1 / 22

Clinical Observations Interoperability (COI): How can Semantic Web Technologies Help?

This article explores how semantic web technologies can help in achieving clinical observations interoperability in the healthcare industry. It discusses the current and goal state of the HCLS ecosystem, use cases and functional requirements, and the role of semantics in data content, exchange, and computable protocol specification.

cmcdermott
Download Presentation

Clinical Observations Interoperability (COI): How can Semantic Web Technologies Help?

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Clinical Observations Interoperability (COI):How can Semantic Web Technologies Help? Vipul Kashyap* vkashyap1@partners.org Senior Medical Informatician, Clinical Informatics R&D Partners Healthcare System http://esw.w3.org/topic/HCLS/ClinicalObservationsInteroperability Clinical Observations Interoperability Session, HCLSIG Face to FaceNovember 8, 2007 Cambridge, MA *Acknowledgments: Members of the HCLSIG/COI Task Force

  2. Outline • Healthcare and Life Sciences (HCLS): A Taxonomy • HCLS Ecosystem: Current and Goal State • Use Cases and Functional Requirements • The Role of Semantics: • Data Content • Data Exchange and Interoperability • Computable Protocol Specification • Data Retrieval • Conclusions • Coming Attractions! • Next Steps

  3. Healthcare and Life Sciences: A Taxonomy Public Health Research Biological Translational Medicine Biosurveillance Clinical Research Practice

  4. HCLS Ecosystem: Current State Characterized by silos with uncoordinated supply chains leading to inefficiencies in the system Patients, Public Patients FDA National Institutes Of Health Center for Disease Control Pharmaceutical Companies Hospitals Payors Clinical Research Organizations (CROs) Universities, Academic Medical Centers (AMCs) Hospitals Doctors Biomedical Research Clinical Practice Patients Patients Clinical Trials/Research Clinical Practice

  5. HCLS Ecosystem: Goal State NIH (Research) FDA CDC Pharmaceutical Companies Universities, AMCs Patients, Public CROs Hospitals Doctors From FDA, CDC Payors Clinical Observations Interoperability will be a Critical Enabler to realize this Vision!

  6. Use Cases and Functional Requirements • X identifies the Use Cases, Systems and Functional Requirement under consideration of the • COI Task Force • Based on the Functional Requirements Specification developed by EHRVA/HIMSS • Available at http://spreadsheets.google.com/ccc?key=pINNryLt_vyDiPyHj11WiDg&hl=en • on Google Spreadsheets

  7. Use Case: Patient Screening - - Research Coordinator selects protocol for patient Clinical Research Protocol screening: Eligibility Criteria: - Inclusion - Exclusion EMR DATA Meds Procedures Research Diagnoses Demographics Coordinator views list of patients and selects which ones to approach in person for … Patient MR # Potentially # Criteria Criteria #1 No Criteria #2 Criteria #3 Eligible for Met / Total (Pass/Fail/ (Pass/Fail/ (Pass/Fail/ evaluation and Protocol Criteria in Researcher Researcher Researcher recruitment. Protocol Needs to Needs to Needs to Evaluate) Evaluate) Evaluate) … 0011111 Yes 6/8 criteria Pass Pass Pass met … 0022222 No 3/8 criteria Pass Fail Pass met Clinical … 0033333 Yes 5/8 criteria Pass Pass Fail Evaluation and met Recruitment … … … … … … … Focus of Next Talk by Rachel Richesson

  8. Role of Semantic Web Technologies: Examples • Data Content • Precise, Generalized and Extensible Specification • Shareable Open Source Models of Clinical Data • Data Exchange and Interoperability • Re-use and alignment of independently developed Industry Standards • Multiple Ways of saying the same thing • Computable Protocol Specification • Precise and Unambiguous Specification • Approximate Matching • Data Retrieval • Ability to identify relevant patients in the absence of explicit data about a clinical condition

  9. Precise, Generalized and Extensible Specifications A Systolic Blood Pressure measurement is a pressure measurement that has a value from 0 to 220 and units are mmHg SystolicBloodPressureMeasurement equivalentClass ClinicalElement that key value “SnomedCodeForSystolicBP” and magnitude only float[>= 0.0, <= 220.0] and units value mmHg A Sitting Systolic Blood Pressure Measurement is a systolic blood pressure measurement taken when a patient is sitting SittingSystolicBloodPressureMeasurement equivalentClass SystolicBloodPressureMeasurement that bodyPosition value Sitting

  10. Shareable Open Source Models of Clinical Data Clinical Observations Clinical Observations • Open • Source • Clinical • Models • DCM • SDTM • BRIDG • Snomed • MedDRA • NCIT • ….. Clinical Trial 1 Healthcare Provider 1 Clinical Trial 2 Healthcare Provider 2 … … Clinical Trial M Healthcare Provider N Focus of Talk by Tom Oniki

  11. Re-use and Alignment of Independently developed Industry Standards SDTM: SystolicBloodPressureMeasurement equivalentClass VSTEST that VSTESTCD value “SYSBP” and VSORRESU value mmHg DCM: SystolicBloodPressureMeasurement subClassOf key value “SnomedCodeForSystolicBP” and magnitude only float[>= 0.0, <= 220.0] and units value HL7:PQ:mmHg Alignment/Mappings: DCM:SystolicBloodPressureMeasurement equivalentClass SDTM:VSTEST that VSTESTCD value “NCITCodeForSYSBP” DCM:key equivalentProperty SDTM:VSTESTCD DCM:units equivalentProperty VSORRESU DCM:magnitude equivalentProperty VSSORRES SDTM:SYSBP sameAs “NCITCodeForSYSBP” “SnomedCodeForSystolicBP” sameAs “NCITCodeForSYSBP” HL7:PQ:mmHg sameAs VSORRESU:mmHg

