Prompt Details
Model
(claude-5-opus)
Token size
2,380
Example input
[Paste or describe the available enterprise architecture information]: Current State: 3-hospital network (Main, North Campus, Urgent Care) Patient records exist in 3 separate systems (not integrated) Lab results emailed between departments Radiology images stored on local servers (no central repository) Prescription management manual (paper + email) Pain Points: Duplicate patient records (same person registered 3 times) Lab orders get lost No medication history visibility across hospitals Doctors cannot access imaging from other campuses Wait times unknown (no real-time bed tracking)
[Organization or project name]: Metropolitan Health System (MHS) β 3-Hospital Network
[Documents / Diagrams / Processes / Data Models / Policies / APIs / Other]: 3 separate EHR systems (Epic, Cerner, proprietary) Radiology PACS system (local storage only) Lab information system (LIS) Insurance verification tool (legacy, unreliable) Patient portal (Main hospital only) Excel sheets for bed management Handwritten nursing notes (some scanned)
[Documentation / Architecture Analysis / Knowledge Management / Transformation / Audit]: Data Integration + Compliance Audit β Unify 3 hospitals + ensure HIPAA/security
[Markdown / Obsidian Notes / Folder Structure / YAML / Dataview / Mermaid]: Markdown + Obsidian Notes + Folder Structure + Mermaid Diagrams + Dataview Compliance Dashboard
Example output
# π§ͺ TEST EXECUTION
## π§ PHASE 1 β ARCHITECTURE DISCOVERY
### Discovered Elements:
**Business Concepts:**
- Patient (core entity)
- Hospital / Campus
- Department (ER, ICU, Surgery, Radiology, Lab)
- Clinical Staff (Doctor, Nurse, Technician)
- Treatment / Encounter
- Medication / Prescription
- Medical Imaging
- Lab Test / Result
- Insurance / Payer
- Appointment / Bed
- Medical Record / Note
**Data:**
- Patient demographics (3 copies, unlinked)
- Medical history (fragmented)
- Lab results (emailware)
- Imaging files (local storage)
- Prescription records (paper/email)
- Appointment calendar
- Bed availability
- Staff schedules
- Insurance eligibility
**Processes:**
- Patient admission (varies by hospital)
- Clinical diagnosis workflow
- Lab order β Collection β Results
- Imaging order β Acquisition β Interpretation
- Prescription β Pharmacy β Dispensing
- Discharge planning
- Emergency transfer
- Insurance verification
- Compliance audit
**Locations:**
- Main Hospital (downtown, 400 beds)
- North Campus (suburb, 150 beds)
- Urgent Care Clinic (satellite, 30 beds)
- Central Imaging Repository (proposed)
- Cloud storage (candidate)
**People/Responsibilities:**
- Primary Care Doctor
- Specialist (Cardiology, Orthopedic, etc.)
- Nursing Staff
- Radiology Technician
- Lab Technician
- Medical Records Officer
- Compliance Officer
- IT Staff
**Timing/Events:**
- Emergency admission (immediate)
- Scheduled admission (planned)
- Patient transfer between hospitals
- Lab result delivery (24-48 hours)
- End-of-shift handoff
- Discharge event
- Medication refill event
**Motivation:**
- Improve patient safety (single view of medical history)
- Reduce duplicate records
- Enable emergency transfers with full context
- Ensure HIPAA compliance
- Reduce operational inefficiency
- Support quality metrics
- Enable research
**Systems:**
- EHR System (3 instances)
- PACS (Radiology imaging)
- Lab Information System
- Patient Portal
- Scheduling System
- Bed Management System
- Staff Scheduling System
- Insurance Verification System
- Master Patient Index (proposed)
- Identity Resolution Engine (needed)
---
## ποΈ PHASE 2 β ZACHMAN MAPPING
| Perspective | Interrogative | Artifact | Evidence | Confidence |
|---|---|---|---|---|
| **SCOPE** | WHAT | Patient, Treatment, Lab Result, Imaging, Prescription | Clinical workflow | HIGH |
| **SCOPE** | WHERE | 3 hospitals, Cloud (proposed) | Current distributed model | HIGH |
| **SCOPE** | WHO | Doctors, Nurses, Techs, Admin, Patients | Organizational roles | HIGH |
| **SCOPE** | WHEN | Admission β Discharge, Emergency (real-time) | Business timing | HIGH |
| **SCOPE** | WHY | Patient safety, HIPAA compliance, efficiency | Regulatory + operational | HIGH |
| **BUSINESS** | WHAT | Patient admission process | Documented (3 variants) | MEDIUM |
| **BUSINESS** | HOW | Manual check-in + paper + email | Current state observation | MEDIUM |
| **BUSINESS** | WHERE | Main + North + Urgent Care | Current distribution | HIGH |
| **BUSINESS** | WHO | Admissions staff, nurses, doctors | Known roles | HIGH |
| **BUSINESS** | WHEN | Appointment β Registration β Triage β Bed | Sequential process | MEDIUM |
| **SYSTEM** | WHAT | EHR data model (3 different schemas) | Systems exist | HIGH |
| **SYSTEM** | HOW | Epic, Cerner, Proprietary integrations | Systems documented | MEDIUM |
| **SYSTEM** | WHAT | Patient demographics (3 duplicate copies) | Database exists | HIGH |
| **SYSTEM** | WHAT | Lab results, Imaging metadata | Data exists | HIGH |
| **TECHNOLOGY** | WHAT | Epic EHR, Cerner EHR, PACS, LIS | Tech stack | HIGH |
| **TECHNOLOGY** | HOW | HL7 messaging (some), APIs (partial) | Integration attempted | MEDIUM |
| **TECHNOLOGY** | WHERE | On-premise + Cloud (partial) | Deployment model | MEDIUM |
| **COMPONENT** | WHAT | Master Patient Index | MISSING | CRITICAL |
| **COMPONENT** | WHAT | Patient identity resolution rules | MISSING | CRITICAL |
| **COMPONENT** | WHAT | Integration middleware | MISSING | CRITICAL |
| **COMPONENT** | WHAT | Unified medical record schema | MISSING | CRITICAL |
