courses × horizon-adjusted NFP QALY/course) + (Child First realized unique families × finite utility bridge × causal transfer)]
$10,239,969 per 10 modeled QALYs; 0.0976565 signed expected QALY, including 0.0050653 from Child First

Model inputs and assumptions
Whole-organization expense and program allocation
$30.712M total; $22.656M NFP / $8.056M Child First allocated (range: Child First is 18.49% of whole expense and 26.23% of direct program expense). Form 990 reports $15.971M direct NFP and $5.679M direct Child First expense. Remaining expense is allocated proportionally so the two programs reconstruct the whole denominator. High accounting; low causal allocation.
Modeled full-course duration
2.5 years (range: fixed). Official NFP care runs from pregnancy until the child turns two. Dividing annual active-family units by 2.5 aligns benefit and cost units but is not measured retention or completion. High program design; low retention/completion observation.
Scenario prior weights
10% harm / 40% null / 30% cautious / 15% central / 5% favorable (range: fixed illustration). The modern primary-composite result was null, the legacy mortality signal is sparse and non-prespecified, and marginal gift leverage is unverified; null plus harm therefore receive 50% weight. Subjective; not empirically calibrated.
Ordinary-gift funding additionality
10% central (range: 2% / 3% / 5% / 10% / 20%). Share of the gift assumed to cause support for an otherwise-unserved family rather than reserves, fee/grant substitution, or planned activity. Changent publishes no priced expansion offer or bottleneck. Very low.
Course realization
90% central (range: 80%–95%). Discounts modeled course-equivalents for incomplete exposure, retention, fidelity, and implementation failure; it is not a measured completion rate. Low.
NFP QALY per realized full course
0.013915 central lifetime (range: −0.010 / 0 / 0.001391 / 0.013915 / 0.111318). Positive follow-up-bound priors (0.001/0.010/0.080) are multiplied by 12.523/9 for the primary lifetime estimate; harm/null are unchanged. Both the original priors and the lifetime persistence factor are judgments informed by sparse, non-prespecified mortality evidence. Very low transfer prior.
Child First QALY per realized unique family
0.010 central (range: −0.005 / 0 / 0.001 / 0.010 / 0.050). Original 157-family RCT was positive; newer 224-family multisite pandemic RCT was null. Values are finite analyst utility bridges after explicit 0%–62.5% causal-transfer factors, not trial QALYs. Very low explicit utility bridge.
NFP mortality horizon
12.523 QALYs/death lifetime judgment (range: 5.713 at 0% to 19.334 at 100% post-age-20 persistence; 9-QALY undiscounted follow-up diagnostic). Uses 68.025 remaining years at age 10, 0.9 utility and 3% discounting to pregnancy/birth baseline. It fully credits 5.713 QALYs through age 20 and credits 50% of the 13.621-QALY post-20 tail to allow mortality catch-up or delay. Very low lifetime extrapolation.
Full delivery resources
$11,700 NFP / $10,834 Child First central (range: NFP $8,580–$14,728; Child First $8,000–$14,643). Resources attach to pre-realization program equivalents. Proportionally allocated Changent support cash is subtracted once, then the donor gift is counted once. Inputs are historical implementation proxies, not current marginal prices. Historical and location-specific.
Bay and San Francisco impact shares
2.3703% / 0.1896% primary lifetime portfolio (range: NFP priors 2.5% / 0.2%; Child First 0%; follow-up diagnostic 2.322% / 0.1858%). Bay means nine counties and SF is nested. No primary evidence verified a local Child First affiliate; NFP shares remain unsourced proportional-allocation priors, not observed reach. Very low analyst allocation.

Illustrative $100,000 ordinary gift to Changent

  • Harm (10% prior weight): −0.012547 modeled QALY. NFP −0.011879 plus Child First −0.000668; no positive donor or gross ratio.
  • Null (40%): No modeled QALYs. Both program pathways have zero QALY benefit; donor and gross ratios are not defined.
  • Cautious positive (30%) — lifetime: $212.32M modeled donor cost per 10 QALYs. 0.004710 portfolio QALY; modeled gross-resource cost $330.76M.
  • Central (15%) — lifetime: $9.99M modeled donor cost per 10 QALYs. 0.100117 portfolio QALY; modeled gross-resource cost $18.57M.
  • Favorable stress (5%) — lifetime: $606,202 modeled donor cost per 10 QALYs. 1.649615 portfolio QALYs; modeled gross-resource cost $1.338M.
  • Signed follow-up-bound diagnostic: $14.03M modeled donor cost per 10 QALYs. 0.071273 expected QALY and $21.23M modeled gross-resource cost. It assigns no survival value after age 20.
  • Primary lifetime judgment: $10.24M modeled donor cost per 10 QALYs. 0.097657 expected QALY and $15.49M modeled gross-resource cost. It uses 12.523 baseline-PV QALYs/death, 50% post-age-20 persistence, 84.46% favorable-tail share and $62.60M donor cost without favorable.

