America’s preventive services work. Completion is what fails. The Prevention Adoption Initiative is a proposed toolkit to close that gap in safety-net settings — beginning with a colorectal-cancer screening pilot. Everything here is illustrative and future-tense; nothing has been deployed.
Proven preventive services exist, but completion fails — colorectal screening sits below target, worst in the safety net. The Prevention Adoption Initiative is a disease-agnostic engine that finds who’s overdue, removes friction, and drives completion — with an explainable, fairness-audited dashboard. I’m looking for a clinic to pilot it and partners to scale it.
An installable toolkit — registry + friction removal + reminders + escalation — that makes recommended preventive care actually get finished. CRC first; disease-agnostic by design.
The failure is adoption, not science
Every lever is already recommended and cost-effective; what’s missing is reliable execution in under-resourced settings — a measurable, economic problem.
A pre-registered single-site pilot with internal controls, an explainable AI dashboard, and a named sustainability path — then replicate across a network.
Only about 8% of U.S. adults 35+ received all 15 high-priority clinical preventive services measured (Borsky et al., Health Affairs, 2018) — directional, but stark.
U.S. colorectal-cancer screening sits at 63.5% (2023) against the Healthy People 2030 target of 72.8% — and has barely moved off baseline.
Federally Qualified Health Centers screened just 44.1% (2018 UDS) — the gap concentrates exactly where the pilot proposes to work.
SOURCES — Borsky et al., Health Affairs 2018; Healthy People 2030 objective C-07 (2023); NCCRT / HRSA UDS 2018. Figures reconciled to current objectives; the older 68.3%/74.4% pair referred to the prior 50–75 age objective.
| Lever | Median increase in CRC screening | Pillar |
|---|---|---|
| Reducing structural barriers (mailed at-home FIT) | +16.1 pp | Service redesign |
| Provider reminder / recall systems | +15.3 pp | Data infrastructure |
| Patient navigation | +13.6 pp | Implementation |
| Client (patient) reminders | +10.9–11.5 pp | Data infrastructure |
SOURCES — CDC Community Guide (CRC screening findings); Coronado et al., STOP CRC cluster-RCT (JAMA Internal Medicine, 2018).
Make the completed path the easiest one — defaults, mailed at-home FIT, one-click scheduling, barrier removal.
Registries and dashboards that surface who is overdue and where patients drop out; FHIR integration; automated reminders; CHW escalation.
Training so a site can run and sustain the toolkit itself — the pillar the evidence says matters most.
The pilot would target a clinic below the UDS colorectal-cancer benchmark, with a FHIR-capable EHR, an existing outreach function, and a quality-improvement sponsor. Evaluation would follow the RE-AIM framework against the site’s own baseline.
Raise a UDS/HEDIS measure inside your existing staff, with the fidelity discipline that makes outreach actually land — low added burden, your clinical team in the lead.
A disease-agnostic engine at roughly $90 per additional person screened, with an equity-stratified design and a path from one site to a network.
A pre-registered stepped-wedge / RE-AIM evaluation with a protocol-first publication — rigor built in from the start, co-authorship on the table.
Designed so reminders and navigation don’t widen gaps for patients with unstable contact or low portal access — benefit first, no data exploitation.
The registry-plus-outreach loop is deliberately disease-agnostic — the same mechanism applies to cervical and breast screening, hypertension, diabetes follow-up, and immunizations. It is also system-portable. National screening systems abroad face the identical problem: Bangladesh, for example, runs a community-clinic cervical and breast screening program with a national registry, yet reports low uptake and loss to follow-up — a completion-and-fidelity gap, not a science gap.
As the initiative matures, the engine is intended to extend to those settings — beginning where the founder already has standing and screening-service experience — so a method proven in one U.S. health center can inform preventive-care completion far more broadly. This global extension is a future direction, in development; the current work is the U.S. demonstration above.
For a private network (a natural first partner such as the founder’s former employer), completion is revenue: more screenings mean more diagnostics, more downstream care, higher patient retention, and stronger corporate-wellness and payer contracts. BDOT lifts the completion rate the business is already trying to grow — solving the health problem and the P&L at once.
Who: the DGHS Non-Communicable Disease Control (NCDC) programme, the national cervical & breast screening leadership, and the Community Clinic Health Support Trust. How: embed the completion loop into the existing ~14,000 community clinics and the national electronic registry to close the uptake and follow-up gaps the programme’s own evaluation reports — piloted in a few upazilas, aligned to the national strategy.
BRAC and Marie Stopes for community reach; icddr,b, BSMMU, and NICRH for evaluation and clinical authority. These partners turn a pilot into publishable, scalable evidence — and open doors to both government and donor funding.
Recruit a Wayne/Washtenaw FQHC via the Michigan Primary Care Association and EMU faculty; lock outcomes and follow-up in the agreement.
Run the loop; evaluate with RE-AIM against baseline (and a comparison where feasible).
Package the toolkit, seek peer review, replicate at a second site.
Distribute via PCA / HCCN / NTTAP networks and professional bodies (NACCHO, ASTHO, APHA, SOPHE).
This is an HSREP initiative in development. All designs are illustrative, all language is future-tense, and no patient data has been collected. Operational materials will be reviewed by counsel before any launch.
I’m seeking the institutional home to lead this pilot — a partnership, a role, or a funded engagement with a health center, a payer, or a research group that owns this problem.