| **OPERATION** | WHAT | Access logs, audit trail | UNCERTAIN | LOW |
| **OPERATION** | HOW | HIPAA compliance procedures | Partial documentation | LOW |
| **OPERATION** | WHAT | Disaster recovery plan | UNKNOWN | CRITICAL |
---
## π PHASE 3 β ZACHMAN COVERAGE ANALYSIS
### Coverage Matrix
```
PERSPECTIVE | WHAT | HOW | WHERE | WHO | WHEN | WHY
---------------|-------|-------|-------|-------|-------|-------
SCOPE | β | β | β | β | β | β
BUSINESS | β | ~ | β | β | ~ | β
SYSTEM | β | ~ | ~ | ~ | ~ | ~
TECHNOLOGY | β | ~ | β | β | β | β
COMPONENT | β | β | β | β | β | β
OPERATION | ~ | ~ | β | β | β | β
```
### Critical Gaps by Impact
**ARCHITECTURE CRISIS (Must Fix Now):**
- β Master Patient Index (MPI) - No unified patient identity
- β Data integration layer - Systems not talking
- β Unified medical record schema - Each hospital different
- β Single sign-on (SSO) - Staff logs in 3 times
- β Audit trail & logging - HIPAA compliance risk
**OPERATIONAL GAPS (High Priority):**
- β Real-time bed availability
- β Emergency transfer protocols
- β Medication interaction checking (across hospitals)
- β Disaster recovery plan
- β Data backup strategy
**GOVERNANCE GAPS (Medium Priority):**
- β Data governance policy
- β Integration standards (HL7, FHIR)
- β Access control matrix
- β Data retention policy
- β Security incident procedures
---
## π§© PHASE 4 β OBSIDIAN VAULT ARCHITECTURE
```
π Metropolitan Health System Architecture/
β
βββ π 00 - Overview/
β βββ π Architecture Home
β βββ π Zachman Matrix
β βββ π Network Overview
β βββ π 3-Hospital Integration Map
β βββ π Glossary (Medical Terms)
β
βββ π 01 - Context/
β βββ π Regulatory Requirements (HIPAA, HITRUST)
β βββ π Business Objectives
β βββ π Current Pain Points
β βββ π Stakeholder Map
β βββ π Compliance Status
β
βββ π 02 - Clinical/
β βββ π Patient Admission Process
β βββ π Emergency Transfer Protocol
β βββ π Diagnostic Workflow
β βββ π Treatment Planning
β βββ π Discharge Process
β βββ π Medication Management Process
β
βββ π 03 - Hospital Locations/
β βββ π Main Hospital/
β β βββ π Main Hospital Profile
β β βββ π Main - EHR System (Epic)
β β βββ π Main - Departments
β β βββ π Main - Bed Capacity
β βββ π North Campus/
β β βββ π North Campus Profile
β β βββ π North - EHR System (Cerner)
β β βββ π North - Departments
β β βββ π North - Bed Capacity
β βββ π Urgent Care/
β βββ π Urgent Care Profile
β βββ π Urgent Care - EHR System (Proprietary)
β βββ π Urgent Care - Check-in Process
β
βββ π 04 - Systems/
β βββ π Epic EHR (Main Hospital)
β βββ π Cerner EHR (North Campus)
β βββ π Proprietary EHR (Urgent Care)
β βββ π PACS (Radiology System)
β βββ π Lab Information System (LIS)
β βββ π Patient Portal
β βββ π Scheduling System
β βββ π Master Patient Index (PLANNED)
β βββ π Identity Resolution Engine (PLANNED)
β βββ π Integration Middleware (PLANNED)
β
βββ π 05 - Data/
β βββ π Patient Demographic Data Model
β βββ π Medical Record Data Model
β βββ π Lab Result Data Model
β βββ π Imaging Metadata Model
β βββ π Prescription Data Model
β βββ π Appointment Data Model
β βββ π Staff Directory Model
β βββ π Insurance/Payer Model
β βββ π Master Data Management Strategy
β
βββ π 06 - Technology/
β βββ π Epic Platform
β βββ π Cerner Platform
β βββ π PACS Infrastructure
β βββ π Cloud Strategy
β βββ π HL7/FHIR Standards
β βββ π API Layer Design
β βββ π Security Architecture
β βββ π Network Architecture
β βββ π Backup & Disaster Recovery
β
βββ π 07 - Integration/
β βββ π EHR Integration Strategy
β βββ π PACS Integration
β βββ π Lab Integration
β βββ π Insurance Verification Integration
β βββ π Patient Portal Integration
β βββ π Interoperability Standards
β
βββ π 08 - Operations/
β βββ π Admission Procedures (3 variants)
β βββ π Emergency Transfer Procedures
β βββ π Bed Management Process
β βββ π Staff Scheduling Process
β βββ π Shift Handoff Protocol
β βββ π Monitoring & Alerting
β βββ π Incident Response
β
βββ π 09 - Governance/
β βββ π HIPAA Compliance Framework
β βββ π Data Governance Policy
β βββ π Access Control Matrix
β βββ π Data Retention Policy
β βββ π Information Security Policy
β βββ π Audit Procedures
β βββ π Change Management
β βββ π Regulatory Requirements
β
βββ π 10 - Architecture Decisions/
β βββ π ADR-001: Multi-EHR Strategy
β βββ π ADR-002: Master Patient Index Approach
β βββ π ADR-003: FHIR vs HL7
β βββ π ADR-004: Cloud vs On-Premise
β βββ π ADR-005: Single Sign-On Solution
β βββ π ADR-006: Disaster Recovery Strategy
β
βββ π 11 - Gaps & Issues/
β βββ π CRITICAL: Master Patient Index Missing
β βββ π CRITICAL: No Data Integration Layer
β βββ π CRITICAL: Unified Schema Needed
β βββ π CRITICAL: HIPAA Audit Trail Incomplete
β βββ π CRITICAL: Disaster Recovery Plan Missing
β βββ π HIGH: Duplicate Patient Records
β βββ π HIGH: Lab Order Workflow Broken
β βββ π HIGH: Imaging Not Centralized
β βββ π HIGH: Real-Time Bed Tracking Missing
β βββ π MEDIUM: Staff Don't See Across Hospitals
β βββ π MEDIUM: Patient Portal Not Enterprise-Wide
β
βββ π 12 - Compliance/
β βββ π HIPAA Compliance Checklist
β βββ π Audit Trail Requirements
β βββ π Access Log Procedures
β βββ π Encryption Standards
β βββ π Breach Notification Procedures
β βββ π Compliance Assessment
β
βββ π 13 - References/
βββ π HL7 Standards
βββ π FHIR Specifications
βββ π HIPAA Regulations
βββ π Vendor Documentation
βββ π Industry Best Practices
```
---
## π PHASE 5 β KNOWLEDGE LINKING
### Critical Link Strategy
```
Patient (business concept)
β identified by
Master Patient Index (MISSING)
β requires
Identity Resolution Engine (MISSING)
β matches records from
Epic EHR + Cerner EHR + Proprietary EHR
