Residency · Residency · Medical Genetics Genomics
Implementing Genomic Medicine in Health Systems
Introduction
Translating advances in genomics from research and specialty clinics into routine clinical care across health systems is the central challenge of the current era of genomic medicine. Implementation requires coordinating technology, informatics, workforce development, clinical decision support, reimbursement, and institutional culture change. This lecture examines frameworks, evidence, and practical strategies for health system-wide genomic medicine implementation.
The Implementation Gap
Where We Stand
Over 75,000 genetic tests are clinically available, yet the majority of patients who would benefit from testing do not receive it. Only about 15-20% of individuals meeting clinical criteria for hereditary cancer genetic testing are referred and tested. Familial hypercholesterolemia, which affects 1 in 250 individuals, is diagnosed in fewer than 10% of cases. Pharmacogenomic testing is available for dozens of gene-drug pairs but implemented systematically at fewer than 20 US health systems. The gap between evidence and practice is widening as genomic discoveries outpace clinical adoption.
Barriers to Implementation
Several interconnected barriers slow implementation. Most physicians receive fewer than 20 hours of genetics education in medical school, creating a provider knowledge deficit. The workforce shortage is stark, with approximately 1,500 clinical geneticists and roughly 5,500 genetic counselors in the US, concentrated in academic centers. EHR systems are not designed for complex genomic data storage, interpretation, and clinical decision support, representing a major informatics infrastructure gap. Reimbursement remains uncertain, with variable insurance coverage and significant prior authorization burden. Many patients are themselves unaware of the availability and relevance of genetic testing.
Implementation Science Frameworks
The RE-AIM Framework
The RE-AIM framework evaluates implementation across five dimensions. Reach asks what proportion of the eligible population is being reached. Effectiveness asks whether the intervention produces the desired outcomes in real-world settings. Adoption examines whether clinical providers and sites are adopting the program. Implementation assesses whether the program is being delivered as intended. Maintenance determines whether the program is sustained over time.
The CFIR (Consolidated Framework for Implementation Research)
The CFIR examines implementation through five domains: intervention characteristics (complexity, cost, evidence strength), outer setting (policy, guidelines, patient needs), inner setting (organizational culture, readiness, resources), individuals involved (knowledge, attitudes, self-efficacy), and process (planning, engaging, executing, evaluating).
The Genomic Medicine Implementation Model
This model is specific to genomic medicine and considers clinical indication and evidence base, test selection and laboratory partnership, results integration and clinical decision support, provider education and competency, patient engagement and shared decision-making, and outcome measurement and quality improvement.
Key Implementation Domains
Clinical Decision Support (CDS)
EHR-integrated alerts can be triggered by prescribing events for pharmacogenomics, diagnosis codes for hereditary cancer criteria, or family history entries. Best Practice Advisories (BPAs) are non-interruptive alerts suggesting genetic testing when clinical criteria are met. Order sets bundle genetic test orders with appropriate counseling and follow-up steps. Integrated knowledge bases link variants to clinical recommendations. Effective CDS requires ongoing maintenance as evidence evolves and new gene-drug or gene-disease associations are established.
Electronic Health Record Integration
Genetic test results must be stored as discrete, computable data elements rather than as PDF attachments to enable downstream use. The HL7 FHIR Genomics Implementation Guide provides standards for representing genomic data in interoperable formats. Variant-level data storage enables automated CDS and reanalysis as knowledge evolves. Structured family history collection tools integrated with genetic risk assessment algorithms improve identification of at-risk individuals. Interoperability is essential so that genetic results remain accessible across health systems as patients move between providers.
Workforce Strategies
The mainstreaming model trains non-genetics specialists to manage straightforward genetic indications, such as oncologists ordering BRCA1/2 testing or cardiologists managing HCM genetic testing, with genetics backup for complex cases. Embedding genetic counselors within specialty clinics in oncology, cardiology, prenatal care, and neurology improves access and integration. Telegenetics extends reach to rural and underserved areas through video-based genetic counseling. Digital pre-test education using chatbots, videos, and interactive tools extends counselor capacity. Mandatory genomic medicine modules in continuing medical education for all specialties help address the knowledge gap.
Laboratory Partnership
Selecting a reference laboratory or developing internal laboratory capacity for genomic testing requires attention to several factors, including test menu breadth, turnaround time, variant interpretation quality, reanalysis policies, and EHR integration capability. Standardization of test selection criteria and ordering processes across the health system ensures consistency, and variant interpretation alignment between the laboratory's framework and institutional clinical practices is important.
