Committee on Geographic Variation in Health Spending and Promotion of High-Value Care Institute of Medicine

1
Committee on Geographic Variation in Health
Spending and Promotion of High-Value
CareInstitute of Medicine

• Testimony of Larry deGhetaldi, M.D.
• Sutter Healths Palo Alto Medical Foundation
• the California Medical Association
• January 17, 2011

2
CMS Mission Statement

• To ensure effective, up-to-date health care
  coverage and to promote quality care for
  beneficiaries
• CMS' Vision
• To achieve a transformed and modernized health
  care system.
• CMS will accomplish our mission by continuing to
  transform and modernize America's health care
  system.
• CMS' Strategic Action Plan Objectives
• Skilled, Committed, and Highly-Motivated
  Workforce
• Accurate and Predictable Payments
• High-Value Health Care
• Confident, Informed Consumers
• Collaborative Partnerships

3
Summary

1. Getting the Geographies Right
2. Getting the Risk/Cost Adjusters Right
3. Start with Existing Anchor System Models
4. Reward Clinical Innovation
5. Reward Sustained Efficiency

4
The Geographies of Dartmouth CMS
HRRs typically resemble and overlap with
MSAs HSAs are sub-county units built around
dominant hospitals
5
1. Getting Geographies Right

• Converge Physician and Hospital Geographies
• Aligns cost risk inputs
• Decreases payment errors
• Size of Geographic Units
• Accountability is lost if the geography is too
  large
• Use MSAs for cost inputs
• Use Dartmouths H.S.A.s for
• Global payments (ACOs and MA payments)
• Value-added features currently in development

6
The Broken Physician Localities

• Physician Payment Localities est. 1966
• No updates since 1996
• GAO (2007)
• Geographic Areas Used to Adjust Physician
  Payments for Variation in Practice Costs Should
  be Revised.
• More than half of the current physician payment
  localities had counties within them with a large
  payment difference of 5 or more percent between
  GAOs measure of physicians costs and Medicares
  geographic adjustment for an area.
• Recommends that CMS revise payment localities
  using a uniform approach and update the
  localities on a periodic basis.

7
Why Are These Maps Different? Congruency Will
Improve Payment Accuracy
(Physicians)
8
Why Are These Maps Different? Input Costs Should
Apply to Measured Geographic Units

• Hospitals
• 12 Distinct MSAs
• High cost DC suburban counties accurately paid
• Physicians
• One statewide physician payment
• Up to 20 underpayments in suburban DC counties
• Small section of VA included in the DC locality

9
Right Sized Geography for Risk Adjustment Los
Angeles Problem

• One-third of CA population
• FFS Consumption MA rates gtgt No CA
• (SF896 LA 1020 Dade 1426)
• Diverse population needs risk adjuster at more
  granular geographic layer
• Commercial rates at or below RBRVS
• Total per capita Medicare exp gtgt Rest of CA
• The parabolic relationship

10
The Los Angeles Problem
The parabolic relationship between income and
expenditures within the Commercial population.
Commercial Population
Average Expenditure Commercial
Poverty also increases expenditures for
Medicare beneficiaries
11
The Los Angeles Problem
Immigrant Status Also Affects Expenditures and
May Contribute Independently to Income
12
Cost Risk Adjuster

• Cost Adjuster
• Set to a national benchmark of 1.0
• Hospital Wage Index is necessary for costs
• Physician GPCIs are necessary cost adjusters
• Use MSAs as the Geography
• Risk Adjuster
• Set to a national benchmark of 1.0
• CMS data for risk used at the beneficiary level
• Apply at a sub-county level reflective of
  community (H.S.A.) risk
• Should capture health, poverty, and immigrant
  status

13
2. Getting the Cost/Risk Adjusters Right

• Step One Measure total Expenditures
• Step Two Remove IME, GME and DSH payments
• Step Three Adjust for Risk (most granular level
  available, typically the H.S.A.)
• Step Four Adjust for Costs (using MSA cost
  inputs for hospitals and physicians)
• Step Five Factor in Historical Consumption
  Inflation (Reward sustained efficiency over time)

14
What are Risk/Cost Adjusters?
15
Why Include Risk/Cost Adjusters?

• Some variations in Medicare expenditures are
  reasonable
• Impact of age, disability status, race/ethnicity,
  poverty, deferred health care, immigrant status
• Higher prevalence of dual eligibles
• Variations in physician practice expenses (GPCIs)
• Variations in hospital core expenses (wage index)
• GME DSH expenditures vary by county/H.S.A.