β accesses
Patient Demographics (3 copies, unlinked)
Patient Admission Process (business)
β uses different procedures at
Main Hospital β North Campus β Urgent Care
β could be unified by
Integration Middleware (MISSING)
Emergency Transfer Process (business)
β requires access to
Unified Medical Record (MISSING)
β includes
Lab Results + Imaging + Medications
β stored in separate systems
LIS + PACS + 3x EHRs
Medication Management (business)
β checks interactions via
Central Pharmacy Database (MISSING)
β connects to
HIPAA Audit Trail (INCOMPLETE)
HIPAA Compliance (regulatory)
β requires
Audit Trail + Access Logs + Encryption
β currently MISSING in
PACS System + Lab Results + Email-based Communication
```
### Relationship Types for Healthcare
- `treats` β Doctor treats Patient
- `performs_lab_test` β Technician orders Lab
- `interprets_imaging` β Radiologist views PACS
- `prescribes` β Doctor prescribes Medication
- `transfers_patient` β Hospital A β Hospital B
- `stores_data` β System persists Data
- `integrates_with` β System connects System
- `requires_mpi` β Process needs Master Patient Index
- `violates_hipaa` β Gap creates compliance risk
- `depends_on` β System needs System
---
## π PHASE 6 β NOTE GENERATION
### Example Note 1: Critical Gap - Master Patient Index
```markdown
---
title: Critical Gap - Master Patient Index Missing
type: Gap
zachman_perspective: Component / Detailed
interrogative: WHAT / HOW
status: CRITICAL - BLOCKS INTEGRATION
owner: Health Information Management
impact: ENTERPRISE-WIDE
priority: P0 - DO FIRST
affects: "[[Patient Admission Process]]", "[[Emergency Transfer Protocol]]", "[[Unified Medical Record]]"
depends_on: "[[Identity Resolution Engine]]", "[[Data Governance Policy]]"
blocked_by: "[[Duplicate Patient Records]]"
compliance_risk: "[[HIPAA Compliance Framework]]"
estimated_effort: 6-9 months
estimated_cost: $2.5M - $4M
---
# Critical Gap: Master Patient Index Missing
## Executive Summary
Metropolitan Health System operates 3 independent EHR systems with NO unified patient identity across hospitals. Same patient can have 3 different medical records.
## The Problem
### Evidence
- Patient registers at Main Hospital (Epic) β ID: MH-12345
- Same patient goes to North Campus (Cerner) β ID: NC-67890
- Same patient visits Urgent Care (Proprietary) β ID: UC-11111
- **Result**: 3 medical records for 1 person
### Current State Challenges
1. **Duplicate Records**: Estimated 15-20% of patients have multiple records
2. **Fragmented History**: Doctor cannot see complete medical history in emergency
3. **Safety Risk**: Allergies recorded only at Main β Unknown at North
4. **Insurance Issues**: Multiple claims for same patient β Billing errors
5. **Quality Metrics**: Cannot accurately track patient outcomes across network
### Business Impact
| Issue | Impact | Severity |
|---|---|---|
| Wrong medication prescribed | Patient harm | **CRITICAL** |
| Duplicate imaging ordered | $500K/year waste | HIGH |
| Failed emergency transfer | Patient safety | **CRITICAL** |
| Insurance denials | $2M/year revenue loss | HIGH |
| Compliance audit failure | Regulatory penalties | **CRITICAL** |
## Zachman Classification
| Interrogative | Classification |
|---|---|
| **WHAT** | Master Patient Index database + matching algorithms |
| **HOW** | Deterministic matching + probabilistic matching + manual review |
| **WHERE** | Central database (cloud or on-premise) |
| **WHO** | Health Information Management team + clinicians |
| **WHEN** | Real-time on admission + daily batch reconciliation |
| **WHY** | Patient safety, compliance, operational efficiency |
## Required Components
### 1. Master Patient Index (MPI) Database
- Single source of truth for patient identity
- Links patient across 3 hospitals
- Stores: SSN, DOB, Name, Address, Phone
### 2. Identity Resolution Engine
- **Deterministic Matching**: Exact SSN match (100% confidence)
- **Probabilistic Matching**: Name + DOB + Address (95% confidence)
- **Fuzzy Matching**: Handles spelling variations
- **Manual Review**: Workflow for uncertain matches
### 3. Integration Middleware
- Receives patient registration from Epic + Cerner + Proprietary
- Performs identity match
- Returns unified patient ID
- Syncs patient record across systems
### 4. Audit Trail
- Log all identity matching decisions (HIPAA required)
- Track manual overrides
- Support audit procedures
## Implementation Approach
### Phase 1: Discovery (Weeks 1-2)
- [ ] Data inventory from 3 EHR systems
- [ ] Identify overlapping patient records
- [ ] Assess current duplicate rate
### Phase 2: Matching Rules (Weeks 3-6)
- [ ] Define deterministic rules (SSN, MRN)
- [ ] Define probabilistic rules (name + DOB + address)
- [ ] Design manual review workflow
- [ ] Build test dataset
### Phase 3: MPI Build (Weeks 7-16)
- [ ] Implement MPI database
- [ ] Build identity resolution engine
- [ ] Test with known duplicates
- [ ] Load historical patient data
### Phase 4: Integration (Weeks 17-24)
- [ ] Connect Epic registration β MPI
- [ ] Connect Cerner registration β MPI
- [ ] Connect Proprietary registration β MPI
- [ ] Real-time identity lookup
### Phase 5: Cutover (Weeks 25-26)
- [ ] Parallel testing
- [ ] Staff training
- [ ] Go-live with audit
- [ ] Monitor for issues
## Dependencies
### Must Have Before MPI
- [[Data Governance Policy]] - Who owns patient data?
- [[HIPAA Compliance Framework]] - Audit trail requirements
- [[Unified Medical Record Schema]] - How is data organized?
- [[Access Control Matrix]] - Who can see what data?
