Lawrence Casalino: what GP consortia might learn from the US

What GP Commissioning Consortia might learn from the development of physician groups in the US: a synthesis of 20 years experience to avoid failure

Lawrence Casalino MD, Ph.D.,Livingston Farrand Associate Professor of Public HealthChief, Division of Outcomes and Effectiveness Research

Weill Cornell Medical CollegeWe Co e ed ca Co egeNew York City

The John Fry Lecture Nuffield TrustOctober 18, 2010

Today’s talk

1. Two organizing frameworks for thinking about GP commissioning g gconsortia

2 U S experience with “consortia” and2. U.S. experience with consortia and commissioning

3 Seven theses on GP commissioning3. Seven theses on GP commissioning4. Suggestions from an outsider

Two views of quality

• the individual physician viewthe individual physician view

h i d i• the organized process view

Two types of things that must beTwo types of things that must be created

• incentives• capabilitiescapabilities

f f(i i• performance = f(incentives + capabilities)

Exhibit 12. Premiums Rising Faster Than Inflation and Wages

Projected Average Family Premium as a Percentage of Median Family Income,

2008–2020

Cumulative Changes in Components of U.S. National Health Expenditures and

Workers’ Earnings, 2000–2009

100

125Insurance premiumsWorkers' earningsC P i I d

Percent Percent

108%

18 18 18 1819 19 19

20 2021 21

22 2223

24

1820

25

50

75

Consumer Price Index

1112

1314

1617

10

15

0

25

5032%

24%0

5

9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0

* 2008 and 2009 NHE projections. Data: Calculations based on M. Hartman et al., “National Health Spending in 2007,” Health Affairs, Jan./Feb. 2009;

02000 2001 2002 2003 2004 2005 2006 2007 2008* 2009* 19

99

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

Projected

THE COMMONWEALTH

FUND

and A. Sisko et al., “Health Spending Projections through 2018,” Health Affairs, March/April 2009. Insurance premiums, workers’ earnings, and CPI from Henry J. Kaiser Family Foundation/Health Research and Educational Trust, Employer Health Benefits Annual Surveys, 2000–2009.Source: K. Davis, Why Health Reform Must Counter the Rising Costs of Health Insurance Premiums, (New York: The Commonwealth Fund, Aug. 2009).

Exhibit 1. National Health Expenditures per Capita, 1980–2007

$Average spending on health per capita ($US PPP)

7000

8000

United States

6000

7000 CanadaFranceGermanyNetherlands

4000

5000 United Kingdom

2000

3000

0

1000

1980 1984 1988 1992 1996 2000 2004

THE COMMONWEALTH

FUND

Data: OECD Health Data 2009 (June 2009).

1980 1984 1988 1992 1996 2000 2004

Quick summary: history of U SQuick summary: history of U.S. “commissioning”

• Anticipated move to “full-risk” contracting did not occur.

• Most physician organizations created to engage in• Most physician organizations created to engage in risk contracting failed– ~ 2000 IPAs created

200 IPAs successful (at the most)– ~ 200 IPAs successful (at the most) • High profile failures of large fund-holding IPAs.• There is now little or no risk contracting in most of

the U.S.• In California and pockets elsewhere, risk

contracting persists in modified forms.contracting persists in modified forms.

Why did risk contracting failWhy did risk contracting fail, overall, in the U.S.?

• policy failurespolicy failures

i i l f il• organizational failures

Policy failures - failure to:• risk-adjust• balance incentives

h i i d i i d i k i b– physicians and patients perceived risk contracting to be about reducing costs

– not about improving quality or patient experienceid i l i f i• provide timely, accurate, transparent information to

the “consortia”• recognize how difficult it is to build competent g p

physician organizations• reduce incentives for specialists and hospitals to

churn high profit serviceschurn high profit services

Organizational failures - failure to:

• invest in:– physician leaders

kill d– skilled managers– IT– adequate staff (e g nurse care managers)adequate staff (e.g. nurse care managers)

• adequately analyze the level of risk • track IBNR (incurred but not reported)( p )• motivate/coordinate their physicians• gain specialist/hospital cooperationg p p p

Flow of funds? NHSNHS

GP Consortium

Hospital

ConsultantsGPs

Consultants

Thesis 1

It will be extremely difficult to create high-performing GP g p gcommissioning consortia. The government should not expect that g plarge numbers of high performing consortia will be formed overnight, g ,or even within 3-5 years.

