GLIF – GELLO – Archetypes Jared Davison Medical-Objects Australia
Key Points GLIF and GELLO implementation using Lymphoma as the clinical example Arden was considered but GELLO felt to be better fit –GELLO was Object-Oriented –Interface with data model –vMR rather than use curly braces –Closely related to Object Constraint Language (OCL)
Key Points GLIF chosen as guideline representation language –Inspiration for GELLO –Specific lacking detail, specifically no structures to accommodate GELLO GELLO specifications had not been implemented –Significant errors and lacked required features eg no standard way of passing parameters to GELLO and no formal definitions of GELLO libraries
Key Points Contact with GELLO authors and researchers –Raised issues for inclusion in second ballot of GELLO
Key Points GELLO requires a VMR for access to patient data –Lack of specifications for vMR –Created abstraction of HL7 v2 model –Could be represented as v3 Used Archetype Definition Language (ADL) to define data structures for user data-entry –Encoded in HL7 v2 OBX using an experimental technique –Enables re-use and re-editing of data –Also used for decision support
GLIF Guideline - Lymphoma
Archetype Editor with GELLO Events GELLO Expressions
Lymphoma Archetype Example
Completed Archetype Example
Editing Lymphoma Type Codes
Editing Help Text (Didactics)
Editing Decision Logic (GELLO)
GELLO Compiler IDE
Setting up SNOMED-CT Codes
Lymphoma Scenario Jack Smith age 63, 5 visits/episodes in all for treatment cycles including registration and early data gathering *2 visits, Diffuse large B-cell lymphoma, increased FLIPI, responds to treatment, bulky stage II, work to guidelines, radiotherapy gets a letter, pathology radiology requests and reports, Rx rituximab, CHOP and radiotherapy, and ABCR notifications
Patient Registration
GLIF Guideline Execution
Record Reported Symptoms
Ordering Blood Tests
Clinical Staging Data
Patient Data Summary
Curative Intent Treatment Plan
Example of Supporting Evidence
Cancer Registry Notification