Leveraging CQL and CDS hook to enhance CDS
summary
A runnable prototype flags type 2 diabetes when any of three computable criteria match: blood glucose >2 g/L, a diabetes-related observation, or a symptom condition. It uses FHIR R4 and CQL compiled to ELM, with CQL libraries stored as Base64 in Library resources and PlanDefinition mapping CQL outputs into card fields. CDS Hooks uses patient-view prefetch plus on-demand FHIR queries and SMART tokens. Implementation is custom, running CQL via open-source Java libraries against an external test FHIR server and returns info cards.
AI-written summary · about
chapters (10)
- 0:00Introduction and speaker intro
- 1:50What is CQL and its purpose
- 7:00CQL architecture and ELM
- 9:09Basic CQL library example
- 14:12Introduction to CDS and CDS Hooks
- 19:58CDS Hooks principles and concepts
- 27:10Building CDS system for diabetes
- 32:57CDS Hooks requests and responses
- 38:24Linking CQL with CDS Hooks
- 41:17Summary and Q&A
