Clinical Observations Interoperability (COI)

1
Clinical Observations Interoperability (COI)
Helen ChenVipul KashyapJyotishman Pathak Eric PrudhommeauxRachel RichessonHolger Stenzhorn
Health Care and Life Sciences Interest Group
(HCLSIG)
2
Motivation

• Clinical trials are conducted to collect safety
  and efficacy data for new drugs or devices
• Depending on the specific trial, investigators
  enroll healthy volunteers or patients
• Finding and recruiting suitable participants is
  important to avoid unnecessary costs

3
Problem

• Hospitals have another standard to encode
  clinical care

• Pharmaceuticalcompanies have onestandard to
  encodeclinical trials
• Waste of money because of two different standards

CDISC SDTM
HL7 DCM
4
Goals

• How can the gap between the two be bridged using
  Semantic Web technologies?
• Map across formats
• Reuse existing data
• Develop a proof of concept application
• Demonstrate the feasibility of the approach
• Get relevant partners involved

5
Use Case Step-Through

• (Textual) specification of the eligibility
  criteria for a given clinical trial
• Ontology-based translation of the eligibility
  criteria into SPARQL queries
• Translation of the SPARQL queries into
  database-specific SQL queries

6
Use Case Step-Through

• Execution of the queries at the database(s)
  results contain all eligible patients
• Return of a list of eligible patients to clinical
  trial administrator

7
Ontology Modeling
8
Ontology Modeling

• Develop ontologies for clinical trials and
  clinical practice based on information models
  CDISC SDTM and HL7 DCM
• Model all necessary patient and medication
  aspects
• Lab results and observations
• Patient vital signs
• Medications

9
Inclusion Criteria

• Type 2 diabetes on diet and exercise therapy or
• monotherapy with metformin, insulin
• secretagogue, or alpha-glucosidase inhibitors, or
• a low-dose combination of these at 50
• maximal dose. Dosing is stable for 8 weeks prior
• to randomization.
• ?patient takes meformin .

10
Exclusion Criteria

• Use of warfarin (Coumadin), clopidogrel
• (Plavix) or other anticoagulants.
• ?patient doesNotTake anticoagulant .

11
Medication in RDF

• M0271 a sdtmMedication splclassCode 6809
  metformin sdtmsubject P0006
  sdtmdosePerAdministration sdtmhasValue
  500 sdtmhasUnit "mg sdtmstartDateTime
  "20070101T000000"xsddateTime
  sdtmendDateTime "2008-0101T000000"xsddate
  Time .

metformin
12
Criteria in SPARQL

• ?medication1 sdtmsubject ?patient
  splactiveIngredient ?ingredient1 .
• ?ingredient1 splclassCode 6809 . metformin
• OPTIONAL
• ?medication2 sdtmsubject ?patient
  splactiveIngredient ?ingredient2 .?ingredient2
  splclassCode 11289 .
  anticoagulant
• FILTER (!BOUND(?medication2))

metformin
anticoagulant
13
Issues

1. Data is not stored only in form of the CDISC SDTM
   but also in HL7 DCM form
2. Data is not stored as RDF but in conventional
   relational databases

14
Solutions

• Transform references to SDTM classes to
  references to HL7 classes via N3 rules
• Transform either
• Relational data into RDF
• SPARQL into SQL via N3 rules
• Second approach chosen

15
SDTM to HL7 Transformation
sdtmMedication sdtmdosePer- Administration


?x a sdtmMedication sdtmdosePer-
Administration ?y gt ?x hl7Substance-
Administration hl7doseQuantity ?y
hl7Substance- Administration hl7doseQuantity
16
Behind the Scenes

• Description of the complete workflow
  athttp//www.w3.org/2008/Talks/
  1027-ISWC/SWObjects