From the Bench All the Way to Bedside Clinical Decision Support: The Role of Semantic Technologies in a Knowledge Management Infrastructure for Translational Medicine Tonya Hongsermeier, MD, MBA Corporate Manager, Clinical Knowledge Management and Decision Support Clinical Informatics R&D Partners Healthcare System

Current State of Translational Medicine 17 year innovation adoption curve from discovery into accepted standards of practice Even if a standard is accepted, patients have a 50:50 chance of receiving appropriate care, a 5-10% probability of incurring a preventable, anticipatable adverse event The market is balking at healthcare inflation, new diagnostics and therapeutics will find increasing resistance for reimbursement

The Volume and Velocity of Knowledge Processing Required for Care Delivery Medical literature doubling every 19 years Doubles every 22 months for AIDS care 2 Million facts needed to practice Genomics, Personalized Medicine will increase the problem exponentially Typical drug order today with decision support accounts for, at best, Age, Weight, Height, Labs, Other Active Meds, Allergies, Diagnoses Today, there are 3000+ molecular diagnostic tests on the market, typical HIT systems cannot support complex, multi-hierarchical chaining clinical decision support Covell DG, Uman GC, Manning PR. Ann Intern Med. 1985 Oct;103(4):596-9

Today’s Health IS Vendor Knowledge Management Capabilities: Knowledge “hardwired” or structured in proprietary modes into applications, not easily updated or shared Little or no standardization of HIT vendors on SNOMED, no shared interface terminologies for observation capture, no standard order catalogues Most EMRs have a task-interfering approach to decision support, sub-optimal usability Knowledge-engineering tools typically edit into transaction, no support for provenance, versioning, life-cycle, propagation, discovery or maintenance Consequently, clinical systems implementations are under-resourced with adequate knowledge to meet current workflow and quality needs Labor of converting knowledge into Clinical Decision Support is vastly underestimated Doesn’t bode well for personalized medicine

Personalized Medicine Knowledge Management and Decision Support Intersection Points in Translational Medicine Diagnostic Test ordering and documentation guidance Patient Encounter (direct care or clinical trial) Clinical Trials Referral Personalized Medicine Decision Support Knowledge Repository Structured Test Result Interpretations Tissue-bank Therapeutic Intervention Ordering and documentation guidance Knowledge Discovery, Acquisition & Management Structured Research Annotations Bench R&D Integrated Genotypic Phenotypic Databases Clinical Trials 1- 4 Pharmacovigilance

Composite Decision Support Application: Diabetes Management Guided Data Interpretation Guided Observation Capture Guided Ordering

What healthcare needs from Semantic Web Technologies… Reduce the cost, duration, risk of drug discovery Data integration, Knowledge integration, Visualization Knowledge representation  New Knowledge Discovery Reduce the cost/duration/risk of clinical trial management Patient identification and referral Trial design (ie to capture better safety surveillance) Data quality and clinical outcomes measurement Post-market surveillance Reduce the cost/duration/liability of knowledge acquisition and maintenance for clinical decision support and clinical performance measurement Knowledge provenance and representation Conversion of “discovery algorithms” into “clinical practice algorithms” Event-driven change management and propagation of change

KM for Translational Medicine: Functional/Business Architecture R&D DIAGNOSTIC Svs LABs CLINICAL TRIALS CLINICAL CARE PORTALS LIMS EHR APPLICATIONS ASSAYS ANNOTATIONS DIAGNOSTIC TEST RESULTS ASSAY INTERPRETATIONS ORDERS AND OBSERVATIONS KNOWLEDGE and WORKFLOW DELIVERY SERVICES FOR ALL PORTAL ROLES Genotypic Phenotypic DATA REPOSITORIES AND SERVICES State Management Services Semantic Inferencing And Agent-based Discovery Services KNOWLEDGE ACQUISITION AND DISCOVERY SERVICES Data Integration Semantics and ontology based approaches Across multiple data sources: Genotypic: LIMS Phenotypic: EHR Actionable Decision Support Semantic Inferences Management of Patient State Knowledge Acquisition, Maintenance and Evolution Ontology-based Definitions Management Ontology Change and Versioning Impact on Knowledge Assets Knowledge Asset Management and Repository Services Workflow Support Services Data Analysts and Collaborative Knowledge Engineers Collaboration Support Services Logic/Policy Domains Meta- Knowledge Knowledge Domains Data Domains NCI Metathesaurus, UMLS, SNOMED CT, DxPlain, ETC other Knowledge Sources Decision Support Services INFERENCING AND VOCABULARY ENGINES

Example: Diabetes Epidemic, associated with obesity Quality measures drive reimbursement of hospitals and physicians Maintain HbA1c <7 (diet, oral agents and/or insulin) If Renal Disease and no contraindication, should be on ACE inhibitor or ARB If lipid disorder and no contraindication, should be on a Statin National problem of non-compliance with these standards of care  compromised patient longevity, quality of life, ability to maintain employment  CMS and employer financial risk

Imagine this CDS Rule: If Renal Disease and DM and no contraindication, should be on ACE inhibitor or ARB Renal disease = Chronic Renal Failure Nephropathy, chronic renal failure, end-stage renal disease, renal insufficiency, hemodialysis, peritoneal dialysis on Problem List (SNOMED) Creatinine > 2 Calculated GFR < 50 Malb/creat ratio test > 30 Diabetes Many variants on the problem list On Insulin or oral hypoglycemic drug Contraindication to ACE inhibitor Allergy, Cough on ACE on adverse reaction list, or Hyperkalemia on problem list, Pregnant (20 sub rules to define this state) K test result > 5

Composite Decision Support Application: Diabetes Management Guided Data Interpretation Guided Observation Capture Guided Ordering

The Maintenance and Propagation Challenge… These “complex” definitions must be identical in rules (if that is how “recognition” is handled), documentation templates for structured data capture, and in reporting systems that drive payor reimbursement The rate of change for contraindication definition today is slow, yet clinical decision support systems are not keeping up… When molecular diagnostics take off, this rate of change could be “daily” or “hourly” Further, when a patient has a molecular diagnostic test result in the EMR that is currently of “unknown significance” and later, with new knowledge, the interpretation of the former result is “contraindication” to a drug, then this “interpretation” must be updated to ensure proper CDS functioning…

Role of Semantic Technologies… Data/Knowledge Integration and Visualization Ontology based approaches Integration across multiple data sources: Genotypic/Phenotypic data from LIMS/HER Knowledge Repositories for data interpretation Clinical Decision Support Inference engines - SWRL Description logics for “recognition” - OWL Knowledge representation Etc. Knowledge Acquisition, Maintenance and Evolution Ontology-based Definitions Management Versioning, life-cycle, propagation into “dependent” objects such as rules, templates orders/documentation, reporting systems Knowledge Provenance Reconciling knowledge representation among different stakeholders (care givers, payors, performance measurement, clinical trials, R&D)

Market Drivers Will Make Semantic Web Technologies an Imperative for Translational Medicine Genomics: personalized medicine will require decision support architectures that can proactively support complex decision making – answering 1,000,000 of questions before run-time These systems will require self-adaptive, machine learning modes of knowledge acquisition, purely human dependent knowledge acquisition will not scale Pharma will need cooperative relationships with HIT vendors to make speed the translational medicine life-cycle