Competition, Prospective Payment, and Outcomes in Post-Acute Care Markets

On September 28, 2010, Neeraj Sood made this presentation at the 2010 Annual Conference. Select to access the PowerPoint® presentation (2 MB). Free PowerPoint® Viewer (Plugin Software Help).

Slide 1

Competition, Prospective Payment, and Outcomes in Post-Acute Care Markets

Neeraj Sood
Schaeffer Center and School of Pharmacy, USC
RAND Corporation

Slide 2

Research Goals

Improve our understanding of:

1. The causes and effects of competition in Post Acute Care (PAC) markets.
2. How have changes in payment policies both influenced competition and modified the effects of competition on resource use and clinical outcomes.

Slide 3

Outline

1. Policy Context
2. Research Team
3. Specific Aims
4. Hypothesis, Data and Approach

Slide 4

Outline

1. Policy Context
2. Research Team
3. Specific Aims
4. Hypothesis, Data and Approach

Slide 5

What is Post Acute Care?

• Health care services received after discharge from acute care hospital that aim to improve patient functioning and transition from hospital to community.
• Care can be received in a wide variety of settings including:
  • Inpatient Rehabilitation Facility (IRF)
  • Skilled Nursing Facility (SNF)
  • Long Term Care Hospital (LTCH)
  • Home Health Care Agencies (HHA)

Slide 6

Post Acute Care is Growing and Large Part of Medicare

• In 2006, 2 out of 5 Medicare beneficiaries discharged from hospitals used post acute care.
• Medicare PAC expenditures more than doubled from 2001 to 2009:
  • Medicare spent $26.6 billion on PAC in 2001
  • Medicare spent $54.4 billion on PAC in 2009

Slide 7

Several Changes in PAC Payment Policy Since 1997

• Move to prospective payment systems (PPSs) for all PAC settings.

Image: A timeline shows the following dates:

• HHA IPS, October 1, 1997
• SNF PPS, July 1, 1998
• HHA PPS, October 1, 2000
• IRF PPS, January 1, 2002

Slide 8

Several Changes in PAC Payment Policy Since 1997

• Move to prospective payment systems (PPSs) for all PAC settings.
• Several adjustments to PPS and other new payment rules.

Image: A timeline shows the following dates:

• HHA IPS, October 1, 1997 (to October 1, 2000)
• SNF PPS, July 1, 1998
• BBRA: SNF PPS (Adjusted April 1, 2000)
• HHA PPS, October 1, 2000
• BBRA: SNF PPS (Adjusted April 1, 2001)
• IRF PPS, January 1, 2002
• LTCH PPS, October 2002
• IRF 75% rule (phased in over 4 years), April 2004
• MMA increased per diem SNF payment, October 2004
• SNF RUG groups increased from 44 to 53, January 2006
• DRA of 2005 Instituted HHA P4P, January 2007
• MMSEA of 2007 permanently reset IRF 75% rule to 60%, December 2007
• MMSEA of 2007 LTCH 25% rule and moratorium, December 2007

Slide 9

Other Payment Changes Expected

• Current payment system has some issues:
  • Same care in different care settings receive different amount of Medicare payment.
  • Lack of incentives to coordinate care.
  • Patients discharged early from one setting might be readmitted to acute care or receive care in other settings.
• Several have advocated for bundled acute care and post acute payment:
  • Health care reform bill includes demonstration project.

Slide 10

The Composition of PAC Markets is Also Changing

• Rapid growth in number of home health care agencies:
  • About 7,000 providers in 2001
  • More than 10,000 providers in 2009
• Number of long term care hospitals has also increased rapidly:
  • About 280 long term care hospitals in 2001
  • About 430 long term care hospitals in 2009
• Number of IRFs and SNFs have remained stable.

Slide 11

How Do Changes in PAC Markets and Payments Impact Medicare?

• Little is known about the impact of changes in PAC markets:
  • Have changes in payments influenced entry and exit of providers and competition in PAC markets?
  • How do changes in number of providers and competition more generally influence patient outcomes and costs?
  • How have changes in payments and competition affected patient outcomes and costs?
• This proposal seeks to address some of these questions.

