Phased Implementation Roadmap
Build a comprehensive pharmacogenomics program in three phases
PHASE 1: START HERE
Generic BPA + CDS Hooks
MODERATE Burden
- Proactive safety foundation
- Moderate Epic IT effort
- Epic-native data
- Smart BPA alerts
- Proves clinical value
PHASE 2: ADD NEXT
Pharmacist Alerts
MODERATE Burden
- Expert mediation layer
- Moderate Epic IT effort
- Pharmacist triage
- Reduced alert fatigue
- Best clinical outcomes
PHASE 3: COMPLETE
iFrame Embedded
LOW IT Burden
- Universal access
- Low Epic IT effort
- Ad-hoc lookups
- Passive check
- Broadest coverage
Detailed Comparison Matrix
| Dimension |
1. Generic BPA + CDS Hooks |
2. Pharmacist Alerts |
3. iFrame Embedded |
| Epic IT Burden |
MODERATE |
MODERATE |
LOW |
| Maintenance Burden |
LOW
Minimal updates |
LOW
Template updates |
ZERO
None |
| Proactive Alerts |
✓ Yes
(order-time) |
✓ Yes
(background) |
✗ No
(passive) |
| Alert Fatigue Risk |
MODERATE
1 smart alert |
LOW
Pharmacist filter |
NONE
No alerts |
| Clinical Experience |
GOOD
Epic-native |
EXCELLENT
Pharmacist support |
GOOD
Embedded |
| Provider Workflow |
Interrupted
(BPA fires) |
Pharmacist contact
(only if needed) |
Manual check
(click tab) |
| Scalability |
EXCELLENT
Easy to add genes |
EXCELLENT
Linear with patients |
EXCELLENT
Helix handles |
Key Tradeoffs: Where Each Approach Falls
Epic IT Burden
3
iFrame
2
Pharmacist
1
Generic BPA
Low
High
Clinical Value & Proactivity
3
iFrame
1
Generic BPA
2
Pharmacist
Passive (manual check)
Proactive (alerts + support)
Recommended Strategy: Implement all three approaches in phases. Start with Approach 1 (Generic BPA + CDS Hooks) to establish proactive safety foundation.
Add Approach 2 (Pharmacist Alerts) for expert mediation and reduced provider burden. Complete with Approach 3 (iFrame) for comprehensive ad-hoc access.
Together, these approaches create a complete pharmacogenomics program serving different clinical needs.
⭐ Recommended Strategy: Phased Implementation of All Three Approaches
These three approaches are complementary, not mutually exclusive. The recommended strategy is to implement all three in phases, building a comprehensive pharmacogenomics program that serves different clinical needs and provider preferences.
Phase 1: Start with Generic BPA + CDS Hooks
- Foundation for Safety: Establishes proactive alerts for highest-risk drug-gene interactions
- Epic-Native Data: Stores genetic results in Epic, enabling all future approaches
- Moderate Setup: Initial implementation creates infrastructure for entire program
- Smart Alerting: Single intelligent BPA catches critical cases at order time
- Immediate Value: Demonstrates clinical impact and builds stakeholder confidence
- Why First: Creates data foundation and proves concept before adding complexity
Phase 2: Add Pharmacist Alerts
- Builds on Phase 1: Uses existing BPA infrastructure plus adds pharmacy workflow
- Human Expertise Layer: Pharmacist reviews and triages alerts before contacting providers
- Reduced Alert Fatigue: Filters recommendations through specialist expertise for complex cases
- Enhanced Clinical Outcomes: Pharmacist judgment improves intervention quality and provider acceptance
- Staffing Required: Dedicated pharmacist support with predictable resource needs
- Why Second: Requires operational maturity and pharmacist training, best added after foundational alerts are stable
Phase 3: Add iFrame Embedded
- Completes the Suite: Provides passive access for ad-hoc lookups and non-CPIC medications
- Quick Addition: Low Epic IT effort, no dependency on Phases 1 & 2
- Broadest Coverage: Enables clinicians to check any medication anytime, beyond alert-driven cases
- Zero Maintenance: No Epic IT burden for updates after initial setup
- Provider Choice: Some clinicians prefer manual lookup over interruption-based alerts
- Why Third: Most clinical value comes from proactive alerts; add this for comprehensive coverage
🎯 End-State Vision: Comprehensive Pharmacogenomics Program
With all three approaches implemented, your organization will have:
- Proactive Safety Net: Generic BPA catches high-risk cases automatically at order time
- Expert Mediation: Pharmacist alerts reduce provider burden while improving recommendation quality
- Universal Access: iFrame enables any clinician to check any medication anytime
- Provider Choice: Different workflows for different preferences and clinical scenarios
- Maximum Coverage: Serves alert-driven, pharmacist-mediated, and ad-hoc lookup needs
- Scalable Foundation: Infrastructure supports program growth and new use cases
Implementation Approach: Each phase proves value before adding the next layer, minimizing risk while building toward comprehensive pharmacogenomics integration.
Next Steps: Planning Your Phased Rollout
- Review all 3 approaches with key stakeholders (CMIO, Pharmacy Lead, Epic Analyst) as complementary components of a comprehensive program
- Assess Phase 1 readiness using the decision framework - Epic IT capacity, data governance requirements, clinical priority
- Create Phase 1 implementation plan - plan Generic BPA + CDS Hooks setup and testing
- Establish success metrics - alert acceptance rates, intervention quality, provider feedback for Phase 1
- Plan Phase 2 resources - identify pharmacist staffing needs, create training plan
- Set Phase 2 trigger - when to add Pharmacist Alerts after Phase 1 is stable
- Budget for Phase 3 - reserve Epic IT capacity for iFrame setup when ready to complete the suite
- Document phased strategy - create roadmap showing all three phases and decision gates
- Secure executive buy-in - present phased approach as risk-minimizing strategy with proven value at each stage
- Begin Phase 1 implementation - start with foundational proactive alerts to prove clinical value early
Key Message: These approaches work together, not against each other. By implementing all three in phases,
you build a comprehensive pharmacogenomics program that serves proactive safety (Phase 1), expert mediation (Phase 2),
and universal access (Phase 3) needs. Each phase proves value before adding the next, minimizing risk while maximizing clinical impact.
The phased approach allows your organization to adjust timing based on resources, priorities, and lessons learned.