  12. Enabling Clinical Observations Interoperability “T1” “T2” “Mr. X” “Mr. X” name name recording_time recording_time systolicBP systolicBP Patient (id = URI1) SystolicBP Measurement1 Patient (id = URI1) SystolicBP Measurement2 magnitude VSORRES key VSTESTCD 120 130 units VSORRESU SnomedCodeForSystolicBP NCITCodeForSYSBP mmHg mmHg EMR Data Clinical Trials Data

  13. Enabling Clinical Observations Interoperability “mmHg” “NCITCodeForSYSBP” “T2” “Mr. X” recording_time name SystolicBP Measurement2 systolicBP Patient (id = URI1) magnitude “T1” 130 recording_time SystolicBP Measurement1 magnitude 120

  14. Merged Patient Data <http:URIForPatient> rdf:type DCM:Patient DCM:name “Mr. X” DCM:systolicBP DCM:systolicBPMeasurements DCM:systolicBPMeasurements rdf:type Collection DCM:units mmHg DCM:key “SnomedCodeForSystolicBP” rdf:_1 _:SystolicBPMeasurement1 rdf:_2 _:SystolicBPMeasurement2 _:SystolicBPMeasurement1 DCM:recordingTime T1 _:SystolicBPMeasurement1 DCM:magnitude 120 _:SystolicBPMeasurement2 DCM:recordingTime T2 _:SystolicBPMeasurement2 DCM:magnitude 130

  15. Multiple ways of saying the same thing! Patient Record 1 Name = Mr. X Recording Time = T1 Weight = 70 Kg WeightType = “Dry” Patient Record 2 Name = Mr. X Recording Time = T2 DryWeight = 70 Kg Are these two weight measurements comparable? Assert the following Mapping: DryWeight equivalentClass Weight that type value “Dry” The OWL Ontology Engine will infer the rest.

  16. Precise Protocol Specification Prophylactic Irradiation to the Contralateral Breast for BRCA Mutation Carriers Undergoing Treatment for Breast Cancer http://www.clinicaltrials.gov/ct/show/NCT00496288?order=2 Ages Eligible for Study:  30 Years   -   90 Years, Genders Eligible for Study:  Female Inclusion Criteria: • Female patient diagnosed with stage I-III breast cancer (AJCC 6), undergoing breast irradiation as part of her adjuvant therapy. • The patient must be a carrier of a deleterious mutation in BRCA 1/2. • … Exclusion Criteria: • Metastatic breast cancer. • Previous irradiation of the breast or chest wall. • Pregnancy. • Patients with active connective tissue diseases are excluded due to the potential risk of significant radiotherapy toxicity. • …

  17. Precise Protocol Specification Patient that hasAge only float[>=30, <=60] and hasGender value “Female” and hasDiagnosis some StageI-IIIBreastCancer and hasTherapy some BreastIrradiation and hasMutation some (mutationType value deleterious and mutationGene value BRCA1/2) andnot (hasDiagnosis some (BreastCancer that cancerType value Metastatic)) andnot (hasTherapy some Irradiation that hasLocation some (Chest or BreastWall)) andnot (hasCondition some Pregnancy) andnot (hasDisease some (Disease that hasLocation some ConnectiveTissue and type value Active)) … Focus of Next Talk by Chintan Patel

  18. Approximate Matching • Subsumption reasoning helps match patients against BreastIrradiationTherapy and it’s subclasses as opposed to a String match with “BreastIrradiationTherapy”. • Systematic query weakening by dropping logical conditions can be used to identify more patients who can later be filtered based on further testing. • “Many ways of stating the same thing”: Infer eligibility based on patient data that doesn’t explicitly assert data satisfying a logical condition.

  19. Identify Patients in the absence of explicit data about a clinical condition Patient Record 1 Name = Mr. X Disease = Hematology Disorder Patient Record 2 Name = Mr. Y Hypereosinophily = 2.0 Lymphosytosis = 6 Blood Lymphocytes = ATypical Which of these two patients have Hematology Disorder? Assert the following Mapping: PatientWithHematologyDisorder equivalentClass Patient that hypereosinophily only float[>1.5] and lymphosytosis only float [> 5] and bloodLymphocytes value ATypical The OWL Ontology Engine will infer the rest.

  20. Conclusions • Critical Need to Align Data, Information and Knowledge across the HCLS Ecosystem • Clinical Observations Interoperability is a critical enabler • Need for: • Open source sharable web based specifications and technologies • Semantic Technologies • Initial Analysis seem to suggest the feasibility and applicability of Semantic Web technologies.

  21. Coming Attractions! • Patient Recruitment Use Case: • Rachel Richesson • Open Source Detailed Clinical Models: • Tom Oniki • Demo: Semantic DB System • Chimezie Ogbuji • Demo: SHER System • Chintan Patel

  22. Next Steps There is a critical need to develop a collaborative framework of various stakeholders such as healthcare providers, pharmaceuticals and IT vendors to address this important problem. Goal: To build consensus and seek participation (time, resources, money!) of these stakeholders to develop a POC that demonstrates feasibility and validates the value proposition. Brainstorming Session: 4:30pm – 5:30pm

More Related