Integrating Quality Data to Improve Care Dan Housman Managing Director and Co-Founder October 20, 2010 Copyright © 2010 Recombinant Data Corp. All rights reserved. 1 Outline/Agenda • About Recombinant • What is driving Clinical Intelligence in healthcare? • Why Recombinant and IBI are partnering • Case study – Piedmont Healthcare • Demonstration • Discussion 2 About Recombinant Best of breed software/analytics company focused 100% on secondary uses of clinical data Core Competencies • Life science data warehousing & integration services • Translational research & quality reporting solutions • Data strategy, governance & compliance consulting • Open Source implementations & extensions Core Values • Pragmatism • Effective Communication • Trust 3 Ecosystem 4 AMC - Personalized Medicine Vision 5 Perfect Storm Opportunity Meaningful Use HITECH/HIPAA Restructuring EHR Adoption (e.g EPIC) Personalized Medicine Accountable Care Organization Why Adopt Now? DW Solution Maturity 10 Years 6 Data Access Needed for Quality Measurement Frameworks PQRI Provider Bonus Program HEDIS/P4P Operations Physician Incentives Balanced Scorecard System Incentives Chronic Disease Mgmt. Preventative Medicine Patient Satisfaction Meaningful Use In Patient/Hospital Ad-hoc queries Report Library Board Reporting Quality Management Core Measures Compliance Six Sigma 7 Data management solution map 8 Why IBI and Recombinant are partnering 9 Why IBI and Recombinant together? Recombinant IBI Capabilities Needs Expertise in healthcare data sets Packaged healthcare modules Packaged data models, reference data (measures), report library Packaged reporting on IBI platform (future) Implementation partners (healthcare focus) Consulting services (governance/strategy) Custom healthcare BI development services Needs Capabilities Packaged development tools Packaged BI software Advanced analytics/ad-hoc data mining Advanced BI (ad-hoc mining) tools Adapters to key source systems Broad usage outside “clinical” Broader capability reach Solution partner/application community Adapter library 10 IBI Recombinant Partnership Information Builders as BI framework Information Builders Recombinant ETL Integration Information Builders / Open Source Integration 11 ACOs/Analytics 12 Bending the curve? ACO models… “In God we trust; all others must bring data.” W. Edwards Deming 2010 2013? Payment Reform 13 Reducing “gaps” from transitions in care Did heart failure visits have sufficient out patient follow-up? Average Cost/Utilization Per Patient Episode Cost 120000 Outpatient 100000 Inpatient Outpatient 80000 60000 40000 20000 0 -8 -7 -6 -5 -4 -3 -2 -1 AP/AR 0 1 Diagnosis 3 4 5 6 7 8 Claims Processing Episode of care Health Need 2 Procedure Results & Outcomes Patient Perception 14 Teams need measures & benchmarks For now what’s “hot” Figure out how to analyze clinancials to reduce costs Reduce 30 day readmits with care co-ordination/six sigma Reduce chronic disease costs for “high risk” (diabetes, CAD) Use groupers that payers use (ETG, MEGGS, Prometheus) or prepare for CMS grouper Join the ACO collaborative led by The Dartmouth Institute and Brookings Near future Sites supporting open benchmarking as it evolves Best practice pathways will evolve from collaborations An open source CMS grouper and open analytics are planned 15 Avoid Reinventing the Wheel 16 Recombinant Goal 17 Data Needed for Typical Report 18 Operations • Denominators (panels, locations) • Practice snapshots 95 400 90 350 85 300 80 Percent 250 200 150 65 60 50 55 0 50 20 03 -1 20 1 04 20 3 04 20 -7 04 -1 20 1 05 20 3 05 20 -7 05 -1 20 1 06 20 3 06 20 -7 06 -1 20 1 07 20 3 07 -7 Avg. Time initialization (hrs) Avg. Time to finalization (hrs) 45 % initialized 72 hr from visit % initialized 120 hr from visit % created total Operations • Visits/utilization 