This is a repository for [email protected] to explore concepts about the EpiCast protocol.
A proposed technical solution to for public health programs to share case information with CDC, each other and the public.
EpiCast is a follow-up project to the earlier Case Surviellance Discover sprint that the CDC/USDS conducted to create a NNDSS sytem that was:
- Excellent on new and emerging diseases and conditions
- Supported current standards and facilited the emergence of new standards
- Supported bidirection/multidirectional collaboration between states and cdc
- Built on cloud-first and DMI principles
- Free from syncronization and aggregate total errors
EpiCast supports the construction of such a system with reasonable costs.
- Human and machine readable - Simple enough for humans to read, structured and standardized enough that machines can acess the data.
- Flexible - Data dictionaries to handle data element changes, local and national, that occur in an new outbreak.
- Scalable - Support pandemic scale. From 1 to 1M data items per month.
- Sustainable - Standardized, open-source reference implementations, free tooling and services.
- Secure - Support both public and private feeds.
In broad sense, EpiCast supports the CDC's NorthStar architecture, the open data movement, and the CDC's DMI direction for NNDSS.
- Publish complete data sets instead of sending individual messages.
- States are the source of truth for data set
- Use CSV because it is understood by both epidemiolgist and data scientists.
- Support for supplemental datatypes including FHIR, and HL7 messages.
- Take important ideas from TESSY including validation language and support from aggregates
- Define a schema language
- Validation done on entry by using a computable language (CQL inspired)
- Logging of events in the feed.
- Use S3 Buckets and Azure Blobs as the base protocol because all cloud vendors can support these protocols
cd demo and follow readme.md