GEM & HL7 Richard N. Shiffman, MD, MCIS, Abha Agrawal, MD, Roland Chen, MD, Bryant Karras, MD, Luis Marenco, MD, Kristi Polvani, BS, Sujai Nath, MD Peter Gershkovich, MD, Aniruddha Deshpande, MD Yale Center for Medical Informatics NOT!! http://ycmi.med.yale.edu/GEM richard.shiffman@yale.edu Use And Satisfaction 8 physicians—not members of the GEM development team (UNC, UAB, Hopkins, Yale) marked up a guideline CONCLUSIONS: Subjects were able to model the content of the guideline using GEM elements. “satisfactory” Improved editing tools would facilitate translation Result: GEM Cutter Karras, Proc AMIA 2001 GEM-Q / GEM-Q OnLine XSL stylesheet extracts info relevant to quality appraisal from GEMified gl Pass to Shaneyfelt and Cluzeau instruments Output is a quality report card Valued by AAP in gl devel process Available for ad hoc reports on WWW Agrawal, Medinfo 2002 GEM to Arden Relevant components for Arden extracted and used to pre-populate MLMs Agrawal, Proc AMIA 2002 Implementation GEMified document can dynamically generate data collection screens and trigger appropriate recommendations based on guideline logic Proof of concept Applied to NHLBI asthma guideline and CDC TB screening guideline Gershkovich, Proc AMIA 2002 Knowledge extractor XSL extracts and formats guideline info relevant to implementation In use ~100 guidelines have been GEMified Groups in US, UK, Germany, Italy, and NZ are using GEM Funded by NLM: To improve the quality and implementability of an AAP guideline w/ feedback during development. (Using GEM-Q) To create tools that transform GEM-encoded guidelines into CDSS. A generic process and software tools will be developed to translate GEM-encoded guidelines into systems that can improve the process of care. To extend and refine the GEM model to serve as a precise, comprehensive, and consistently applied ontology of guideline-related concepts. (Logic, link, algorithm elements; application of advanced X-technologies) Logical Analysis with Highlighters Recommendation 3 If an infant or young child 2 months to 2 years of age with unexplained fever is assessed as being sufficiently ill to warrant immediate antimicrobial therapy, a urine specimen should be obtained by SPA or bladder catheterization; the diagnosis of UTI cannot be established by a culture of urine collected in a bag. (Strength of evidence: good) Urine obtained by SPA or urethral catheterization is unlikely to be contaminated... UTI Recommendation in XML <decision.variable id=dv1>age</decision.variable> <value>2 months to 2 years</value> <decision.variable id= dv2>unexplained fever </decision.variable> <decision.variable id=dv3>sufficiently ill to warrant immediate antimicrobial therapy </decision.variable> <action id=a1>obtain urine specimen by SPA</action> <action id=a2>obtain urine specimen by catheterization</action> <reason>the diagnosis of UTI cannot be established by a culture of urine collected in a bag</reason> <evidence.quality>Good</evidence.quality> <logic>IF (dv1=2m-2y) AND dv2 AND dv3 THEN a1 OR a2</logic> <link>after: Recommendation 2</link> <link>Diagnosis section</link> Adding guideline meta-information • Operationalizing abstract constructs Sufficiently ill to warrant immediate antimicrobial therapy or Febrile • Interactive Tolerating oral fluids Determining when to collect data, when to deliver advice (site-specific) Whither GEM in HL7 • GEM users asking why HL7 is creating a new architecture HS Identity Developer Purpose Audience Method Knowledge Testing Revision INF Title Citation Release Date Availability Contact Status Companion Document Adaptation Developer Name Committee Name Funding Endorser Comparable Guideline Health Practices Category Target Population Rationale Objective Available Options Implementation Strategy Health Outcomes Exceptions Care Setting Clinician Users Evidence Collection Evidence Time Period Evidence Grading Combining Evidence Specification of Harm/Benefit Quantification of Harm/Benefit Value Judgment Patient Preference Qualifying Statement Cost Analysis Recommendation Conditional (decision variable) . Action . Logic . Reason . Strength of Recommendation . Evidence Quality . Cost . Certainty . Algorithm Eligibility Definition External Review Pilot Testing Expiration Date Scheduled Review . . GEM: Distinguishing