Achieving clinical outcomes improvement through data-driven decision-making and process engineering Seth Bata Director, Clinical & Business Analytics May 6, 2016 Mission Analytics I May 6, 2016 I 1
Key Questions Who is Mission Health? Why change? How do we use analytics to drive improvement? What are our results so far? What s next? Mission Analytics I May 6, 2016 I 2
Mission Health System Based in Asheville, NC 130-year history 10,700 employees and 2,000 volunteers Over 1,000 physicians on staff Six hospitals and numerous outpatient, surgery, and post-acute care facilities Mission Hospital is the flagship, licensed for 763 beds Mission Analytics I May 6, 2016 I 3
Organizational Priorities Culture of excellence Aim: Get each patient to the desired outcome, first without harm, also without waste, and with a great experience for the patient and family Truven Top 15 health system 4 consecutive years Mission Hospital Truven Top 100 hospital 5 consecutive years Mission Analytics I May 6, 2016 I 4
Mission Analytics I May 6, 2016 I 5 Clinical Redesign
Clinical Redesign Population Health Evidence-based Decisions Clinical Programs Mission Analytics I May 6, 2016 I 6
Evidence-based Decisions Scientific Research Global Research Local Evidence Informed Decision Organizational Perspectives Culture Experience & Judgment Patient Preference Patient Needs & Values Mission Analytics I May 6, 2016 I 7
Reengineering through Clinical Programs Organizing and delivering care by disease processes, procedure type, or preventive service Care Process Models are the building blocks, founded in evidencebased best practice Mission Analytics I May 6, 2016 I 8
Decision tree / algorithm of care Care Process Models Education for provider, patient, and family Hard-wire into the EHR workflow Key performance metrics Source: https://intermountainhealthcare.org/ext/dcmnt?ncid=520257347 Mission Analytics I May 6, 2016 I 9
CPM Core Team Physician Data Architect Nurse CPM Quality Improvement Advisor Informaticist Mission Analytics I May 6, 2016 I 10
Mission Analytics I May 6, 2016 I 11 Mission Analytics
Mission Analytics Vision Where are we headed? Data accessibility & integrity Predictions Recommendations Mission Analytics I May 6, 2016 I 12
Complexity of Decisions Functionality Sophistication What is the best course of action? Optimization Strategic How can I influence the future? What are likely future outcomes? Forecasting Predictive What-if Analysis Advanced Analytics Why did this occur? Statistical Analysis Operational Query Drill-down Alerts & Triggers What action is needed? What exactly is the problem? Business Intelligence Ad hoc Reports How many, how often, where? Reporting Tactical Standard Reports What happened? Data Information Insight Sophistication of Intelligence Mission Analytics I May 6, 2016 I 13
Analytics Support Service Mission Analytics functions as a support service for all of Mission Health Mission Analytics provides analytical solutions in the form of systems, tools, training, and consulting services Primary Care Acute Medicine Trauma & Surgery Analytics Finance HR Quality Mission Analytics I May 6, 2016 I 14
Analytics Infrastructure Technology Enterprise Data Warehouse Data Visualization Platform Science Engine People Knowledge Engineer Data Architect BI Developer Data Scientist Process Opportunity Identification Prioritization Implementation Data Governance Mission Analytics I May 6, 2016 I 15
Value Assessment Measuring value is crucial in a resourceconstrained environment Accountability for delivering value drives ownership and engagement 1) Estimate Anticipated Value 2) Measure Actual Value Mission Analytics I May 6, 2016 I 16
Change Request Production Process 1. Idea Use Cases Return On Investment Study Identify High-level Metrics Prototype Mockup Business Requirements 7. Production 2. Scope/Design Approval and Formal Signoff Publish to Production Environment Training Support Administration Outline Potential Future Enhancements Project Scope Specify Functional Requirements Acceptance Criteria Capacity Planning Scope Approval and Formal Signoff Functional Requirements 6. QA Logic Validation Testing / Data Verification to Source Publish to Development Environment Cycle Testing 3. Data Discovery Specify Technical Requirements Feasibility Data Availability Already mapped to EDW? Level of Effort defined Design Specifications Yes Approved 5. App Visualization Development 4. Data Modeling Design Approved Features Frozen No Parking Lot Review with Project Team Develop Review with Architect Team Develop Known Issue, Enhancement Request, or Future Release Evaluate Evaluate & Validate & Validate Mission Analytics I May 6, 2016 I 17
Mission Analytics I May 6, 2016 I 18 Analytics Applications
Bowel Surgery Focused on standardizing the bowel surgery care process model, beginning with preoperative procedures 20 percentage-point increase in bundle compliance 7% decrease in LOS Mission Analytics I May 6, 2016 I 19
Renal Focused on standardizing the renal care process models, including both acute and chronic conditions 13 percentage-point increase in hyperkalemia bundle compliance 0.7 percentagepoint decrease in readmission rate Mission Analytics I May 6, 2016 I 20
Sepsis Focused on standardizing the sepsis care process model, beginning with treatment bundle 12 percentagepoint increase in bundle compliance 3 percentagepoint decrease in mortality rate Mission Analytics I May 6, 2016 I 21
Sepsis 2.2 increase in the odds of survival for septicshock cases Mission Analytics I May 6, 2016 I 22
Mission Analytics I May 6, 2016 I 23 Key Learnings
Clearly articulate vision Engage stakeholders Key Learnings Define ownership and accountability Measure value, and communicate it Iterate Bite-sized, prioritized pieces Adjust Mission Analytics I May 6, 2016 I 24
Mission Analytics I May 6, 2016 I 25 What s Next
What s Next? Continue to develop the Enterprise Data Warehouse Continue to develop the data-visualization platform Build advanced analytics functionality, focused on the continuum of care Mission Analytics I May 6, 2016 I 26