16
Risk and Cost Adjusters Applied (GAF
Geographic Adjustment Factor)
Dartmouth
CMA Derived
CMS Data
County
Remove DSH et al
Adjust for Risk
Adjust for Cost Inputs
Baselines should reflect cost AND risk
variations and should incent favorable
inflationary trends.
Getting the Benchmarks Right
17
Without Cost Adjustment CA 101 of U.S.
18
With Cost Adjustment CA 90 of US
19
The Los Angeles Problem

• The CMS County Risk Adjuster of 1.12 applied
  countywide
• Underpays low income/high risk communities (HSAs)
• Overpays high income/low risk communities
• Flattens the parabolic curve in large, diverse
  populations
• Could exacerbate health care disparities as it
  unintentionally redistributes Medicare payments
  away from at-risk populations, by reducing the
  desirability of serving Medicare beneficiaries
  in those areas
• CMS has risk data calculated at the FFS
  beneficiary level which could be adapted to
  H.S.A. risk adjusters

20
3. Start with Existing Anchor System Models

• Northern California Medical Group Innovators
• Kaiser Permanente the original disruptive
  innovator
• Sutter Healths Palo Alto Medical Foundation
• And other multi-specialty medical groups
  integrated with hospitals
• Southern California IPA Innovators
• IPAs at risk for total cost of care
• Leveraging the Medicare Advantage programs total
  risk opportunity for physician incentives
• California
• Physician Organizations adopting quality,
  service, and efficiency targets
• Wide adoption of Pay for Performance by
  physicians
• Advanced systems for managing chronic diseases
• Advancing Physician-Hospital Organizations
• Physicians Leading Transformation
• High Hospital Wage Index yet low hospital
  utilization

21
Success with Medicare Advantage Improves FFS
Medicare Performance

• 1,664 FFS Bed-Days versus 982 Bed-Days for MA is
  the CA opportunity
• Los Angeles IPAs have MA Bed-Days in the 600s
• Lower MA Bed-Days derive from coordinated
  physician-led care

22
4. Rewarding Innovation Transparency

• Palo Alto Medical Foundation
• 1,000 physicians, multi-specialty medical group
• 750,000 pts -30 capitated 70 FFS
• Three fully aligned hospitals four partner
  hospitals
• Physician led
• Community Board populated by Silicon Valley
  Industry Leaders
• A culture of Innovation Excellence

23
Sutter Healths PAMF

• Disruptive Innovations Underway
• Group Practice since 1930 Non-profit since 1981
  (Sutter Health Affiliate)
• Research, Education and HealthCare Delivery Meld
  (Physician Led)
• Early e HR and Patient Portal Adoption
  eMessaging/eVisits Free to Pts.
• Interoperable e HR with FQHC (Santa Cruz pilot)
  Specialty care for the safety net (Community
  Need)
• Beyond Pt Centered Medical Home primary care
  redesign (Primary Care Resurrection)
• Physician Dashboards/Internal Transparency
  (Accountability)
• External Quality/Service Transparency (Quality
  Service Standards)
• Shared Decision Making Modules (Patient
  Preference, not MD Preference)
• Shared Medical Appointments (Enhancing
  Compliance)
• Longitudinal Outcomes Analysis partnering with
  hospitals (Value)
• Research Institute Focus on Healthcare Delivery
  Reforms
• David Druker Innovation Center (established 2010)
  (Paul Tang, MD as lead)
• Physician/Admin Pairings at all levels of the
  organization

24
Two Key Innovations Underway at PAMF

• Clinical Variation Reduction
• Physician driven physician defined, specialty
  specific
• Physician-defined Quality/Outcomes metrics
• Measurable Savings to Payers and Patients
• Palliative Care Ambulatory Medical Home
• Leveraging current high performance on end of
  life expenditures
• Patient Family Centered in an ambulatory
  environment with dedicated Palliativists and care
  team
• Contrast

25
How Can CMS Incent Innovation?

• CA Private Sector paid for HIT Quality
  Innovation
• Provide Resources to Physician Groups to support
  Clinical Variation Reduction
• Provide same to individual physicians based on
  physician specialty-defined clinical protocols
• Provide payments for palliative, end-of-life
  medical home care

26
5. Reward Sustained Efficiency
27
Medicares Geographies in CA
Small homogeneous counties like Santa Cruz could
be treated as a single HSA
28
Recommendations

• Adopt MSAs for
• Hospital and Physician Cost inputs (hospital wage
  index MD GPCIs)
• Adopt H.S.A.s for
• Value Based incentives efficiency, service,
  quality, innovation
• Global Payment Targets
• Medicare Advantage Rates
• ACO Benchmarks
• Abandon
• Current 89 Physician Fee Schedule Areas
• County Based MA rates and move to H.S.A. cost
  risk adjustments
• Develop
• Risk Adjustment data at the H.S.A. level
• Cost Risk Adjusted targets for each H.S.A.
• Quality and Service Metrics/targets at the H.S.A.
  level
• Medicare Advantage and ACO targets should be
  applied using congruent geographies with similar
  incentives
• Longitudinal incentive for inflation control
• Incent Innovation
• Pay us (groups solos) to innovate and we will

29
Summary

• Geographies Need
• Update the Physician localities and make
  consistent with Hospital localities
• Risk and cost adjusters at the level of the
  community (H.S.A.) which can be accountable to
  payers and patients
• Risk Cost Adjusters are Needed
• Accurate risk, cost and Value Metrics Actionable
  at the physician and medical group level
• Actionable value metrics at both the physician
  and medical group level
• Anchor System Models
• Exist due to market influences that should not be
  ignored
• Should be early adopters
• Incentivize Innovation
• Especially around data sharing leading to
  clinical variation reduction
• And focused on improvements in end-of-life care
• Inflations Impact Compounds
• SF 2.36 over 15 years
• McAllen 8.31