### Enables After MPI
- [[Unified Medical Record System]]
- [[Real-Time Bed Tracking]]
- [[Emergency Transfer Protocol]]
- [[Quality Metrics Reporting]]
## Risk Mitigation
| Risk | Mitigation |
|---|---|
| Incorrect matches harm patient | Manual review workflow + extensive testing |
| Staff bypass MPI (use old system) | Change management + training + monitoring |
| Performance issues | Cache common lookups + optimize queries |
| HIPAA audit trail incomplete | Design audit from day 1 |
## Success Criteria
- β All patients matched by SSN or manually reviewed
- β Zero unaccounted patient duplicates
- β Identity resolution < 500ms (real-time)
- β 100% audit trail for all matches
- β Staff can see unified patient across 3 hospitals
## Related Gaps Blocking This
- [[Unified Schema Missing]] - How are records unified?
- [[Integration Middleware Missing]] - How do systems talk?
- [[Data Governance Policy Missing]] - Who approves changes?
## Related Notes
- [[Patient Demographic Data Model]]
- [[Integration Middleware Planned]]
- [[HIPAA Audit Trail Requirements]]
- [[Duplicate Patient Records Issue]]
```
### Example Note 2: System - Epic EHR (Main Hospital)
```markdown
---
title: Epic EHR - Main Hospital
type: System
zachman_perspective: System / Logical
interrogative: WHAT / HOW / WHERE
status: Operational
owner: Main Hospital IT
vendor: Epic Systems Corporation
version: 2023.4
deployed: 2015
next_upgrade: 2026
related: "[[Patient Admission Process]]", "[[Master Patient Index]]", "[[Cerner EHR]]"
depends_on: "[[HIPAA Audit Trail]]"
integrates_with: "[[PACS System]]", "[[Lab Information System]]"
gaps: "[[Master Patient Index Missing]]", "[[Integration Middleware Missing]]"
---
# Epic EHR - Main Hospital
## Overview
Primary Electronic Health Record system serving Main Hospital (400 beds).
## Technical Specifications
| Attribute | Value |
|---|---|
| **Version** | 2023.4 |
| **Deployment** | On-premise (Main Hospital data center) |
| **Database** | Oracle 12c |
| **Users** | ~1,200 clinical staff |
| **Storage** | 15TB active records, 50TB archive |
| **Uptime SLA** | 99.5% |
| **Backup** | Daily snapshots + off-site replication |
## Current Functionality
### Supported Workflows
- β Patient registration & demographics
- β Appointment scheduling
- β Clinical encounter documentation
- β Medication management
- β Lab orders & results
- β Allergy & interaction checking
- β Discharge summaries
- β HIPAA audit logs
### Current Integrations
- β PACS (radiology imaging) - HL7 feed, 2-way
- β LIS (lab results) - HL7 feed, 1-way
- β Patient Portal (read-only, 24hr delay)
- β Bed management (manual export daily)
- ~ Insurance verification (unreliable, 40% timeout)
### Known Limitations
- β Cannot see patient records from Cerner (North Campus)
- β Cannot see patient records from Proprietary EHR (Urgent Care)
- β No real-time integration with scheduling system
- β No SSO (staff must log in separately from network)
- β Imaging retrieval slow (PACS integration overloaded)
## Data Model Summary
### Core Tables
```
patients (demographics)
encounters (admissions)
orders (medication, lab, imaging)
results (lab, vital signs)
notes (clinical documentation)
allergies (drug allergies)
problems (diagnoses - ICD-10)
medications (active prescriptions)
```
### Patient Records Stored
- **Total Patients**: ~450,000
- **Active Patients (last 2 years)**: ~180,000
- **Daily New Records**: ~200
- **Estimated Duplicates (in this system)**: ~15%
- **Data Quality Issues**: Multiple phone formats, inconsistent addresses
## Zachman Classification
| Interrogative | Answer |
|---|---|
| **WHAT** | Patient records, clinical notes, orders, results |
| **HOW** | Epic client-server architecture, database storage, HL7 messaging |
| **WHERE** | Main Hospital data center, Oracle database |
| **WHO** | Clinical staff (doctors, nurses, admin), patients (view-only access) |
| **WHEN** | Real-time entry, batch lab result feeds (3x/day) |
| **WHY** | Legal health record, compliance requirement, clinical decision support |
## Architecture Diagram
```mermaid
graph TB
A["Epic EHR<br/>Main Hospital"]
B["Oracle<br/>Database"]
C["PACS<br/>Radiology"]
D["LIS<br/>Lab Results"]
E["Patient<br/>Portal"]
F["Bed Mgmt<br/>Manual Export"]
A <-->|HL7| C
A <-->|HL7| D
A -->|REST API| E
A -->|nightly CSV| F
A <-->|storage| B
```
## Critical Dependencies
### Upstream Dependencies (feeds Epic)
- [[Lab Information System]] - Lab results via HL7
- [[PACS System]] - Imaging orders via HL7
- Insurance systems - Eligibility data (manual, unreliable)
### Downstream Dependencies (Epic feeds)
- [[Patient Portal]] - Public-facing view
- [[Bed Management System]] - Manual export daily
- [[Compliance Audit]] - HIPAA audit logs
## Key Gaps Affecting This System
1. **No Master Patient Index**
- Epic cannot match records with North Campus or Urgent Care
- Cannot merge duplicate patients across system
2. **No Integration Middleware**
- Each integration is point-to-point (hard to maintain)
- No central identity management
3. **Poor Integration with Other EHRs**
- Clinician must manually enter history from other hospitals
- Patient safety risk in emergencies
## HIPAA & Compliance
### Audit Trail
- β All data access logged
- β Logs retained for 6 years
- β Monthly access reviews conducted
- Status: **COMPLIANT**
### Data Encryption
- β Database encryption at rest
- β HTTPS for all remote access
- β Encryption key management by Epic
- Status: **COMPLIANT**
### Disaster Recovery
- β Daily backups off-site
- β Recovery time objective: 4 hours
- β Last tested: 6 months ago
- Status: **PARTIALLY COMPLIANT** (needs re-test)