Uncertainty. Funding additionality, realization, both QALY bridges, Child First uniqueness, mortality horizon, geography, delivery-cost transfer and weights are analyst priors. NFP's modern age-two composite and Child First's newer multisite primary outcomes were null; the age-20 mortality signal was sparse and non-prespecified. Zero funding additionality returns zero QALYs and null donor, gross, Bay and SF ratios.

5. Funding and previous grants

Obtain unrestricted-gift allocation across NFP, Child First, policy, and reserves; SF unique active, new, and completing families; unserved eligible referrals; marginal national-office and local delivery costs; cofunding; fidelity; and a preregistered local outcome plan.

Health-producing marginal RFMF is unverified and appears constrained. Changent accepts ordinary gifts to EIN 20-0234163, but does not promise NFP- or SF-restriction or publish a next-$100,000 family target, local cofunding gap, waitlist, or binding national-office constraint. FY2024 total net assets of $111.212M were 3.6 times expense, but are not demonstrated liquid or unrestricted; pledge restrictions, timing, and liquidity are unavailable.

This review does not establish a verified marginal funding offer or a complete history of grants.

Donate

Opens the organization’s giving page. A general donation may not fund the specific activity modeled here; confirm allocation with the recipient.

6. Sources

  1. 2024 Year in Review. Changent. Official organization output report. Published: 2024 reporting year; retrieved: 2026-09-10.
  2. FY2024 public-inspection Form 990. Changent. Official IRS filing and program-expense split. Published: Tax year ending 2024-09-30; retrieved: 2026-09-10.
  3. Nurse-Family Partnership. Changent. Official program model and duration. Published: Undated/current; retrieved: 2026-09-10.
  4. Child First. Changent. Official program model, staffing and duration. Published: Undated/current; retrieved: 2026-09-10.
  5. Support Our Work. Changent. Official donation route; Changent EIN 20-0234163 verified. Published: Undated/current; retrieved: 2026-09-10.
  6. 2025 Year in Review. Changent. Official organization reporting and philanthropic-commitment accounting boundary. Published: 2025 reporting year; retrieved: 2026-09-10.
  7. Effect of Prenatal and Infancy Nurse Home Visitation on Mortality of Mothers and Children. JAMA Pediatrics. Peer-reviewed randomized-trial follow-up. Published: 2014; retrieved: 2026-09-10.
  8. Nurse-Family Partnership and a Primary Composite of Adverse Outcomes at Age 2 Years. JAMA Pediatrics. Peer-reviewed modern randomized controlled trial. Published: 2025; retrieved: 2026-09-10.
  9. A randomized controlled trial of Child FIRST. Child Development. Peer-reviewed original randomized controlled trial. Published: 2011; retrieved: 2026-09-10.
  10. Child First home visiting impacts during the COVID-19 pandemic. Journal of Family Psychology. Peer-reviewed multisite randomized controlled trial. Published: 2026; online 2025; retrieved: 2026-09-10.
  11. Child First. ACF Prevention Services Clearinghouse. Federal evidence clearinghouse. Published: Current evidence record; retrieved: 2026-09-10.
  12. Child First. Blueprints for Healthy Youth Development. Evidence and historical implementation-cost profile. Published: Current profile; retrieved: 2026-09-10.
  13. Costs of Nurse-Family Partnership Scale-Up. Prevention Science. Peer-reviewed national replication cost study. Published: 2015; retrieved: 2026-09-10.
  14. NFP Multidomain Impact in California. Nurse-Family Partnership. Official California impact and historical cost profile. Published: 2019; retrieved: 2026-09-10.
  15. California 2024 State Profile. Nurse-Family Partnership. Official California capacity and partner profile. Published: Late-2023 data in 2024 profile; retrieved: 2026-09-10.
  16. Changent filing record, EIN 20-0234163. Internal Revenue Service, mirrored by ProPublica Nonprofit Explorer. IRS Form 990 filing history. Published: FY2024 filing received 2025-07-18; retrieved: 2026-09-10.
  17. 2022 period life table. U.S. Social Security Administration. Actuarial remaining-life sensitivity anchor. Published: 2025 supplement; 2022 mortality; retrieved: 2026-09-10.

Annual expenses: years and sources

Average annual expenses (three consecutive fiscal years): Not available. Organization size is separate from the modeled cost-effectiveness of a donation.

Changent / Nurse-Family Partnership

Three years of organization-level expenses have not yet been verified for this report. No other organization or fiscal sponsor budget has been substituted.