Population Health in the Accountable Care Environment Thomas H. Lee, MD Network President, Partners HealthCare System Professor of Medicine, Harvard Medical School Associate Editor, New England Journal of Medicine May 23, 2013
- 2 - Health Care Reform Unfolds in Three Phases 1. Insurance reform who gets covered, where the money comes from 2. Payment reform how that money is paid to providers 3. Delivery system reform how the care is delivered
- 3 - Health Care Reform Unfolds in Three Phases 1. Insurance reform who gets covered, where the money comes from 2. Payment reform how that money is paid to providers 3. Delivery system reform how the care is delivered Those second and third phases are getting underway They will be disruptive: They will create pressure for consolidation and integration Hospital-hospital Hospital-physician Payer-provider
4 PCHI and Partners HealthCare System, Inc. Dana-Farber/ Partners Joint Venture Partners HealthCare System, Inc. Newton- Wellesley Health Care System, Inc. Brigham and Women s/ Faulkner Hospitals, Inc. The Massachusetts General Hospital North Shore Medical Center, Inc. Partners Community HealthCare, Inc. Affiliated Pediatric Practices, Burlington Medical Associates, Cape Ann Medical Center, Cambridge Health Alliance, Charles River Medical Associates, Emerson PHO, Hallmark Health System, Hawthorn Medical Associates, Pentucket Medical Associates, Plymouth Medical Group, Prima Care, TriCounty Medical Associates Partners Clinical Performance Management
20 Key Tactics and Top Immediate Focuses Longitudinal Care Episodic Care Primary Care Specialty Care Hospital Care Access to care Patient portal/physician portal Extended hours/same day appointments Expand virtual visit options Optimize site of care Reduced low acuity admissions Defined process standards in priority conditions (multidisciplinary teams, registries) Design of care High risk care management Required patient decision aids Re-admissions Hospital Acquired Conditions 100% preventive services Appropriateness Hand-off standards Continuity visit EHR with decision support and order entry Incentive programs (recognition, financial) Measurement Variance reporting/performance dashboards Quality metrics: clinical outcomes, satisfaction Costs/population Costs/episode 5
Key Strategic Payment Decisions 2011-2012 Reopen Existing Commercial Contracts Reduction in Contracted Rate Increases ($345M/four years) Shared Savings in Primary Care Populations Work toward Bundled Payments for Referral Populations Pioneer ACO agreement with CMS Merged with Medicaid MCO/payer (~230,000 lives) Small Commercial Population Exchange experience 6
3 Phases of Work for Improving Population Health Phase 2 Phase 3 Phase 1 1 Primary care: The hub for managing populations: preventive services, chronic illness, high risk 2 3 Specialty care: Where a large fraction of costs are incurred, especially in commercial populations Patient engagement: Involving patients in better self-management of care 4 Wellness Promotion Programs to prevent or delay the progression of illness Ongoing: IS, analytics and central infrastructure 7
Why We Get Stuck We don t have clarity on what we are trying to accomplish We don t have a performance framework We don t have data on outcomes that matter We don t have data on costs We are not organized into functional teams We are not paid to improve value A Path Forward: 1. Porter Strategy 2. Gawande Tactics 3. Bohmer Operations - 8 -
- 9 - Three Colleagues and Key Themes From Their Work Michael Porter, Harvard Business School Value as defined by patient outcomes and costs as an organizing strategic framework for health care delivery organizations Atul Gawande, Brigham and Women s Hospital Checklists that define key processes for which clinician teams should be reliable Richard Bohmer, Harvard Business School Teams that permanently own the work of improvement
- 10 - Porter s Value Framework Ideas Whose Time Has Come Fundamental question What are we trying to do? Clinicians and other stakeholders need shared overarching goal Otherwise, interactions devolve into gamesmanship with goal of protecting each stakeholders interests Traditional concepts of quality put focus on clinicians reliability in performing processes Our goal must be more than defending ourselves from criticism Goal has to be one that invokes problem solving for issues of patients Our shared goal should be improvement of value as defined by the outcomes that matter to patients and costs over meaningful episodes of care
- 11 - Care Redesign Deliverables Phase I Value Dashboard Care Redesign Plan Version 1.0: outcomes, processes, service metrics, and cost available currently Version 2.0 (Future Aspirations): measures that matter to patients, e.g., outcomes Care Plan: pause points and key interventions Incentives: bundled payment program planning Performance Metrics: measures to monitor implementation and value
