Summary
What do they do? St. Anthony’s A3 initiative describes outreach and insurance reconnection for patients with poorly controlled diabetes. A January2025 report says quarterly testing increased from61% to86%; this process change is not a causal QALY estimate. Clinic financing includes public/sliding-fee arrangements. Healthy SF is not insurance. We model additional SF outreach, not all Foundation services. More
Why this approach interests us
- Outreach could lead to otherwise missed indicated treatment and fewer later complications.
Our main reservations
- More testing or insurance paperwork alone does not establish additional health; dose and case mix differ from the external trial.
What do you get for your dollar?
Our provisional best guess is $10,666,667 per better life: 10 additional QALYs. The health bridge comes from an external modeled support package, not measured A3 outcomes. A more intensive peer-supported redesign is considered separately and does not determine this ranking. Inspect the models →
- outreach best guess
- $10,666,667 — per10 incremental QALYs; very low confidence
- outreach scenarios
- $2,370,370–$266,666,667 — Same external CHW-only anchor; not confidence bounds
- separate redesign
- $96,618 — Conditional peer support, not current A3 effectiveness
- funding room
- Unverified — Existing workflow and public care are counterfactuals
1. What do they do?
St. Anthony’s A3 initiative describes outreach and insurance reconnection for patients with poorly controlled diabetes. A January2025 report says quarterly testing increased from61% to86%; this process change is not a causal QALY estimate. Clinic financing includes public/sliding-fee arrangements. Healthy SF is not insurance. We model additional SF outreach, not all Foundation services.
Define the offered cohort
Include disconnected adults offered the episode, with failed contact and noncompletion costs. Do not price only successes while importing intention-to-treat health gains.
Reconnect to indicated care
Track actual care and medication access, not just tests or enrollment forms. Clinical targets depend on individual circumstances.
Measure the difference
Compare sustained outcomes and harms with concurrent usual care, including existing A3 capacity and alternative financing.
Scope of this review. The proposed USD400 buys a constructed outreach episode, not a demonstrated replica of an18-month trial package. Continuing support needs continuing costs.
2. Monitoring and information sharing
Diabetes support versus enhanced usual care
REACH Detroit trial-based economic simulation. Table2 reports .0015 QALY for CHW-only and .0276 for CHW-plus-peer support over20years.
Our assessment. Already includes waning and3% annual discounting. Source intervention costs are2018 USD:234 usual care,820 CHW,1599 CHW+peer over18months; not local donor prices.
Safety-net adults; excluded homelessness, serious illness and impending moves
Seattle randomized community-health-worker trial. Overall A1c and health-related quality-of-life differences were not significant; high-A1c subgroup was more favorable.
Our assessment. About49% potentially eligible people could not be contacted. These selection limits matter for disconnected SF adults.
300 adults;215 followed up
Miami randomized intensive support trial. Adjusted additional A1c reduction about .51 percentage points over a year.
Our assessment. Home visits and repeated calls are more intensive than a reminder. This is not the QALY coefficient.
3. Qualitative assessment
Outreach could lead to otherwise missed indicated treatment and fewer later complications.
Key reservations
- At central health assumptions donor cost must be below USD3.75 per episode for the100K/10Q target, versus proposed400.
- The separate peer redesign changes treatment intensity and financing; its optimistic threshold result is not established local cost-effectiveness.
- Outcome association, enrollment targets, clinic visits and individual patient stories cannot identify donation-caused QALYs.
- Source rounded table increments do not reproduce printed ICERs exactly; we preserve displayed values rather than invent precision.
- No universal lower-A1c target is recommended. Extra net harms must be assessed relative to alternative clinical care.
Benefits not included in our estimate
- Future healthcare savings as donor cash
- Income and enrollment itself
- Extra lifetime multiplier
- Unmeasured caregiver benefits
4. What do you get for your dollar?
$10,666,667 per10 QALYs
Central: USD400 per offered episode and .0015 external modeled QALY gain, multiplied by .50 delivery/population transfer and .50 financing additionality. Net extra harm is initially assumed zero, not known absent. This gives .000375 donor-attributable QALY. The small positive estimate is an explicit uncertain prior, not evidence that the clinic is ineffective.
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
DOLLARS PER BETTER LIFE: 10 × donor cost ÷ {funding additionality × [external integrated QALYs × local transfer − extra net harm]}
$10,666,667 per10 QALYs
Model inputs and assumptions
- Donor cost per offered episode
- 400 (range: 200 / 400 / 1000). No second lifetime, duration or generic completion multiplier. Analyst judgment.
