Using a Rule-Based Approach Questionnaires in a FHIR-first Chatbot Application
summary
A FHIR-first chatbot implements a rule engine that listens for questionnaire saves or rule updates, extracts structured resources (Condition, Observation, MedicationStatement) from QuestionnaireResponse via Structured Data Capture, and runs SNOMED ECL match queries to trigger follow-up questions. Rules live as custom FHIR resources, are evaluated on events, and populate dynamic questionnaires without code changes. Production deployment uses the Snowstorm terminology server for synonyms and implicit value sets, is low-code (authors write ECL), and already runs about 90 rules.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 0:27The Situation
- 1:25The Goal
- 2:39RuleBased Approach
- 3:10FireFirst Approach
- 4:39Medical Terminology
- 6:13Assumptions
- 6:43Terminology
- 7:24Medical Rule
- 8:52Create a Rule
- 12:18How it works
- 13:08Additional questions
- 14:57Results
- 15:25Questions Answers
related talks
50mMedical Chatbot and a Physician Rating System
21mUsing FHIR Workflows to deliver personalized health experiences
09mEnhance Healthcare Chatbot Transparency
34mTechnical: Harvesting Trust: Methods for Validating Chatbots in
47mTesting Conversational AI
46mStream-enabled chronic disease management at population scale
30mFHIR in the W.H.O.L.E.: Graphing Clinical Data for Smarter AI Systems
39mAI, FHIR & the future of healthcare data