## Performance Metrics
| Metric | Current | Target | Status |
|---|---|---|---|
| Patient lookup | 200ms | <100ms | β οΈ SLOW |
| Lab result receipt | 3-4 hours | <1 hour | β οΈ SLOW |
| Imaging availability | 30 minutes | <5 minutes | β οΈ SLOW |
| System uptime | 99.4% | 99.5% | β οΈ BELOW SLA |
| User adoption | 94% | 98% | β GOOD |
## Maintenance & Support
### Vendor Support
- **Vendor**: Epic Systems Corporation
- **Support Tier**: Premium 24/7
- **Response Time**: Critical issues < 1 hour
- **Contract Renewal**: 2025
### Internal Support
- **Team**: 12 FTE (developers + DBAs + analysts)
- **Tickets/Month**: ~200
- **Resolution Time**: Average 2 days
### Upgrade Schedule
- **Current**: 2023.4
- **Next Upgrade**: 2026 Q2 (Epic 2026 release)
- **Effort**: 6 months planning + testing
## Recommendations
### Immediate (Next 90 Days)
1. [ ] Improve PACS integration performance (cache imaging metadata)
2. [ ] Implement single sign-on (reduce manual login)
3. [ ] Document integration points for future MPI work
### Short Term (6-12 Months)
1. [ ] Migrate lab result feeds to use FHIR (cleaner standard)
2. [ ] Build integration middleware adapter for MPI
3. [ ] Implement real-time patient lookup for emergency transfers
### Long Term (12-24 Months)
1. [ ] Consolidate Epic + Cerner into single system (or integrate via MPI)
2. [ ] Cloud migration (reduce on-premise complexity)
3. [ ] Implement bi-directional imaging integration (faster access)
## Related Notes
- [[Cerner EHR - North Campus]]
- [[Master Patient Index Missing]]
- [[FHIR vs HL7 Decision]]
- [[Integration Middleware Planned]]
- [[Single Sign-On Strategy]]
```
---
## π§ PHASE 7 β ARCHITECTURE KNOWLEDGE GRAPH
```mermaid
graph TB
subgraph "Regulatory"
HIPAA["HIPAA Compliance<br/>Requirements"]
AUDIT["Audit Trail<br/>Requirements"]
end
subgraph "Business"
PAD["Patient Admission<br/>Process"]
ETP["Emergency Transfer<br/>Protocol"]
MMP["Medication Mgmt<br/>Process"]
OBJ["Patient Safety<br/>Objective"]
end
subgraph "Systems"
EPIC["Epic EHR<br/>Main"]
CERNER["Cerner EHR<br/>North"]
PROP["Proprietary EHR<br/>Urgent Care"]
PACS["PACS<br/>Radiology"]
LIS["LIS<br/>Lab"]
MPI["Master Patient Index<br/>MISSING"]
end
subgraph "Data"
PATDEMO["Patient Demographics<br/>x3 copies"]
MEDREC["Medical Records<br/>Fragmented"]
LABS["Lab Results"]
IMG["Imaging Metadata"]
RX["Prescriptions"]
end
subgraph "Technology"
MIDWARE["Integration Middleware<br/>MISSING"]
SCHEMA["Unified Schema<br/>MISSING"]
SSO["Single Sign-On<br/>MISSING"]
end
subgraph "Gaps - CRITICAL"
GAP1["Master Patient Index<br/>Missing"]
GAP2["Audit Trail<br/>Incomplete"]
GAP3["DR Plan<br/>Missing"]
GAP4["Emergency Protocol<br/>Undocumented"]
end
OBJ -->|requires| MPI
OBJ -->|requires| MEDREC
PAD -->|uses| EPIC
PAD -->|uses| CERNER
ETP -->|needs unified access| MPI
ETP -->|needs| MEDREC
MMP -->|checks interactions| EPIC
MMP -->|needs unified view| MPI
EPIC -->|stores| PATDEMO
CERNER -->|stores| PATDEMO
PROP -->|stores| PATDEMO
EPIC -->|integrates| PACS
EPIC -->|integrates| LIS
EPIC -->|needs| MPI
CERNER -->|needs| MPI
HIPAA -->|requires| AUDIT
AUDIT -->|needs| MPI
AUDIT -->|needs| MIDWARE
GAP1 -->|blocks| MPI
GAP1 -->|affects| ETP
GAP1 -->|affects| OBJ
GAP2 -->|affects| HIPAA
GAP3 -->|affects| OBJ
MIDWARE -->|connects| EPIC
MIDWARE -->|connects| CERNER
MIDWARE -->|connects| MPI
SCHEMA -->|unifies| PATDEMO
```
---
## π PHASE 8 β GAP & INCONSISTENCY ANALYSIS
| Issue | Area | Evidence | Impact | Priority | Action |
|---|---|---|---|---|---|
| **Master Patient Index Missing** | Data / Identity | 3 separate EHRs, no patient matching | Patient safety risk, billing errors, duplicates (15-20%) | **P0** | Design MPI + identity resolution |
| **Data Integration Layer Missing** | Systems | Point-to-point integrations, manual exports | Operational inefficiency, data stale, errors | **P0** | Build integration middleware + APIs |
| **Unified Medical Record Schema Missing** | Data | Each EHR has different data model | Cannot merge records, incomplete history | **P0** | Define FHIR-based schema |
| **HIPAA Audit Trail Incomplete** | Compliance | PACS not fully audited, email communication untracked | Compliance risk, regulatory penalties | **P0** | Implement central audit logging |
| **Disaster Recovery Plan Missing** | Operations | No formal DR procedure documented | Business continuity risk, patient care interrupted | **P0** | Document DR plan + test quarterly |
| **Single Sign-On Missing** | Operations | Staff log in to 3 separate systems | Staff frustration, password fatigue, security risk | **P1** | Implement enterprise SSO |
| **Duplicate Patient Records (15-20%)** | Data Quality | Manual registration, no matching | Billing errors, wrong care, safety risk | **P1** | Deduplicate + implement MPI |
| **Lab Order Workflow Broken** | Operations | Orders emailed between departments, results lost | Patient care delayed, missed diagnoses | **P1** | Build LIS integration + alerts |
| **Imaging Not Centralized** | Operations | PACS local storage only, images not accessible from other hospitals | Cannot access imaging in emergencies, duplicate orders | **P1** | Centralize PACS + cloud backup |
| **Real-Time Bed Tracking Missing** | Operations | Bed management via spreadsheets, updated hourly | Wait times unknown, bed utilization poor | **P2** | Build real-time bed management system |
| **Staff Cannot See Across Hospitals** | Operations | Must manually look up each EHR | Inefficient, error-prone | **P2** | Connect EHRs via MPI + middleware |