Characteristics Conceived and built in XML Multi-platform Open standard Human-readable yet can be processed by machine DTD/schema allows file validation Markup can be performed by non-programmers … GEM passed balloting as a standard (ASTM E2210-02) Goals Comprehensive – capable of expressing all the knowledge contained in guidelines. Health service models cannot express recommendations in sufficient detail; informatics models inadequate to model constructs that express and support guideline validity Goal 2 Expressively adequate to convey the complexities and nuances of clinical medicine while remaining informationally equivalent to the original guideline; tagged elements store actual language Goal 3 Flexible – must be able to deal with variety and complexity of guidelines; permit modeling at high and low levels of granularity Goal 4 Comprehensible – the model should match the stakeholder’s normal problem-solving language and allow domain experts to describe their knowledge with little effort; markup should not require a background as a programmer Goal 5 Shareable across institutions Goal 6 Reusable - across all phases of the guideline lifecycle GEM: Major Components Guideline Document Header Identity Developer Document Body Purpose Method of Development Intended Audience Testing Target Population Revision Plan Knowledge Components Recommendation Pilot Testing Target Popul’n Intended User Method of Dev Developer Identity Unit of implementability Identity Identity Title Citation Length Release Date Availability Electronic Print Status Contact Companion Adaptation Document Patient Resource Developer Developer Developer Name Developer Type Committee Name Committee Expertise NGC Controlled Vocabulary Committee Member Member Expertise Funding Endorser Comparable Guideline Purpose Purpose Main Focus Category Rationale Objective Available Option Implem’n Strategy Health Outcome Exception Intended Audience Intended Audience User Clinical Specialty Professional Group Care Setting Method of Development Method of Development Descrip’n Evid Evidence Time Collection Period Method Evid Collect Number Source Docs Method Descrip’n Spec’n Quant Role Role Cost Evidence Evidence Harm Harm Value Pt Anal Grading Combinat’n Benefit Benefit Judgmt Pref Rating Scheme Method Evidence Combinat’n Qualifying Statement Target Population Target Population Eligibility Inclusion Criterion Exclusion Criterion Age Sex Testing Testing External Review Review Method Pilot Testing Revision Plan Revision Plan Expiration Scheduled Review Knowledge Components Knowledge Components Recommendation Conditional Imperative Algorithm Definition Term Term Meaning Action Step Condit’l Step Branch Step Sync Step Conditional Knowledge Components Recommendation Conditional Dec Var Value Action Reason Dec Variable Descripn Sensitivity Evid Quality Test Param Specificity Recmdn Strength Dec Var Cost Flexblty Action Benefit Predictive Value Action Risk Harm Logic Cost Action Descripn Link Ref Action Cost Certainty Conditional Knowledge Components Recommendation Conditional Dec Var Action Value Sensitivity Reason Dec Variable Descripn Specificity Evid Quality Test Param Recmdn Strength Dec Var Cost Predictive Value Flexblty Action Benefit What Action Risk Harm How Much Logic Cost Action Descripn Where Link Ref Action Cost When Who Certainty Actions What Medication Lab test Procedure Consultation Pt Education Disposition How (much) Where When Who GEM Cutter Knowledge customization • • Add meta-information necessary for implementation, e.g. • Identifier, clinical source, interface, prompt, mechanism of actions Local adaptation • Translation of national recommendations into systems that operate at a local level • Must account for legitimate variations in clinical settings, populations served, and resources available • Danger: protection of professional habit or economic self-interest Strengths of GEM Hierarchy is relatively intuitive Elements are derived from published models Value-added applications have been developed Stable >1 year Designation as a standard GEM Limitations Has been “frozen” > 1 year Not comprehensive (as demonstrated by CPGA) Need guidelines for extension GEM file only as good as guideline document Requires training to use correctly Need to develop <link>, <logic>, and <algorithm> elements