Tracking an Outbreak (Ebola, COVID-19 and What Comes Next)
Sector: Ministries of health, public health institutes
The scenario
When a serious outbreak starts, the first thing that fails is not the hospitals. It's the information.
Imagine a public health institute facing a fast-moving outbreak. Clinics and labs report cases by spreadsheet, phone and fax, each with a different form. The case definition changes as the disease becomes better understood: new symptoms, new lab criteria, a different incubation window. The same person can be reported by a clinic, then a lab, then a hospital. Contact tracing lives in a separate tool. Leadership wants the numbers every morning.
What goes wrong without the right tool
- Every change of the case definition means reissuing forms, and facilities adopt it on different days, so counts from different weeks can't be compared.
- The same person appears several times with slightly different spellings, and analysts spend their days deduplicating.
- "Who did this person meet?" and "what are their results?" live in different places.
- Daily numbers arrive days late because someone has to merge the spreadsheets.
- When the numbers are challenged, nobody can show who reported what and when.
How Atomik can be used
- Define the case report once. Symptoms, onset date, exposure, test results and outcome as an openEHR template, shared by all reporting facilities, who send reports through the REST API or through an integration layer.
- Change the definition without breaking history. A new case definition is a new template version. Reports created under the previous version stay valid under it, and queries can be written to span versions when the fields are compatible.
- One identity per person. Reports about the same person from a clinic, a lab and a hospital attach to the same demographic record, and the Master Patient Index capabilities help resolve duplicates.
- Record contacts as relationships. Atomik supports relationships between demographic records, so the exposure network of a person can be stored next to their clinical data.
- Ask the daily questions as stored queries. For example: new confirmed cases by region in the last 7 days, contacts of confirmed cases without follow-up, patients with an open outcome. When the definition changes, edit the query.
- Keep the audit trail. Each report records who submitted it and when.
- Reuse it for the next one. For a new pathogen, define a new case report template and reuse the same identity, relationships and reporting mechanisms.
How CaboLabs can help
Designing the case report template and connecting each facility are modeling and integration work that CaboLabs, the builder of Atomik, can help with: see openEHR Implementation and HIS Integration. Atomik doesn't require CaboLabs services.
What this gives you
A case definition you can change without losing comparability, one connected view of people, exposures and results, and daily questions answered from live data.
Keep in mind
Atomik is the data and query layer. Dashboards, maps, notifications and the reporting workflow with facilities are built around it, and mapping each facility's data is integration work.
Why Atomik fits
In an outbreak the data model changes faster than software can. Atomik lets the public health team change what is collected by changing a definition, keeps old and new reports side by side, and treats people, their exposures and their clinical data as one connected picture.