12 Version 1.1: AMI AMI: STEMI Value Dashboard 1.1 Diagnosis: Inpatients Diagnosed with STEMI (ST Elevation Myocardial Infarction) (except where otherwise stated) Reporting Period: FY11 Mid-year Report, Oct 2010 - Mar 2011 (except where otherwise stated) Data Measures BWH MGH FH Source NWH NSMC PHS Benchmark/ Goal PHS Color Shading Relative to Benchmark/Goal VOLUME UHC Vol STEMI Cases UHC 74 116 0 4 65 259 7,935 NA NA NA OUTCOMES Mortality (All Cause) In-Hospital (Risk Adjusted) UHC Avg 95% statistical significance Index (Observed/Expected Ratio) 1.5 1.1 No Cases 0.0 0.3 1.1 1.2 Worse Similar Better UHC Observed (N) 12 9 No Cases 0 1 22 619 NA NA NA 30-Day Post-Discharge PHS Goal Relative to Prior Performance* Observed Rate 1.6% 0.9% No Cases 0.0% 3.2% 1.7% Cntrl. Meas.* no change UHC Observed (N) 1 1 No Cases 0 2 4 NA NA NA NA Returns to Hospital UHC Avg 95% statistical significance 30-Day Readmission Rate: Unplanned Readmissions 8.1% 15.0% No Cases 50.0% 7.9% 11.5% 8.0% Worse Similar Better Observed (N) 5 16 No Cases 1 5 27 NA NA NA NA UHC ED visit within 30 Days of Discharge, All Cause 0.0% 1.9% No Cases 0.0% 7.9% 3.0% Avail 2012 NA NA NA Observed (N) 0 2 No Cases 0 5 7 NA NA NA NA PHS Goal % of Patients Discharged Home UHC 82% 90% No Cases 25% 83% 85% Baseline NA NA NA Nat'l Avg Relative to Nat'l Avg/90th %ile CMS AMI Composite Quality Measure FY10 (All AMI Patients) CMS 98% 99% 100% 100% 100% 99% 97% < 97% 97% 98% Rapidity of Treatment of AMI Nat'l Avg Relative to Nat'l Avg/90th %ile PCI within 90 Minutes FY10 (All AMI Patients) CMS 90% 94% NA NA 98% 93% 90% < 90% 90% 98% UHC Median 95% statistical significance Median Time (Minutes) to PCI UHC 67 57 NA NA 61 61 63 Worse Similar Better RESOURCE USE (Acute Hospital Index Stay) Resource Use (Uncomplicated STEMI Patients) PHS Goal Relative to Prior Performance* Percent of Short Stay Patients ( 5 days) with More than One Catheterization 2.7% 5.9% No Cases 0.0% 0.0% 3.2% Cntrl. Meas.* no change Percent of Short Stay Patients ( 5 days) with More than One Echocardiogram 5.4% 1.5% No Cases 0.0% 0.0% 1.9% Cntrl. Meas.* no change TSI/EPSi Average Nursing Acuity Per Patient Day 2.2 2.2 No Cases 1.6 1.7 2.1 Desc. Meas. NA NA NA Average ICU Day(s) Per Case 2.0 2.2 No Cases 0.2 1.8 2.0 Cntrl. Meas.* no change Average Length of Stay (Risk Adjusted) UHC Avg 95% statistical significance Index (Observed/Expected ratio) 1.5 1.4 No Cases 0.5 1.0 1.3 1.1 Higher Similar Lower UHC Observed (N) 7.0 6.6 No Cases 2.0 5.2 6.2 5.2 NA NA NA Cost (Uncomplicated STEMI Patients) Trend Average Direct Cost Per Case $17,425 $20,901 No Cases $2,646 $12,893 $17,073 NA NA NA NA TSI/EPSi FY11 (Q1&2) Versus FY10: increase (decrease) ($4,678) $4,681 No Cases ($331) $2,414 $806 NA NA NA *Control Measure: color scoring is based on comparison to entity's prior period score, not raw value score
- 13 - Version 2.0 Selected Recommendations (Stroke) Days spent at home 90 days post-stroke Functional Independence Measures (FIMs) upon admission and at follow-up: Self-care Sphincter control Transfers Locomotion Communication Social cognition
- 14 - Version 2.0 Selected Recommendations (Stroke) Days spent at home 90 days post-stroke Functional Independence Measures (FIMs) upon admission and at follow-up: Self-care Sphincter control Transfers Locomotion Communication Social cognition We have begun testing new methods for collection of PROMs (Patient Reported Outcome Measures) including patient portals, tablet computers, and interactive voice response systems.
- 15 - Performance Dashboard 1.0 for End of Life Care Diagnosis: Patients receiving end-of-life care. Measures Data Source BWHPO MGHPO NW PHO NSHS/ NSPO DFCI Benchmark VOLUME Cases FYxx RPDR OUTCOMES Documentation of HCP Documentation of resuscitation preferences and EOL care goals ACP dicussion with family and patient Completion of advance directive or durable power of attorney forms QDW Presence of DNR/DNI orders Referral to hospice Time from DNR orders to death Time from referral to hospice to death RESOURCE USE Hospital days ICU days Length of stay ED visits Readmissions Chemotherapy in last 2 weeks of life Number of doctors seen, excl. Palliative Care Inpatient hospice days Costs of care RESTRICTIONS Inclusions: Adult Patients (age >18) who died during fiscal year 20xx with qualifying diagnoses and who were managed as an outpatient by a Partners physician. Resource use counted for 6 months prior to death. Exclusions: Age <18 UHC RPDR/ UHC/ TSI TSI
- 16 - Bohmer Operations Teams that have responsibility for improvement of value forever Data on outcomes and costs collected as routine part of care Incentives (financial and otherwise) for improvement A leader who thinks constantly about improving performance Formal and informal interactions among team members
- 17 - Words That Resonate Through Their Work Value Measurement Outcomes Efficiency Improvement Learning Collaboration Humility Discipline Teamwork