For the CMS Rural Health Transformation Program

The curated data layer for rural health

A $50B program. A 5-year window. One missing layer.

Predoc ingests rural health medical records from HIE networks, EHRs, pharmacy data, imaging networks, direct provider outreach, fax, and non-connected facilities. It returns a standardized, longitudinal record that solves the rural health interoperability gap downstream analytics, AI, and care models depend on.

Get In Touch

HIE Networks

Pharmacy Records

EHR Connections

Imaging Networks

Direct Provider Requests

Fax-Based Providers

Non-Connected Facilities

HIE Networks

Pharmacy Records

EHR Connections

Imaging Networks

Direct Provider Requests

Fax-Based Providers

Non-Connected Facilities

HIE Networks

Pharmacy Records

EHR Connections

Imaging Networks

Direct Provider Requests

Fax-Based Providers

Non-Connected Facilities

HIE Networks

Pharmacy Records

EHR Connections

Imaging Networks

Direct Provider Requests

Fax-Based Providers

Non-Connected Facilities

INGESTING SOURCESCURATED DATA LAYERSYNCING…

Patient cohort: rural health

Lab Results

Medications

Problems

Provider Facility Notes

Pathology

The rural data problem

Rural health data is fragmented by design.

Three structural problems undermine every downstream initiative, from RHT planning to AI-assisted care.

Fragmented rural data

Rural patients move across HIE-connected and non-connected facilities, solo practitioners, and regional hospitals. No single network sees the full record.

Fax-driven workflows

Records still arrive by fax, paper, and direct provider outreach. The data exists. It's just not in a system anyone downstream can consume.

Downstream AI starvation

Population analysis, risk adjustment, and AI-assisted care depend on a curated longitudinal record. Without one, every model underperforms.

The curated data layer

Where rural health data becomes infrastructure.

Seven source types (including the fax-based, direct-outreach, and non-connected facility flows that define rural ingestion) converge into one queryable longitudinal record.

01Clinical Data Sources

Multi-format, unstructured

02AI / ML Parsing Engine

Extraction, normalization, curation

03Curated Data Layer

Structured, aggregated, indexed

Digital sources

HIE Networks

Pharmacy Records

EHR Connections

Imaging Networks

Direct outreach (PDFs)

Direct Provider Requests

Fax-Based Providers

Non-Connected Facilities

Extraction

OCR and clinical parsing across PDFs, imaging, exchange feeds, and handwritten notes.

Normalization

Codes, units, and concepts mapped to standard medical vocabularies.

Curation

Resolved, deduplicated, clinically organized into the Predoc patient model.

Lab Results

lab-results.json

{

"loinc": "718-7"

}

12345678

{

"loinc": "718-7",

"name": "Hemoglobin",

"value": "13.2 g/dL",

"ref_range": "12.0-15.5",

"effective": "2025-08-14",

"source": "EHR / Epic"

}

Medications

medications.json

{

"rxnorm": "861009"

}

12345678

{

"rxnorm": "861009",

"name": "Lisinopril 10 MG",

"status": "active",

"started": "2024-11-02",

"dose": "10 mg PO daily",

"source": "Pharmacy / NCPDP"

}

Problems

problems.json

{

"icd10": "I10"

}

12345678

{

"icd10": "I10",

"snomed": "59621000",

"name": "Essential hypertension",

"status": "active",

"onset": "2023-04",

"source": "EHR Problem List"

}

Provider Facility Notes

provider-facility-notes.json

{

"type": "progress_note"

}

12345678

{

"type": "progress_note",

"author": "Dr. M. Chen",

"facility": "Rural Health Clinic",

"date": "2025-09-12",

"pages": 4,

"source": "Direct provider outreach"

}

Pathology

pathology.json

{

"snomed": "M-87200"

}

12345678

{

"specimen": "Skin, left forearm",

"diagnosis": "Benign nevus",

"snomed": "M-87200",

"date": "2025-06-30",

"pathologist": "Dr. R. Patel",

"source": "Non-connected facility"

}

Imaging Notes

imaging-notes.json

{

"cpt": "71250"

}

12345678

{

"modality": "CT",

"study": "Chest w/o contrast",

"cpt": "71250",

"impression": "No acute findings",

"date": "2025-07-21",

"source": "Imaging Network"

}

Procedures

procedures.json

{

"cpt": "45378"

}

12345678

{

"cpt": "45378",

"snomed": "73761001",

"name": "Colonoscopy, diagnostic",

"date": "2024-12-08",

"facility": "Regional Hospital",

"source": "Fax-Based Providers"

}

Family History

family-history.json

{

"icd10": "E11"

}

12345678

{

"relation": "mother",

"condition": "Type 2 Diabetes",

"icd10": "E11",

"age_onset": 52,

"verified": "patient-reported",

"source": "Direct provider outreach"

}

01Clinical Data Sources

Multi-format, unstructured

Digital sources

HIE Networks

Pharmacy Records

EHR Connections

Imaging Networks

Direct outreach (PDFs)

Direct Provider Requests

Fax-Based Providers

Non-Connected Facilities

02AI / ML Parsing Engine

Extraction, normalization, curation

Extraction

OCR and clinical parsing across PDFs, imaging, exchange feeds, and handwritten notes.

Normalization

Codes, units, and concepts mapped to standard medical vocabularies.

Curation

Resolved, deduplicated, clinically organized into the Predoc patient model.