- External integrated20-year QALY gain
- 0.0015 (range: 0.0015 / 0.0015 / 0.0015). No second lifetime, duration or generic completion multiplier. Primary Table2 economic-model increment.
- Delivery and population transfer
- 0.5 (range: 0.75 / 0.5 / 0.1). No second lifetime, duration or generic completion multiplier. Analyst judgment.
- Genuinely additional financing
- 0.5 (range: 0.75 / 0.5 / 0.25). No second lifetime, duration or generic completion multiplier. Analyst judgment.
- Extra net harm per additional episode
- 0 (range: 0 / 0 / 0). No second lifetime, duration or generic completion multiplier. Analyst judgment.
Hypothetical USD4,000 for ten offered episodes—not a verified tranche
- Favorable outreach: $2,370,370 per10 QALYs. 0.0008438 modeled QALY per offered episode.
- Central: $10,666,667 per10 QALYs. 0.0003750 modeled QALY per offered episode.
- Pessimistic: $266,666,667 per10 QALYs. 0.00003750 modeled QALY per offered episode.
Uncertainty. No change in indicated care or complete financing replacement gives no finite positive price. Hypoglycemia and treatment burden can produce negative net health.
Peer-supported redesign: a different bet, not the same outreach program
A separate conditional design uses the source peer-support package’s .0276 QALY increment, .75 local transfer, .75 financing and a low USD150 donor contribution supported by donated capacity. Its result is about USD96.6K/10Q. It changes intervention and cost structure, so it is not the favorable bound for existing A3 and is excluded from the main ranking.
SEPARATE REDESIGN: 10 ×150 / (.0276 ×.75 ×.75)
$96,618 per10 QALYs
QALY conversion assumptions
- Donor cost per offered episode
- 150 (range: Separate design, not outreach uncertainty). The additional peer support and low donor cost must both be demonstrated. Redesign judgment.
- External integrated20-year QALY gain
- 0.0276 (range: Separate design, not outreach uncertainty). The additional peer support and low donor cost must both be demonstrated. External peer-package model.
- Delivery and population transfer
- 0.75 (range: Separate design, not outreach uncertainty). The additional peer support and low donor cost must both be demonstrated. Redesign judgment.
- Genuinely additional financing
- 0.75 (range: Separate design, not outreach uncertainty). The additional peer support and low donor cost must both be demonstrated. Redesign judgment.
- Extra net harm per additional episode
- 0 (range: Separate design, not outreach uncertainty). The additional peer support and low donor cost must both be demonstrated. Redesign judgment.
- USD400 instead of150: $257,649 per10 QALYs. Same peer-health assumptions; threshold crossing is sensitive to donor price.
Weekly peer support is more intensive than reconnection. Donated capacity, full-resource costs and actual local delivery are unverified. Do not borrow the source’s persistent-effect sensitivity or subtract its future medical savings from donor cash.
5. Funding and previous grants
Need the current marginal delivery plan, completed care, causal outcomes, full payer costs and treatment harms.
USD400 is five staff hours ×USD60 plusUSD100 access/admin allocation, including failed outreach. Not a provider quote. Visits, drugs and other payers’ resources are excluded, not free. Trial noncompletion is embedded in the package anchor; do not add an arbitrary completion haircut. Additional local recruitment losses need separate evidence.
This review does not establish a verified marginal funding offer or a complete history of grants.
We have not verified a suitable donation route for this reviewed activity. Confirm the legal recipient and intended allocation before donating.
6. Sources
- A3 diabetes reconnection. St. Anthony Foundation. Provider before/after process report. Published: 2025-01-30; retrieved: 2026-09-07.
- Ye2021 REACH Detroit economic model, Table2. Diabetes Care. Primary modeled20-year health gains. Published: 2021; retrieved: 2026-09-07.
- Randomized trial of community health worker diabetes support. CDC Preventing Chronic Disease. Primary randomized trial; important nulls. Published: 2017; retrieved: 2026-09-07.
- Miami Healthy Heart Initiative. JAMA Internal Medicine. Primary intensive-support randomized trial. Published: 2017; retrieved: 2026-09-07.
- Clinic and payment FAQs. St. Anthony Foundation. Payer and eligibility disclosure. Published: Undated; retrieved: 2026-09-07.
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.
St. Anthony Foundation
Unaudited annual-report expense distributions, including administration or support services; not yet a comparable three-year series.
Three consecutive full-year expense totals are not verified.
FY2024 covers July 1, 2023–June 30, 2024. The FY2023 report labels a fiscal year but its exact dates remain unconfirmed. The older report covers calendar 2022 and could overlap FY2023, so it is excluded rather than averaged with these two years. A compatible third year and FY2023 date confirmation are still needed.