Slide 12

Outline

1. Policy Context
2. Research Team
3. Specific Aims
4. Hypothesis, Data and Approach

Slide 13

Research Team

• Neeraj Sood, PhD. (Principal Investigator)
  • Associate Professor at USC and Senior Economist at RAND
  • Expertise in empirical economics and health economics
• José J. Escarce, M.D., Ph.D (Co-PrincipaI Investigator)
  • Professor at UCLA and Senior Natural Scientist at RAND
  • Expertise in post acute care markets and health care financing and organization

Slide 14

Research Team

• John Romley, PhD. (Investigator)
  • Assistant Professor at USC and Economist at RAND
  • Expertise in industrial organization and competition
• Peter Huckfeldt, Ph.D (Investigator)
  • Associate Economist at RAND
  • New PhD (will become expert in post acute markets and competition by the end of the project!)

Slide 15

Outline

1. Policy Context
2. Research Team
3. Specific Aims
4. Hypothesis, Data and Approach

Slide 16

Specific Aims (1 & 2): Examine the Effects of Competition

1. Examine the effects of competition on resource use and clinical outcomes for PAC episodes among patients discharged from an acute care hospital after a stroke, fracture, or LEJR, and how these effects are mitigated or enhanced by payment policies.
2. Assess how the effects of competition on resource use and clinical outcomes for PAC episodes differ by care setting.

Slide 17

Specific Aim (3): Examine Determinants of Competition

3. Describe the variation across geographic market areas and trends over time in PAC competition, and assess the factors that influence these variations and trends, especially changes in PAC payment policies

Slide 18

Outline

1. Policy Context
2. Research Team
3. Specific Aims
4. Hypothesis, Data and Approach

Slide 19

Specific Aims (1 & 2): Examine the Effects of Competition

Study Hypothesis:

1. Competition increases costs of care but has ambiguous effect on patient outcomes.
2. Switch to PPS reduces effects of competition on costs but effect on patient outcomes is ambiguous.
3. Under PPS, the influence of competition on cost and clinical outcomes is smaller for providers whose payment levels are more constrained.

Slide 20

Specific Aims (1 & 2): Examine the Effects of Competition

Data:

1. Medicare claims data from 1996 to 2008 for patients with stroke, hip replacement, & lower extremity joint replacement patients.
2. Mortality data linked to claims data above.
3. Administrative data on nursing home stays (MDS) linked to claims data above.
4. Medicare cost report data to estimate resource use for different providers.

Slide 21

Specific Aims (1 & 2): Examine the Effects of Competition

Empirical Models:

• Outcome Variables:
  • Cost per PAC episode (60, 120, 180 day episodes)
  • Mortality and/or institutionalization at episode end
• Key explanatory variables:
  • Competition among PAC providers
  • Competition interacted with payment policy
    • Cost based versus PPS
    • Generosity of payments and Medicare share of provider
• Covariates include demographics, health status measures, acute hospital and market characteristics

Slide 22

Specific Aims (1 & 2): Examine the Effects of Competition

Empirical Models:

• Key challenge is accounting for bias due to the fact that competition measures based on actual patient flows might induce spurious correlation between outcomes and competition.
• Address this concern by constructing competition measures using predicted patient flows that only rely on geographic distribution of patients and providers in a market.

Slide 23

Specific Aim (3): Determinants of Competition

Study Hypothesis:

1. More generous Medicare payment for PAC services within a market leads to more competition.
2. Greater demand within a market for PAC services leads to more competition.
3. Lower costs of providing PAC services leads to more competition.

Slide 24

Specific Aim (3): Determinants of Competition

Data:

1. Medicare cost report data and provider of service files to estimate number of providers in a given market.
2. Area Resources File, Bureau of Economic Analysis and Medicare claims data to measure demand side factors such as age distribution of population, income, acute care discharges, etc.
3. Hospital wage index and Bureau of Labor Statistics data to measure input costs
4. Data on state policies from a variety of sources.

Slide 25

Specific Aim (3): Determinants of Competition

Empirical Models:

• Outcome Variables:
  • Number of providers
  • Hirschmann-Herfindahl Index
• Key explanatory variables:
  • Payment policy: payment regime and generosity
  • Demand side factors: population income, age, health status and access to acute care hospitals
  • Supply side factors: labor costs and rents
  • Other state policy such as certificate of need laws, minimum staffing rules and Medicaid policy

Slide 26

Potential Policy Implications

• Should we promote competition in health care markets?
• Understand the extent to which geographic variation in costs and outcomes is influenced by the level of competition.
• Understand how effects of payment policy changes might vary across markets or regions with different competitive environments.

Slide 27

Thank You!
Questions?

Neeraj Sood
Schaeffer Center and School of Pharmacy, USC
RAND Corporation