Thomas H' Lee, MD, MSc'

1
The Nash Equilibrium Breaks Down

• Thomas H. Lee, MD, MSc.
• Network President, Partners HealthCare System
• Professor of Medicine, Harvard Medical School
• Professor of Health Policy and Management,
  Harvard School of Public Health
• Associate Editor, The New England Journal of
  Medicine
• July 23, 2009

2
The Good News in Massachusetts
Enrollment Since April 2006
MA now has lowest uninsured rate in U.S. (2.6)
But MA didnt make healthcare a right we made
it a responsibility. And that has unmasked a
major problem
3
Health Care Affordability Is Now a Middle Class
Problem
Cumulative increase 2000-2007
Employer Health Benefits 2007 Annual Survey
(7672), The Henry J. Kaiser Family Foundation
HRET, September 2007 This information was
reprinted with permission from the Henry J.
Kaiser Family Foundation. The Kaiser Family
Foundation, based in Menlo Park, California, is a
nonprofit, private operating foundation focusing
on the major health care issues facing the nation
and is not associated with Kaiser Permanente or
Kaiser Industries.
4
Why We May Be Hitting Generositys Brick Wall
5
Willingness of Healthier and Wealthier to
Subsidize Care for Sicker and Poorer is Weakening
Harris Survey question Do you agree or
disagree? The higher someones income is, the
more he or she should expect to pay in taxes to
cover the cost of people who are less well off
and are heavy users of medical services.
Implication We shouldnt expect help from
taxpayers.
http//www.harrisinteractive.com/news/allnewsbydat
e.asp?NewsID1076
6
The Bad News Progress Raises Costs and
Generates Chaos

• Flood of progress and knowledge imposed on
  fragmented delivery system leads to
• Individual clinicians feel less knowledgeable
• Super-specialization, which means
• More MDs involved in care
• Physicians knowing more and more about less and
  less until they know everything about nothing or
• less and less about more and more until they
  know nothing about everything
• Physicians approaching patient with question of
  Is this what I do?
• Too many people, too much to do, no one with all
  the responsibility or all the information

7
No Bad Guys to Blame for Our Issues

• Why are healthcare costs rising?
• Surprisingly small contributions from
• Profits of drug/device companies
• Administrative costs
• Malpractice
• Aging of the population
• Life-style choices
• Personnel
• The dominant factor progress (60-70) is main
  driver of rising costs
• Safety and reliability issues are attributable to
  turbulence in the wake of progress as well.

8
Reason for Optimism
9
John Nash
10
John Nashs Nobel Prize Work

• Nobel Prize for Economics in 1994 for describing
  an equilibrium concept for non-cooperative
  games in which binding agreements cannot be
  written.
• Nash Equilibrium -- Multiple parties frozen in
  current relationships because no party can change
  its strategies while the other parties keep their
  strategies unchanged.
• Nash Equilibriums break down when pain of status
  quo for multiple parties exceeds fear of unknown.

11
An Optimistic Long-term Perspective
Finale Breakthrough
High
Act III Clinical Reengineering
Large annual gains in quality and affordability
Act II Performance Sensitivity
Faster uptake discovery of better, faster,
leaner care delivery innovations
Efficiency of Health Benefits Spending (Health
Gain / )
Act I Transparency
Performance-sensitive health plan design and/or
provider payments
Universal hospital MD performance transparency
Low
2005
2015
Evolutionary Path
Slide used with permission of Arnold Milstein,
MD, of Mercer
12
Working to Achieve the Vision EMR Adoption
Percent of PCPs Using EMR

• Success in adoption has allowed focus to shift to
  effective use
• Rate of Computer Generated Prescriptions among
  PCPs is 85-88.
• Rate among community specialists has exceeded 70.