Necessary capabilities for GPNecessary capabilities for GP consortia

• leadership• organized processes to improve care (not g p p (

just to commission it)• sophisticated information collecting and

processing– and people with the time and skills do do

thi ith th i f tisomething with the information– sophisticated financial capabilities, including

both accounting and modelingg g

Necessary capabilities for GPNecessary capabilities for GP consortia (more)

• ability to create and manage relationships with many external entities

• ability to pay claims??• a culture of cooperation and quality

improvement– not only within the GP consortium, but

with outside entities as well

Even with perfectly designed incentivesEven with perfectly designed incentives, the risk of failure is high

• inadequate supply of GP leaders• GP consortia likely to underinvest in

management• takes time to develop culture

b diffi lt t i ti• may be very difficult to gain cooperation from consultants and hospitals

• GP consortia will be more like IPAs than• GP consortia will be more like IPAs than multispecialty medical groups or integrated systems

Thesis 2

It will be necessary to create incentives for cooperation at multiple levels within the health care delivery system.

GP i- GP consortium- GP practice/individual GP- consultant/specialist physiciansconsultant/specialist physicians- hospitals- and others

To gain support from rank and fileTo gain support from rank and file GPs:

• GPs must believe that changes will significantly improve some or all of g y pthe following:– quality of care for their patientsquality of care for their patients– quality of their workday– respect from their peersrespect from their peers– physician income

Ways to influence physicians withinWays to influence physicians within an organization

• develop an organizational culture• include only physicians compatible with the desired

cultureculture• educate/persuade/develop guidelines• show physicians in the organization data on:

– the organization’s performance – the performance of practices/individual MDs within the

organization• choose payment methods to reward desired

behavior• require prior approval for certainrequire prior approval for certain

referrals/procedures (for some physicians?)

Thesis 3

Incentives should not focus primarily on generating savings/reducing the g g g gcost of care. They should be balanced among quality, patient g q y, pexperience, and cost-control.

Thesis 4

Incentives should be neither too strong nor too weak.

Should have:

• risk-adjustment• moderate upside and smaller downside risk, p ,

gradually increasing over time– threat to close a consortium not likely to be

h h i b hi ienough when consortium membership is required for GPs

• risk modifiers e g stop loss insurance for• risk modifiers - e.g. stop-loss insurance for outlier patients

Thesis 5

It will be critically important to find ways to foster collaboration among y gGPs, specialist physicians, and hospitals.p

What’s in a name?

• GP Commissioning is likely not an ideal name

• Why not call it “GP Dominance?”• Why not call it GP Dominance?

Other barriers

• basically impossible to form a multispecialty groupp y g p

• incentives not aligned: Payment by• incentives not aligned: Payment by Results

We’ll know the system is workingWe ll know the system is working when:

• GPs and consultants frequently discuss patients on the telephonep p

• Phone conversations often replace• Phone conversations often replace visits to consultants

Thesis 6

Don’t skimp on funds for consortium management!g

Management costs

• critical to have:– skilled clinical and lay leaders– infrastructure support (people and data)– data in itself is useless

th t b l d h l i j b i t• there must be leaders whose only or main job is to help the GP group improve the care provided

• left to themselves, GPs will under-invest in ,management– (at least until they see a reliable ROI)

Thesis 7

• GP commissioning is likely to result in the transfer of a large amount of NHS gfunds to the private sector– (for better or for worse)(for better or for worse)

UK advantages (1)

• “single payer” gives the opportunity to:– collect comprehensive data – risk adjust– balance incentives (cost, quality, patient

i )experience)– invest in the development of physician leaders– invest in management costs in GP consortiainvest in management costs in GP consortia

UK advantages (2)

• public acceptance of GPs

• savings perceived as going to NHS, not to corporate executives andnot to corporate executives and shareholders

UK advantages in developingUK advantages in developing physician leaders

• NHS can pay GP leaders• NHS can provide training for GPNHS can provide training for GP

leaders• NHS can provide an attractive career• NHS can provide an attractive career

track for GP leaders

Suggestions (1)

1. anticipate failures; don’t overinflate expectations for rapid, widespread change b d f d l f2. budget for gradual performance improvement by GP consortia

- provide upside and downside incentives- provide upside and downside incentives - with incentives increasing over time

3. balance incentives: cost, quality, patient , q y, pexperience

Suggestions (2)

5. make it possible for GP consortia to have financial leverage vis-à-vis member physicians/practicesphysicians/practices

6. seek ways to create substantial financial incentives for hospitals and consultants toincentives for hospitals and consultants to cooperate with GP consortia

7. seek ways to make it attractive for consultants to join with GPs in creating multispecialty medical groups

Suggestions (3)

8. provide substantial ring-fenced management funds to GP consortia for 4 years then blend into their budget (and ?years, then blend into their budget (and ? reduce the funds)

9. consider a name other than “GP9. consider a name other than GP commissioning”

10. invest in developing GP and consultant leadership