100 75 70 20 03 -1 1 20 04 -3 20 04 20 -7 04 -1 1 20 05 -3 20 05 20 -7 05 -1 1 20 06 -3 20 06 20 -7 06 -1 1 20 07 -3 20 07 -7 hours • EHR adoption Percentage of notes initialized within N hours Visit to note cycle tim e Chronic Disease Mgmt. Preventative Medicine Patient Satisfaction In Patient/Hospital Ad-hoc queries Report Library 19 Chronic Disease Management • Diabetes • Asthma • CAD Operations • CHF Chronic Disease Mgmt. Preventative Medicine Patient Satisfaction In Patient/Hospital Ad-hoc queries Report Library 20 Dashboard Measures 21 Preventative Medicine • Adult/Pediatric immunizations • Cancer screening (Pap, Mammogram, Colonoscopy) Operations • Lab order completion Chronic Disease Mgmt. Preventative Medicine Patient Satisfaction In Patient/Hospital Ad-hoc queries Report Library 22 Recombinant Data Model Overview 23 Data Trust Subject Area Model Interventions Observations Events High Content Vocabulary Equipment Provider Organization Devices Medications Procedures Clinical Decision Support Problems Diagnosis Notes Biomarkers Narratives Labs Orders Incidents Registration Experiments Images Biomaterial Insurance Membership Location Patient Reported Outcomes Appointments Clinical Trials Meta Data Master Data Patient Omics Compiled Finance Internal Analysis Core Measures External Analysis Good Citizenship Claims Billing 24 Data Trust Data Model: Entities 25 Case Study – Piedmont Health 26 Piedmont Clinic – Current State • 4 Acute Care Hospitals • 671 Physicians • 250 Employed Primary Care • 90 Employed Cardiology • Other Employed Specialists • Hospitalists, Transplant Surgeons, CT Surgery • Balance of Independent Practices • No health plan 27 Piedmont Clinic (Brief History) • Traditional PHO formed in 1980s - contracting • Part of larger Atlanta Holding Company • (Promina) • Participated in Full-risk capitation • Left Promina around 2002 • Clinic restructured around Financial Integration Model • Experience with measuring quality essentially mirrored hospitals • Added Patient Satisfaction (Office based) • 20% financial risk 28 Evolution to Clinical Integration • New CEO 2008 • Studied Memorial Hermann • MDs frustrated with quality data (limited to hospital) and financial risk • Required new governance • Adopted PQRI to improve quality both inpatient and outpatient • 3 Quality Specialists “detailed” practices on CPT-II codes 29 Phase I (Jan-June 2010) • Jan 2010 launched new web portal and CIT • Combination of Clinical Integration and Financial Integration • Collect claims data from each MD in clinic and associated CPT II (PQRI) codes • All claims, all patients • 670 docs represent 150+ DIFFERENT practice management systems, Billing clearing houses • 4 technical support + Recombinant to develop “back end” claims loading process, coding metrics, etc 30 Clinical Data Agents 31 Clinical Integration Trust (CIT) – Phase 1 32 Phase II • Add Population Management Dashboards • Integrate data across providers (PQRI, EMR) and hospitals (EPSI) • Had to “map” patients to providers • 2 visits in 2 years • Building risk adjustment model • Adding metrics related to Efficiency and Cost / Utilization • Based on CMS metrics and methodology 33 Clinical Integration Trust (CIT) – Phase 2 34 Results Metric Community Acquired Pneumonia (Hospital) Congestive Heart Failure (Hospital) SCIP (Hospital) Colon Cancer Screening (Ambulatory) Pneumovax age > 65 (Ambulatory) Diabetes BP Control (Ambulatory) % Improvement 78% to 92% since 2007 83% to 90% since 2007 80 to 93% since 2007 (100 pts per yr) +10% since Jan 2010 (958 more pts) +25% since Jan 2010 (50 more pts) +12% since Jan 2010 (700 more pts) 35 Questions/Feedback discussion 36 Thank You Recombinant Data Corp. 255 Washington Street, Suite 235, Newton, MA 02458 Tel: (617) 243-3700 Fax: (617) 243-3788 www.recomdata.com 37