Value, Suffering, and 10 Things I Didn t Know Before My New Job Thomas H. Lee, MD October 28, 2013 2 1
Why We Are Stuck 3 Getting Unstuck 2
Step One: Clarifying the Overarching Goal In the absence of an overarching goal, interactions among providers devolve to gamesmanship, in which everyone tries to optimize their own portfolios The goal of health care is more than the reduction of health care spending Not fully captured by mortality, or measures of health either Traditional definitions of quality (e.g., HEDIS) focus on provider reliability not meeting the needs of patients Improving value for patients (i.e., the outcomes that matter to patients versus the costs required to deliver those outcomes) is the one goal that is embraced by all stakeholders in health care. 5 Four Key Success Factors for Providers 1. What we get paid 2. What it costs us 3. Market share of patients 4. Market share of personnel Improvement of value (outcomes/costs) favorably influences all four major success factors for providers and is robust regardless of mixture of payment model. 6 3
Six Component Strategy for Improving Value 1. Build real teams of Integrated Practice Units (IPUs) 2. Measure outcomes/experience and costs for all patients 3. Episode-based incentives for value 4. Integration of care across delivery systems 5. Growth of excellent IPUs 6. Enabling information technology platform These components are interdependent and synergistic not a menu from which 1-2 should be sampled. 7 1. IPUs Real Teams v. Pick-up Teams Traditional management structure aimed at optimizing efficiency of use of resources (e.g., Hospital beds, MD time) Need: Real teams focused on improvement of outcomes/efficiency for segments of patients Actionable: Patients with similar shared needs, so that teams can meet most of them Not focused factories Palliative care service with oncology clinicians Focus: Full cycle of care Personnel: Multidisciplinary, meet regularly, co-located Data: Outcomes and costs 8 4
IPU Example: Virginia Mason Spine Clinic Key features 1. One phone number 2. Same-day visits 3. MD physiatrist and physical therapist see patient as team 4. PT often started first day 5. Lower costs, radiology testing, time lost from work 6. More patients seen in same physical space 9 2. Measure Outcomes and Costs on Every Patient Key points: 1. Tier 1 outcomes most important, but difficult (or impossible) to differentiate on basis of superior performance. 2. Tier 2 outcomes (experience of care) are critical differentiators 10 5
3. Bundled Payments Neither fee-for-service nor capitation directly drive improvement of value (which is the only robust strategy across payment models) Bundled Payments seem complex but it is actually the provider organization that is complicated Bundled payments are spreading, and early adopters have been rewarded with market share. Examples: Transplant programs Walmart cardiac and spine surgery programs 11 Virtuous Cycle Driven by Increased Volume Bundled payments combined with excellent outcomes and positive patient experience leads to more market share, which leads to ability to cultivate more effective teams, which leads to better outcomes and efficiency, etc. Volume in a Medical Condition Enables Value The Virtuous Circle of Value Improving Greater Patient Reputation Volume in a Medical Better Results, Condition Adjusted for Risk Rapidly Accumulating Experience Faster Innovation Better Information/ Clinical Data Costs of IT, Measure- ment, and Process More Fully Improvement Spread Dedicated Teams over More Patients Greater Leverage in More Tailored Facilities Purchasing Rising Process Wider Capabilities in Efficiency the Care Cycle, Including Patient Better utilization of Engagement Rising capacity Capacity for Sub-Specialization Volume and experience will have an even greater impact on value in an IPU structure than in the current system 2012.03.26 VBHCD Core Concepts 9 Copyright Michael Porter 2012 Porter/Teisberg, Redesigning Healthcare (HBS Press, 2006) 12 6
Strategic Components 4-6 4. Integrate care across system Consolidate care where it can be done at highest value Be willing to stop delivering care at sites if it can be done at higher value elsewhere 5. Grow excellent IPUs Regional spread of satellites Deep affiliations 6. Enabling information technology 13 Deconstructing Suffering AvoidableSuffering Due to Dysfunction in Care System Unavoidable Suffering Due to Treatment Unavoidable Suffering Due to Diagnosis 14 7
Translating Good Intentions into Action Max Weber s Four Models for Social Action 1. Tradition e.g., Mayo Dress Code 2. Self-interest e.g., Performance bonuses 3. Affection e.g., Peer pressure 4. Shared purpose e.g., Reducing suffering 15 Translating Good Intentions into Action Max Weber s Four Models for Social Action 1. Tradition e.g., Mayo Dress Code 2. Self-interest e.g., Performance bonuses 3. Affection e.g., Peer pressure 4. Shared purpose e.g., Reducing suffering We need to press all four levers -- but the first lever that must be pressed is creation of Shared Purpose. In isolation, any of the other three levers is ineffective or potentially perverse. But in pursuit of a shared purpose, all three other levers can be embraced. 16 8
Appreciative Inquiry as a Tool to Create Shared Purpose Focus on positive, not errors What went right? What characterizes the cases that made us proud? Identify that features that characterize care at its best and try to make those things happen reliably. Deconstruct great care and focus organization on delivering it. Emphasis always ends up on patients. Examples: Cleveland Clinic appreciative inquiry exercises for all personnel Commitments to make appointments available on same day basis when patients request it. 17 Data Drives the Other Three Levers Affection (Peer Pressure) Individual MD-level data on quality/efficiency drives improvement: Fosters learning Creates pressure Self-interest (Financial Incentives) Patient experience data increasingly being used in compensation programs. Tradition Clinicians who fail in attempts to improve their performance may be asked to leave organizations Implication: The stakes are high, so you need good data on metrics that really matter and lots of it. 18 9
10 Things I Didn t Know 1. Likelihood to recommend/overall rating NOT driven by amenities (food, parking) 2. Coordination of care is HUGE driver of likelihood to recommend 3. Nursing communication to patients is the biggest single driver of inpatient likelihood to recommend 4. When clinicians bad-mouth others to patients, the patients lose confidence in everyone including the criticizing clinician. 5. On the other hand, when clinical personnel are supportive of other clinicians, patients trust that they are in good hands increases. 6. Even if MD s expertise is greatly respected, patients likelihood to recommend falls if care not coordinated and concerns not heard. 7. Patients admitted through ED give lower hospitalization ratings 8. Pain is common on medical service 9. Patients are generous graders meaning when they don t give top rating, we have disappointed them. 10. E-surveying of every patient is way of the future. 19 Discussion 20 10