03Curated Data Layer

Structured, aggregated, indexed

Lab Results

lab-results.json

{

"loinc": "718-7"

}

12345678

{

"loinc": "718-7",

"name": "Hemoglobin",

"value": "13.2 g/dL",

"ref_range": "12.0-15.5",

"effective": "2025-08-14",

"source": "EHR / Epic"

}

Medications

medications.json

{

"rxnorm": "861009"

}

12345678

{

"rxnorm": "861009",

"name": "Lisinopril 10 MG",

"status": "active",

"started": "2024-11-02",

"dose": "10 mg PO daily",

"source": "Pharmacy / NCPDP"

}

Problems

problems.json

{

"icd10": "I10"

}

12345678

{

"icd10": "I10",

"snomed": "59621000",

"name": "Essential hypertension",

"status": "active",

"onset": "2023-04",

"source": "EHR Problem List"

}

Provider Facility Notes

provider-facility-notes.json

{

"type": "progress_note"

}

12345678

{

"type": "progress_note",

"author": "Dr. M. Chen",

"facility": "Rural Health Clinic",

"date": "2025-09-12",

"pages": 4,

"source": "Direct provider outreach"

}

Pathology

pathology.json

{

"snomed": "M-87200"

}

12345678

{

"specimen": "Skin, left forearm",

"diagnosis": "Benign nevus",

"snomed": "M-87200",

"date": "2025-06-30",

"pathologist": "Dr. R. Patel",

"source": "Non-connected facility"

}

Imaging Notes

imaging-notes.json

{

"cpt": "71250"

}

12345678

{

"modality": "CT",

"study": "Chest w/o contrast",

"cpt": "71250",

"impression": "No acute findings",

"date": "2025-07-21",

"source": "Imaging Network"

}

Procedures

procedures.json

{

"cpt": "45378"

}

12345678

{

"cpt": "45378",

"snomed": "73761001",

"name": "Colonoscopy, diagnostic",

"date": "2024-12-08",

"facility": "Regional Hospital",

"source": "Fax-Based Providers"

}

Family History

family-history.json

{

"icd10": "E11"

}

12345678

{

"relation": "mother",

"condition": "Type 2 Diabetes",

"icd10": "E11",

"age_onset": 52,

"verified": "patient-reported",

"source": "Direct provider outreach"

}

Four capabilities make the layer possible

Master Patient Index

Resolve a single patient identity across HIEs, EHRs, and unconnected facilities.

Learn more

Terminology Mapping

Normalize codes and free-text concepts to standard ontologies (RxNorm, LOINC, SNOMED).

Learn more

Unstructured Data Extraction

Parse PDFs, faxes, and scanned documents into structured clinical data.

Learn more

Data Standardization

Normalize fragmented sources to FHIR, HL7, and DICOM canonical formats.

Learn more

CMS RHTP fit

Three of the five CMS pillars depend on a curated data layer.

The Rural Health Transformation Program targets five outcomes: Make Rural America Healthy Again, Sustainable Access, Workforce Development, Innovative Care, and Tech Innovation. Predoc maps directly to three of them.

primary fit

Tech Innovation

Predoc is core technology infrastructure for the program: a curated medical-record data layer underneath every AI and analytics initiative.

Predoc’s role

The RHT Program funds technology innovation, but most innovation lives at the application layer: AI clinical decision support, population analytics, predictive risk models. These tools are only as good as the data they consume.

Predoc sits below the application layer as shared infrastructure. Every funded innovation benefits from the same curated longitudinal record, so funding allocated to Predoc compounds across every downstream tool rather than each vendor rebuilding rural-health ingestion from scratch.

secondary fit

Sustainable Access

Shared data infrastructure across rural facilities reduces duplicated investment and keeps services viable as facilities consolidate or partner.

Predoc’s role

Rural facility consolidation is a one-way trend. Solo practices close, regional hospitals absorb their patients, telehealth fills gaps. Each transition fragments records further.

A shared data layer maintains patient continuity through those transitions. When a closing rural clinic transfers patients to a regional hospital, the new provider's first encounter starts with the full record, not a paper binder. Infrastructure decoupled from facility-level volatility is what makes “sustainable” sustainable.

tertiary fit

Innovative Care

Measurement, screening, and AI-assisted care models all depend on the underlying curated record. Predoc is the measurement layer dependency.

Predoc’s role

Value-based care, AI-assisted screening, and remote patient monitoring all share a hidden dependency on a complete, standardized patient record to measure against. In rural settings, that measurement substrate doesn't exist by default.

Predoc is the substrate. The curated record makes quality reporting, risk stratification, and trial recruitment possible without each program rebuilding its own ingestion pipeline. The fit is tertiary because Predoc isn't the care-model innovation itself; it's the precondition that makes care-model innovation measurable.

Predoc is listed by CMS as a Friend of the Health Tech Ecosystem, pledged to support connected, secure health data access: the same infrastructure backbone the RHT Program depends on.

How this fits depends on where you sit.

Proof at scale

The architecture transfers.

The same source mix that defines rural ingestion (Non-Connected Facilities, Fax-Based Providers) is what the architecture already handles end-to-end at a major specialty practice.

Care Provider

The Oncology Institute Saves ~$2M Annually with Predoc

~$2M

Annual savings

Faster record retrieval

0%+

Record retrieval rate

“I've never seen anything that matches the quality of Predoc. One of the cofounders comes from the medical field; they really understand what we're up against.”
— Kiran Annavarapu, Regional President, The Oncology Institute

Read the full case study

Let's talk about your RHT plan.

Predoc partners with state teams, technology integrators, and rural health systems to build the curated data layer underneath every RHT initiative.

Get in touch