| **Patient Portal Not Enterprise-Wide** | Operations | Portal only on Main hospital, 24hr data delay | Patient engagement poor, compliance risk | **P2** | Extend portal to all hospitals + real-time |
| **Emergency Transfer Protocol Undocumented** | Operations | Ad-hoc process, no formal procedure | Patient safety risk, inconsistent care | **P2** | Document protocol + train staff |
| **Medication Interaction Checking Gaps** | Operations | Each EHR checks independently, cannot see other hospital meds | Drug interaction missed across hospitals | **P2** | Build unified medication repository |
| **Integration Standards Not Defined** | Governance | Mix of HL7, CSV, manual, no FHIR | Technical debt, hard to maintain | **P3** | Define FHIR + HL7 standards |
| **Data Governance Policy Missing** | Governance | No clear ownership of patient data, no retention policy | Compliance risk, data quality poor | **P3** | Create data governance charter |
| **Access Control Matrix Incomplete** | Governance | Unknown who can access what data | HIPAA risk, potential breaches | **P3** | Document access rules |
| **Incident Response Procedures Missing** | Governance | No documented breach response process | Regulatory violations, slow response | **P3** | Document incident procedures |
---
## βοΈ PHASE 9 β ARCHITECTURE CONSISTENCY CHECK
| Check | Result | Evidence | Gap |
|---|---|---|---|
| Business Processes β Systems | β MISALIGNED | Patient admitted at 3 different places differently | Standardize admission process |
| Systems β Data | β MISALIGNED | 3 EHRs, 3 patient schemas, duplicates | Build Master Patient Index |
| Emergency Protocol β System Support | β MISSING | No protocol, no unified access | Document + build infrastructure |
| Medication Management β Interaction Checking | β INCOMPLETE | Only checks within 1 hospital | Unify across 3 hospitals |
| Lab Ordering β Result Delivery | β οΈ PARTIAL | Some automated, some manual, email-based | Improve LIS integration |
| Imaging β Accessibility | β INCOMPLETE | PACS local, not accessible from other hospitals | Centralize + cloud backup |
| Compliance Requirements β Audit Logging | β INCOMPLETE | HIPAA requires full audit trail, PACS not fully logged | Complete audit infrastructure |
| Staff Access β Identity Management | β MISSING | 3 separate logins required | Implement SSO |
| Disaster Recovery Planning β Backup Strategy | β οΈ PARTIAL | Backups exist but DR plan undocumented/untested | Document + test DR |
| Data Ownership β Governance | β MISSING | No clear data steward | Assign data governance roles |
---
## π PHASE 10 β OBSIDIAN NAVIGATION SYSTEM
### Main Navigation Hub
```markdown
# π₯ Metropolitan Health System Architecture
## π¨ CRITICAL ISSUES (Read First)
### P0 - MUST FIX NOW
1. **[[Critical Gap - Master Patient Index Missing]]** β START HERE
- 3 EHRs, 1 patient = 3 records = PATIENT SAFETY RISK
- Est. cost: $2.5M-4M, Est. time: 6-9 months
2. **[[Critical Gap - No Data Integration Layer]]**
- Systems talking point-to-point, no central hub
- Est. cost: $1.5M-2.5M, Est. time: 4-6 months
3. **[[Critical Gap - HIPAA Audit Trail Incomplete]]**
- PACS not fully audited, email untracked
- Compliance risk, regulatory penalties
4. **[[Critical Gap - Disaster Recovery Plan Missing]]**
- No formal procedure, no recent testing
- Business continuity at risk
---
## π QUICK ACCESS BY ROLE
### For Chief Medical Officer
- [[Patient Safety Gaps]] - What's at risk?
- [[Emergency Transfer Protocol]] - Do we have one?
- [[Medication Interaction Checking]] - Can we see across hospitals?
### For IT Leadership
- [[System Architecture Overview]]
- [[Integration Technology Stack]]
- [[Infrastructure & Cloud Strategy]]
### For Compliance Officer
- [[HIPAA Compliance Checklist]]
- [[Audit Trail Requirements]]
- [[Data Governance Policy]]
### For Hospital CFO
- [[Architecture Investment Roadmap]] - What will it cost?
- [[ROI Analysis]] - What's the payoff?
- [[Risk Mitigation Priority]] - What's most urgent?
---
## π― BROWSE BY TOPIC
### Business & Clinical
- [[Patient Admission Process]] (3 different procedures)
- [[Emergency Transfer Protocol]] (missing!)
- [[Diagnostic Workflow]]
- [[Medication Management Process]]
- [[Discharge Process]]
### Systems Landscape
- [[Epic EHR - Main Hospital]] (operational)
- [[Cerner EHR - North Campus]] (operational)
- [[Proprietary EHR - Urgent Care]] (legacy)
- [[PACS System]] (imaging, local storage)
- [[Lab Information System]] (LIS)
- [[Patient Portal]] (Main hospital only)
### Data & Master Data
- [[Patient Demographic Data Model]] (3 copies, unlinked)
- [[Medical Record Data Model]] (fragmented)
- [[Master Patient Index Planned]] (CRITICAL gap)
- [[Unified Schema Needed]] (required for integration)
### Technology & Infrastructure
- [[Epic Platform]] - Deployed 2015, next upgrade 2026
- [[Cerner Platform]] - Deployed 2017
- [[PACS Infrastructure]] - On-premise, not centralized
- [[Cloud Strategy]] - Candidate infrastructure
- [[HL7 vs FHIR Standards]] - Which to use?
### Integration & Interoperability
- [[Integration Middleware Planned]] (CRITICAL gap)
- [[HL7 Integration Points]] (current)
- [[FHIR Roadmap]] (future state)
- [[Single Sign-On Strategy]] (needed)
### Operations & Procedures
- [[Bed Management Process]] (spreadsheets!)
- [[Staff Scheduling Procedure]]
- [[Shift Handoff Protocol]]
- [[Incident Response Procedures]] (missing!)
### Governance & Compliance
- [[HIPAA Compliance Framework]]
- [[Data Governance Policy]] (missing!)
- [[Access Control Matrix]] (incomplete)
- [[Data Retention Policy]] (missing!)
- [[Audit Procedures]] (partial)
### Architecture Decisions
- [[ADR-001: Multi-EHR Strategy]] - Keep all 3 or consolidate?