Exemplar Programs
Geisinger MyCode Community Health Initiative
The Geisinger MyCode program is a population-level genomic screening initiative with more than 300,000 participants. Exome sequencing is performed with return of ACMG secondary findings variants and pharmacogenomic results. Pathogenic variants have been identified in approximately 3.5% of participants, and most had no prior awareness of their genetic risk. The program has demonstrated the feasibility of integrating population-scale genomic screening into a learning health system, with clinical interventions such as cascade testing, cancer screening, and statin therapy implemented following results.
| Program | Institution/Region | Model | Scale | Key Achievement |
|---|---|---|---|---|
| MyCode | Geisinger | Population exome screening + ACMG SF return | >300,000 participants | ~3.5% with actionable PV; cascade testing implemented |
| eMERGE | Multi-site NIH network | EHR-integrated genomic CDS research | Multiple health systems | Developed tools for CDS and diverse population return |
| NHS Genomic Medicine Service | England (national) | Centralized test directory + regional lab hubs | National | Standardized genomic testing across NHS |
| All of Us | NIH (US) | Population-scale research + result return | >1 million | Returning hereditary risk + PGx to diverse participants |
| IGNITE | Multi-site NIH network | Genomic medicine implementation research | Multiple sites | Evaluated strategies for genomic integration |
eMERGE (Electronic Medical Records and Genomics) Network
The eMERGE Network is an NIH-funded network of health systems implementing genomic medicine. It has developed tools for EHR-integrated genomic CDS, clinical trial recruitment, and patient engagement. The fourth phase, eMERGE IV, focused on returning genomic results in diverse populations with integrated clinical decision support.
NHS Genomic Medicine Service (England)
England has implemented national genomic testing through the NHS Genomic Medicine Service, featuring a centralized test directory with standardized clinical indications and regional Genomic Laboratory Hubs providing comprehensive genomic testing. A national data infrastructure supports variant interpretation and research.
All of Us Research Program
The All of Us Research Program is returning hereditary disease risk and pharmacogenomic results to more than one million diverse participants. It is testing the model of population-scale genomic medicine outside of traditional clinical indications and building infrastructure for genomic data integration across multiple health systems.
Governance and Quality
Institutional Governance
Genomic medicine committees provide multidisciplinary oversight of test selection, result reporting, CDS content, and clinical protocols. Clinical pathways standardize workflows from test indication through ordering, counseling, result return, and follow-up management. Quality metrics track referral-to-test completion rates, time to results, variant reclassification, and patient outcomes.
Outcome Measurement
Process metrics include the number of patients tested, turnaround time, and CDS alert firing and override rates. Clinical outcomes encompass disease detection rates, management changes, and health outcomes such as cancer stage at diagnosis. Patient-reported outcomes capture understanding, satisfaction, psychological impact, and decisional regret. Economic outcomes assess cost per diagnosis, downstream healthcare utilization, and cost-effectiveness.
Health Economics and Sustainability
Genetic testing costs range from approximately $250 for a single gene test to over $5,000 for genome sequencing. The economic case for genomic medicine rests on ending the diagnostic odyssey by reducing unnecessary tests, referrals, and procedures; early detection and prevention through cancer screening, cardiac surveillance, and familial hypercholesterolemia management; precision prescribing that avoids adverse drug reactions and treatment failures; and reduced hospitalizations and emergency care. Value-based models including outcomes-based reimbursement for high-cost gene therapies and installment payment plans are being explored as sustainable approaches. Insurance coverage and prior authorization remain the most significant practical barriers.
Future Directions
The field is moving from indication-based to population-based genomic screening approaches for conditions like familial hypercholesterolemia and hereditary cancer syndromes. Genomic learning health systems will establish continuous cycles of data generation, analysis, implementation, and evaluation. AI-augmented clinical genomics will automate phenotyping, variant prioritization, and CDS generation. Adapting implementation models for resource-limited settings represents an important global challenge. Patient-facing genomic tools will provide direct access to genomic results with decision support and educational resources.
Clinical Pearls
The primary barrier to genomic medicine implementation is not technology but rather the integration of genomic data into clinical workflows, provider education, and sustainable reimbursement models. Effective clinical decision support integrated into the EHR is the single most important enabler of genomic medicine at scale, because without it, genetic results remain siloed in specialty clinics. The mainstreaming model, which trains non-genetics specialists to manage straightforward genetic testing with genetics backup, is essential for scaling genomic medicine beyond the genetics workforce capacity. Population-scale programs like Geisinger MyCode demonstrate that systematic genomic screening can identify clinically actionable findings in approximately 3-5% of unselected individuals.
References
- Manolio TA, Rowley R, Williams MS, et al. Opportunities, resources, and techniques for implementing genomics in clinical care. Lancet. 2019;394(10197):511-520.
- Williams MS, Buchanan AH, Davis FD, et al. Patient-centered precision health in a learning health care system: Geisinger's genomic medicine experience. Health Affairs. 2018;37(5):757-764.
- Damschroder LJ, Aron DC, Keith RE, et al. Fostering implementation of health services research findings into practice: a consolidated framework for advancing implementation science. Implementation Science. 2009;4:50.
- Stark Z, Dolman L, Manolio TA, et al. Integrating genomics into healthcare: a global responsibility. American Journal of Human Genetics. 2019;104(1):13-20.