Percent of Specialists Using EMR
13
Prospect Theory Explains Why Relatively Small
Incentives Can Produce Major Change
Prospect Theory, Kahneman and Tversky,
Econometria 1979
14
Evolving Reimbursement and Care Models
Full Capitation
Closed System
Sub-Capitation
Team-Based Care
Case Rates
Disease Management
Evolution of Supporting Systems
PAYMENT METHODOLODY
P4P (Robust)
EMR
P4P (Lite)
Registries
Fee-for-Service
Non-MD Clinicians
Solo MD Practices
Multi-Specialty Group Practices
Integrated Delivery System
Clinic Model
Group Practices
STAGE OF EVOLUTION
15
Disease Management Averts a CHF Admission
MD notified of weight gain. Patient called, and
MD learned she had stopped taking furosemide
twice daily. Regular regimen restored
16
The Real Agenda Two Revolutions

• Industrial Revolution in which clinicians adopt
  systems that reduce errors of over-use,
  under-use, and mis-use.
• Cultural revolution
• Teamwork instead of MD as the lone cowboy
• Focus on care of populations over time
• Chronic diseases like diabetes, heart failure
• Complex, high risk patients with multi-system
  disease

17
Medicines Cultural Revolution

• New types of responsibilities
• Responsibility for non-visit care of patient
• Responsibility for population of patients
• Evolving concepts of professionalism
• Not just highest possible individual standards of
  excellence
• Ability and willingness to work with teams that
  can assume new responsibilities. Examples
• Use of EMR
• Computerized prescribing
• Medication reconciliation at discharge
• Opt out approach to team care
• Exploration of variation in practice patterns

18
Variation A Challenge and Opportunity

• Issues for which there is a clear right and wrong
  (.e.g, ASA for AMI) constitute minority of
  medical decisions.
• Most decisions are gray zone issues for which
  there is no clear right thing to do.
• But if there is a bell-shaped distribution of
  what rational professionals (e.g., your
  colleagues) are doing in that gray zone, wouldnt
  you want to know if you are at one end or the
  other?

19
Variation is Greatest When Right Thing to Do Is
Less Clear
Variation in rates of care across 306 Medicare
regions (2000-01).
Hip Fracture
Back Surgery
Implications
Procedure Rates of 306 Regions
Adapted from The Dartmouth Atlas of Health
Care Jack Wennberg presentation 2005.
20
The approach to managing variation differs
depending on the existence of a standard of care
Standard of care exists

• Gather and feed back data
• Set guidelines, standards or protocols
• Consider explicit financial/non-financial
  incentives
• Provide analytic services and peer support

Variation in clinical practice

• Success requires mindset that variation is
  undesirable even without a willingness to define
  a group norm
• Describe variation and agree to internal
  standards
• Can form basis for research to define standard of
  care
• Success requires very high will from clinicians
  to reduce variation (bottom-up nature of project
  is even more crucial)

No standard of care exists
?
21
Data on Variation Are Reaching Individual MDs
22
Why Does Variation Exist Within Small Groups?

• Clinicians are overwhelmed with information, and
  have gaps in knowledge
• Experts tend to get to answers in fewer iterative
  cycles
• Clinicians vary in tolerance of risk/uncertainty
• Experts can often live with greater level of
  uncertainty
• Clinicians are isolated, and do not have way to
  develop group consensus
• Clinicians are influenced by local norms from
  where they trained and their current environment
• But ironically, clinicians often dont know how
  they compare with local norms

23
Variation in physician risk thresholds drive
individual propensity to act regardless of
patient risk
Pearson et al., Triage Decisions for Emergency
Department Patients with Chest Pain. J Gen Intern
Med. 1995 10557-564.
24
Taking on Variation PHS Strategies

• Develop guidelines, and disseminate them
• Attack the gray zone where specific guidelines
  cannot be described yet
• Increase group-ness and increase conversations
• Show data
• Ideally, in unblinded, ranked formats for
  practice
• Also provide data proximate to time to ordering
  of tests/drugs
• Follow-up on data with chart reviews
• Cultivate individual accountability through 1-1
  meetings and pairing of clinicians for chart
  reviews

25
Can We Address Right Side of Curve?
Rogers EM. Diffusion of Innovations, 1983
26
Conclusions

• An important root cause of our challenges in
  healthcare is tremendous progress imposed on a
  fragmented delivery system
• Result is chaos leading to inefficiency and
  disappointing reliability and safety
• Regardless of how healthcare is financed,
  important strategy is for healthcare providers to
  become organized and adopt systems that improve
  quality and efficiency -- over episodes of care
  that matter to patients
• Organization as a goal poses challenges and
  opportunities for medicines leadership