- [[ADR-002: Master Patient Index Approach]] - Centralized or distributed?
- [[ADR-003: FHIR vs HL7]] - Which standard for future?
- [[ADR-004: Cloud vs On-Premise]] - Where will MPI live?
- [[ADR-005: Single Sign-On Solution]] - Which platform?
- [[ADR-006: Disaster Recovery Strategy]] - How to protect patient data?
---
## π― START HERE: 3-Month Priority
**IF YOU ONLY READ 3 PAGES, READ THESE:**
1. [[Critical Gap - Master Patient Index Missing]]
β Why patient safety depends on this
2. [[Enterprise Architecture Roadmap - Year 1]]
β What to build and in what order
3. [[Zachman Matrix]]
β Where this all fits together
---
## π BROWSE BY HOSPITAL LOCATION
### Main Hospital (400 beds)
- [[Main Hospital Profile]]
- [[Main Hospital - Epic EHR]]
- [[Main Hospital - Departments]]
- [[Main Hospital - Bed Capacity]]
### North Campus (150 beds)
- [[North Campus Profile]]
- [[North Campus - Cerner EHR]]
- [[North Campus - Departments]]
- [[North Campus - Bed Capacity]]
### Urgent Care Clinic (30 beds)
- [[Urgent Care Profile]]
- [[Urgent Care - Proprietary EHR]]
- [[Urgent Care - Check-in Process]]
---
## π GLOSSARY
Medical & Technical Terms
- [[Patient]] - Individual receiving care
- [[MPI]] - Master Patient Index (unified identity)
- [[EHR]] - Electronic Health Record
- [[PACS]] - Picture Archiving & Communication System (imaging)
- [[LIS]] - Lab Information System
- [[HL7]] - Health Level 7 (messaging standard)
- [[FHIR]] - Fast Healthcare Interoperability Resources (modern standard)
- [[HIPAA]] - Health Insurance Portability & Accountability Act
- [[HITRUST]] - Health Information Trust Alliance (compliance certification)
```
---
## π PHASE 11 β DATAVIEW COMPLIANCE DASHBOARD
### Query 1: Critical Gaps by Priority
```markdown
---
title: Critical Gaps Dashboard
---
# π¨ Critical Gaps Requiring Action
## P0 - MUST FIX (Do in next 90 days)
```dataview
TABLE priority, impact, estimated_effort, owner
FROM "11 - Gaps & Issues"
WHERE priority = "P0"
SORT priority ASC
```
## P1 - HIGH (Do in next 6 months)
```dataview
TABLE priority, impact, owner, blocked_by
FROM "11 - Gaps & Issues"
WHERE priority = "P1"
SORT impact DESC
```
---
## Compliance Risk Summary
```dataview
TABLE status, affected_systems, regulatory_impact
FROM "12 - Compliance"
WHERE status = "AT RISK"
```
```
### Query 2: System Integration Status
```markdown
---
title: System Integration Status
---
# System Integration Inventory
## Current State
```dataview
TABLE deployed, integration_status, data_quality
FROM "04 - Systems"
GROUP BY vendor
```
## Missing Integrations
```dataview
TABLE FROM "04 - Systems"
WHERE integration_status = "MISSING"
```
## Integration Debt (To Fix)
```dataview
TABLE current_integration, issue, priority
FROM "07 - Integration"
WHERE issue = "fragmented" OR issue = "manual"
SORT priority DESC
```
```
### Query 3: Duplicate Patient Records Tracking
```markdown
---
title: Data Quality - Duplicate Records
---
# Patient Deduplication Status
```dataview
TABLE hospital, estimated_duplicates, data_quality_issue
FROM "04 - Data"
WHERE file.name CONTAINS "Patient"
```
## Impact
- **Estimated Duplicates**: 15-20% of patient database
- **Estimated Affected Patients**: ~67,500 of 450,000
- **Manual Reconciliation Effort**: 6-12 months
- **Business Impact**: $2M/year revenue loss (duplicate billing)
```
---
## π§Ύ PHASE 12 β FINAL ARCHITECTURE PACKAGE
### Executive Summary
**Organization**: Metropolitan Health System (3 hospitals)
**Scope**: Enterprise Healthcare IT Architecture
**Assessment Date**: Today
**Current State**:
- β 3 functional EHR systems (Epic, Cerner, Proprietary)
- β Basic imaging system (PACS)
- β Lab information system (LIS)
- β NO unified patient identity
- β NO data integration layer
- β NO enterprise-wide processes
- β INCOMPLETE compliance infrastructure
- β UNDOCUMENTED emergency procedures
**Critical Risks**:
1. **Patient Safety** - Doctor cannot see complete medical history in emergency
2. **Compliance** - HIPAA audit trail incomplete, regulatory penalties likely
3. **Operational** - Staff inefficiency, manual workarounds, data errors
4. **Financial** - Duplicate billing, wasted imaging/lab tests, revenue loss (~$2M/year)
---
### Recommended Investment (3-Year Roadmap)
| Phase | Timeline | Objectives | Estimated Cost | Business Value |
|---|---|---|---|---|
| **Phase 0** | Months 1-3 | Master Patient Index Planning | $250K | Enables all other work |
| **Phase 1** | Months 4-12 | Master Patient Index Implementation | $3.5M | Patient safety, deduplication |
| **Phase 2** | Months 10-18 | Integration Middleware + APIs | $2M | System connectivity, automation |
| **Phase 3** | Months 16-24 | Unified Medical Record + Schema | $1.5M | Single source of truth |
| **Phase 4** | Months 22-30 | Emergency Transfer Protocols + Training | $800K | Patient safety, compliance |
| **Phase 5** | Months 25-36 | Analytics & Insights Platform | $1.2M | Quality metrics, operational data |
**Total Investment**: ~$9M over 3 years
**Estimated ROI**: 12-18 months (from duplicates eliminated + efficiency gains)
---
### Obsidian Vault Status
β **Vault Structure Created**: 13 folders
β **Notes Generated**: 45+ architecture notes
β **Relationships Defined**: 120+ links between concepts
β **Dataview Queries**: 8 compliance dashboards
β **Mermaid Diagrams**: System integration graphs
β **Gap Analysis**: 18 issues identified by priority
---
## π IMPLEMENTATION ROADMAP (3-Year Plan)
### PHASE 0: Discovery & Planning (Months 1-3)
**Objectives**:
- Assess all patient records across 3 hospitals
- Quantify duplicate rate
- Define MPI requirements
- Secure executive sponsorship + budget
**Key Tasks**:
- [ ] Data inventory from Epic, Cerner, Proprietary
- [ ] Duplicate patient analysis (sample 10%)
- [ ] Vendor RFP for MPI solutions
- [ ] Governance model for patient data ownership
- [ ] HIPAA impact assessment
**Success Criteria**:
- β Board approval for $3.5M MPI investment
- β Duplicate rate quantified
- β MPI vendor selected
- β Governance structure established
---
### PHASE 1: Master Patient Index Build (Months 4-12)
**Objectives**:
- Deploy working MPI across all 3 hospitals
- Eliminate duplicate patient records
- Enable unified patient lookup in real-time
**Key Tasks**:
- [ ] Build MPI database (centralized or cloud)
- [ ] Implement identity resolution engine
- [ ] Connect Epic registration β MPI
- [ ] Connect Cerner registration β MPI
- [ ] Connect Proprietary registration β MPI
- [ ] Deduplicate historical records
- [ ] Train staff on new workflows
- [ ] Go-live with parallel operations
**Success Criteria**:
- β 100% of new patients use MPI
- β 95%+ of existing records deduplicated
- β Duplicate patient records reduced from 15% to <1%
- β Identity lookup < 500ms
- β Staff adoption > 95%
- β Zero patient safety incidents
---
### PHASE 2: Integration Middleware (Months 10-18)
**Objectives**:
- Build central integration layer
- Stop point-to-point integration chaos
- Enable enterprise-wide API ecosystem
**Key Tasks**:
- [ ] Select integration platform (Mulesoft, Boomi, etc.)
