Client Industry
A healthtech company focused on clinical trial recruitment, serving contract research organizations, academic medical centers, & others.
Epic Integrated Clinical Trial Patient Matching Solution
A clinical research technology company collaborated with AkshatMedTech Digital Health to build an Epic EHR integration solution that extracts bulk patient profiles and automatically matches them against clinical trial inclusion and exclusion criteria, reducing manual chart review and accelerating trial enrollment timelines.
Request a DemoEnhancements
- Automated bulk patient data extraction from Epic
- Rule-based matching engine aligned to clinical trial criteria
- Structured patient data normalization
- Provider-facing candidate dashboard embedded within Epic
Services Delivered
- Epic FHIR API integration for bulk data retrieval
- Clinical trial eligibility engine development
- HIPAA-compliant data pipeline and de-identification framework
- Epic listing readiness consulting & interoperability validation
The Tech Stack
- Epic FHIR R4 APIs
- Python / FastAPI
- PostgreSQL
- Apache Kafka
- AWS (EC2, S3, Lambda)
- Docker
- React.js
Project Overview:
Patient Matching DashboardRun Matching
Patients screened12,847Candidates3,291Pending review248Active trials18
Patient Match Queue
IDAge · SexDiagnosisMatchStatus
PT-482158 · MNSCLC Stage IIIB92%Eligible
PT-481763 · FBreast ca, HER2+87%Review
PT-480949 · MT2DM, CKD 381%Eligible
PT-480271 · FHeart failure44%Excluded
Matching Summary
64%
Top matched trialOncology Phase 3 · Lung
80%of trials miss enrollment timelines
Auto-matched via Epic FHIRClinical trial enrollment remains one of the biggest challenges in drug development, with nearly 80% of trials missing enrollment timelines. In many hospitals, manual review of patient charts, lab results, medications, and diagnoses inside the EHR takes place to find eligible participants. This process is slow, difficult to scale, and often causes qualified patients to be missed during recruitment.
In health systems using Epic, the challenge becomes even larger because patient information is spread across encounters, problem lists, and Flowsheets. AkshatMedTech Digital Health developed an Epic-integrated clinical trial matching platform that uses Epic FHIR APIs to pull and organize structured patient data against trial eligibility criteria. The platform automatically identifies matching candidates inside Epic, helping research teams speed up recruitment and reduce the manual effort involved in patient screening.
Challenges Faced
01
Research coordinators manually reviewed individual patient charts within Epic to assess trial eligibility
Solutions We Delivered
Bulk Patient Data Extraction via Epic FHIR APIs
AkshatMedTech developed a FHIR-based data extraction pipeline that programmatically retrieves patient profiles from Epic at scale. The integration pulls structured data across multiple FHIR resources, like Patient demographics, Conditions (problem list), MedicationRequests, Observations (lab results), Encounters, and Procedures. Then it is aggregated into unified candidate profiles. Instead of coordinators opening one chart at a time, the system queries thousands of records in batches, building a searchable patient dataset that reflects the hospital’s current population.
Patient Matching DashboardExport Report
Active trials12Screened today1,420Potential matches84FHIR API statusConnected
● Live extraction job · Phase 3 Oncology TrialBulk data: 12,458 syncing…
FHIR resources pulled
Patient
Condition
MedicationRequest
Observation
Encounter
Procedure
GET /Patient?_count=200 → 200 OK
GET /Condition?patient=… → 200 OK
Batch 62/66 · 12,458 records
Automated Eligibility Matching Engine
Because patient profiles feed into a rule-based matching engine that evaluates each candidate against a trial’s specific inclusion and exclusion criteria. The engine supports multi-condition logic like age ranges, ICD-10 diagnosis codes, laboratory thresholds, active medication classes, and procedure history. This produces a scored eligibility assessment for each patient. Coordinators receive a ranked list of candidates rather than an undifferentiated wave of charts, cutting the screening-to-identification cycle from days to seconds.
