This project implements a Public Health Emergency Surveillance System using virtual machines (VMs) configured with OpenEMR and a centralized HAPI-FHIR server for real-time data sharing. It is designed to simulate how regional hospitals can integrate Electronic Health Records (EHRs) with a central Health Information Exchange (HIE) for public health monitoring and disease outbreak management.
Use Case: Communicable Disease Surveillance across multiple hospital nodes in Michigan’s Upper Peninsula region.
- Deploy a virtualized hospital network architecture for health data interoperability.
- Configure EHR systems (OpenEMR) for each hospital node.
- Set up a centralized FHIR-based HIE (HAPI FHIR Server) using Docker.
- Enable RESTful communication using FHIR standards between hospitals and HIE.
- Facilitate real-time surveillance dashboards using Google Looker.
| Layer | Technology/Tool |
|---|---|
| Virtualization | VMware vCenter Cluster (MTU CCC) |
| EHR Systems | OpenEMR (v6.0.0) |
| Interoperability | HAPI FHIR Server (Dockerized) |
| API Testing | Postman, Swagger UI |
| Public Health Dashboards | Google Looker |
| OS | Ubuntu Server |
| Networking | Custom VLAN (pv-stu-v3XXX) |
| IP Addressing | Static IPs per VM |
The architecture includes four hospital nodes and one central HIE node, simulating the data flow for public health reporting.
- Each hospital runs OpenEMR on a dedicated VM.
- HL7 messages are transformed into FHIR resources.
- FHIR data is pushed to the UPHIE (HAPI-FHIR Server).
- Data is visualized via dashboards using Google Looker.
| Hospital / HIE | Software | IP Format | HAPI-FHIR | OpenEMR | Pinged All VMs |
|---|---|---|---|---|---|
| Aspirus | OpenEMR | 192.168.17.218 | ❌ | ✅ | ✅ |
| Portage Health | OpenEMR | 192.168.17.219 | ❌ | ✅ | ✅ |
| BCMH | OpenEMR | 192.168.17.220 | ❌ | ✅ | ✅ |
| MGH | OpenEMR | 192.168.17.221 | ❌ | ✅ | ✅ |
| UPHIE (Central HIE) | HAPI FHIR | 192.168.17.222 | ✅ | ❌ | ✅ |
- 🛠️ OpenEMR Installation using LAMP stack on Ubuntu
- 🐳 HAPI-FHIR Setup using Docker with custom
application.yaml - 🌐 REST API Integration via Postman (CRUD on FHIR resources)
- 📊 Google Looker Dashboards for disease outbreak tracking
- 🔄 Full FHIR Interoperability between VMs
- Verified REST API endpoints with Postman.
- Confirmed VM-to-VM network connectivity using
ping. - Successfully performed CRUD operations on FHIR Practitioner resources.
- Confirmed UI access to both OpenEMR and HAPI-FHIR interfaces.
Displays regional concentrations of reported COVID-19 cases across Michigan and surrounding areas.
Compares true/false COVID-19 case flags across counties including Marquette, Houghton, Baraga, and Calumet.
- Docker networking issues during initial HAPI FHIR deployment.
- Manual configuration of Apache for OpenEMR across multiple VMs.
- Ensuring all nodes could successfully ping each other over the private VLAN.
- Syncing FHIR formats between OpenEMR and HAPI-FHIR endpoint.
The following Jupyter notebook was used to combine and preprocess data received from all OpenEMR nodes before sending it to Looker:
- Merges data across VMs
- Performs basic cleansing and feature engineering
- Outputs structured datasets ready for dashboard ingestion
- Achieved successful integration of a multi-node EMR system with a central HIE.
- Enabled secure and standardized data sharing via the FHIR protocol.
- Visualized public health data with cloud-based dashboards.
- Gained hands-on experience with healthcare interoperability standards and architecture.
- Aakash Walavalkar
This project is part of SAT5424 (Spring 2025) at Michigan Technological University. For academic use only.