- [ ] Design API gateway for all systems
- [ ] Build Epic β Middleware β LIS connector
- [ ] Build Epic β Middleware β PACS connector
- [ ] Implement HL7 to FHIR translation
- [ ] Deploy API management + security
- [ ] Migrate legacy integrations
**Success Criteria**:
- β Central integration hub operational
- β All system integrations go through middleware
- β API response time < 200ms
- β System uptime > 99.5%
- β Reduced integration maintenance effort by 40%
---
### PHASE 3: Unified Medical Record Schema (Months 16-24)
**Objectives**:
- Define standard data model for all hospitals
- Map existing data to standard schema
- Enable record merge and unification
**Key Tasks**:
- [ ] Define FHIR-based data model
- [ ] Map Epic schema β FHIR
- [ ] Map Cerner schema β FHIR
- [ ] Map Proprietary schema β FHIR
- [ ] Implement data transformation rules
- [ ] Build unified patient record view
- [ ] Migrate historical data
**Success Criteria**:
- β Single FHIR-compliant schema in use
- β All patient data accessible in unified format
- β Data quality > 95%
- β Clinician can view unified record < 2 seconds
---
### PHASE 4: Emergency Transfer Protocols (Months 22-30)
**Objectives**:
- Document emergency transfer procedures
- Enable real-time access to patient history
- Train staff on new workflows
**Key Tasks**:
- [ ] Define emergency transfer protocol
- [ ] Build emergency access workflow (override rules)
- [ ] Implement real-time patient summary
- [ ] Train clinical staff (500+ people)
- [ ] Define escalation procedures
- [ ] Document legal liability
**Success Criteria**:
- β Protocol documented + approved
- β Staff trained (100% certification required)
- β Emergency access < 30 seconds
- β Patient history available at receiving hospital
- β Zero patient safety incidents
---
### PHASE 5: Analytics & Insights (Months 25-36)
**Objectives**:
- Enable data-driven decision making
- Measure quality outcomes
- Support clinical research
**Key Tasks**:
- [ ] Build data warehouse (patient outcomes)
- [ ] Implement quality metrics dashboards
- [ ] Enable clinical research data access
- [ ] Build physician performance scorecards
- [ ] Support population health management
**Success Criteria**:
- β Real-time quality dashboards
- β Identify opportunities (e.g., unnecessary procedures)
- β Support clinical improvement initiatives
- β Enable research on patient outcomes
---
---
# β
TEST VALIDATION
## Differences from Sample Test 1 (E-Commerce)
| Aspect | Test 1: E-Commerce | Test 2: Healthcare |
|---|---|---|
| **Domain** | Retail order processing | Clinical patient care |
| **Scope** | Product level (Order Management) | Enterprise level (3-hospital network) |
| **Key Challenge** | API documentation + inventory sync | Patient identity + data integration |
| **Systems Count** | 6 systems | 10+ systems + medical devices |
| **Regulatory Pressure** | Low | **CRITICAL** (HIPAA compliance) |
| **Safety Impact** | Operational delays | **Life and death** decisions |
| **Data Complexity** | Simple order records | Complex medical records (imagery, genotype data, device readings) |
| **Integration Pattern** | Mostly SaaS APIs | Mix of on-premise + SaaS + medical device protocols |
| **Cost of Failure** | Revenue loss | Patient harm / regulatory penalties |
---
## Sample Test 2 Demonstrates
1. **Enterprise Scope** - 3 separate organizations (hospitals)
2. **Complex Data** - Medical records, imaging, lab results, pharmaceuticals
3. **High Stakes** - Patient safety + HIPAA compliance + life-or-death decisions
4. **Real Gap Analysis** - 18 critical gaps identified with business impact
5. **Healthcare-Specific Glossary** - Medical terminology + HL7/FHIR standards
6. **Compliance Dataviews** - HIPAA-specific dashboards
7. **3-Year Roadmap** - Phased enterprise transformation
8. **Executive Summary** - Hospital board-level ROI analysis
---
**Ready for your own scenario?** Provide your 7 inputs and I'll build a complete Zachman architecture! ποΈ
By purchasing this prompt, you agree to our terms of service
CLAUDE-5-OPUS
Enterprise architecture knowledge is often scattered across documents, diagrams, spreadsheets, and disconnected notes, making it difficult to maintain a clear relationship between business requirements, systems, data, processes, and technical implementation.
This prompt transforms AI into a Zachman-Obsidian Bridge Architect that converts Zachman Framework information into a structured Obsidian knowledge system, maps architecture artifacts across perspectives and interrogatives, identifies
...more
Updated 3 weeks ago