Eligibility Matching EngineRe-run
Inclusion criteria
Age 18–75
Confirmed diagnosis (ICD-10 C34)
ECOG 0–2
Adequate organ function
Exclusion criteria
Active pregnancy
eGFR < 30
CNS metastases
Ranked candidates
PT-4821 · 92%
PT-4817 · 87%
PT-4809 · 81%
PT-4795 · 66%
Standardized Data Normalization and De-identification
Patient data extracted from Epic arrives in varying formats and coding standards across different hospital implementations. The platform builds a normalization layer that maps incoming FHIR data to standard clinical vocabularies: SNOMED CT for conditions, RxNorm for medications, and LOINC for lab observations, ensuring consistent matching regardless of how individual Epic instances encode their clinical data. A parallel de-identification framework protects health information before any data leaves the trusted environment, maintaining full HIPAA compliance throughout the pipeline.
Normalization & De-identificationHIPAA ✓
Vocabulary mapping
SourceStandardCode
E11.9SNOMED CT44054006
metformin 500RxNorm860975
HbA1cLOINC4548-4
I50.9SNOMED CT84114007
Pipeline
ExtractNormalizeDe-identifyMatch
name → ▮▮▮▮▮▮
MRN → tok_9f3a…
dob → 1960 (year)
Provider-facing Candidate Dashboard
Matched candidates are surfaced through a provider-facing dashboard designed to fit within existing Epic clinical workflows. Research coordinators and principal investigators can review candidate lists, filter by trial, view supporting clinical evidence for each match, and initiate outreach without leaving Epic. The dashboard provides real-time visibility into the enrollment pipeline, replacing spreadsheets and manual tracking with a centralized, auditable recruitment interface.
Candidate Priority QueueFilter by trial
Candidates · Trial 2024-X
Marcus WeiNSCLC92%Outreach
Elena R.Breast ca87%Contacted
Sarah J.T2DM81%New
EvidenceDx: C34.90 · ECOG 1 · eGFR 78
Enrollment pipeline
Identified 248Reviewed 121Contacted 64Enrolled 22
Epic Readiness and Listing
The integration aligned with Epic’s interoperability standards, authentication protocols, and App Orchard onboarding requirements. The solution leverages Epic’s OAuth 2.0 backend authentication, supports SMART on FHIR launch contexts, and conforms to Epic’s data access governance policies. This positions the client’s product for listing within the Epic ecosystem, making it accessible to the 300+ million patient records managed across Epic’s installed hospital base.
Epic Readiness & ListingSubmit
Checklist
OAuth 2.0 backend auth
SMART on FHIR launch
FHIR R4 conformance
Data access governance
Listing statusInteroperability validation
Epic ecosystem reach
HospitalsAcademic centersCROs
Impact
Patient Matching DashboardRun Matching
Patients screened12,847Candidates3,291Pending review248Active trials18
Patient Match Queue
IDAge · SexDiagnosisMatchStatus
PT-482158 · MNSCLC Stage IIIB92%Eligible
PT-481763 · FBreast ca, HER2+87%Review
PT-480949 · MT2DM, CKD 381%Eligible
PT-480271 · FHeart failure44%Excluded
Matching Summary
64%
Top matched trialOncology Phase 3 · Lung
Up to 70% faster patient identification
via automated bulk extraction and rule-based matching, reducing the overall time.
5x more eligible patients
surfaced that manual review consistently missed, significantly expanding the eligible patient pool per trial.
40% reduction in coordinator workload
via automated eligibility assessment.
Epic Marketplace Positioning
that gives the client commercial access to hospitals and academic medical centers running the Epic EHR platform.
Looking to develop an AI-powered diagnostic app or scale your healthcare innovation?
Contact Us NowOur Tech Stack
Contact Us TodayEpicEpic
🔥FHIR
djDjango
PgPostgreSql
awsAWS
EC2EC2
S3S3
NNode JS
DkDocker
λLambda
FAFastAPI
KfApache Kafka