Health Care Sector Efficiency: Measuring and improving it Christophe André OECD, Economics Department Institute of Health Economics – Alberta, Canada Becoming.

Health Care Sector Efficiency:Measuring and improving it

Christophe AndréOECD, Economics Department

Institute of Health Economics – Alberta, CanadaBecoming the Best: Building Sustainable Health Systems

High Performing Health Systems

Edmonton, April 15th, 2011

Health at a Glance and Health Data

Health Care Systems: Efficiency and Policy Settings

OECD Economic surveys: Canada, 2010

Improving Value in Health Care – Measuring quality

Value for Money in Health Spending

Obesity and the Economics of Prevention: Fit not Fat

Websites:Health

The Economics of Health

The OECD’s work on healthSelected publications and weblinks

Outline of the presentation

1. Measuring health care outcomes…

2. …and inputs

3. Deriving efficiency indicators

4. Reaping efficiency gains: the impact on public spending

5. Performance and institutions

A positive link between health care spendingand outcomes but with country differences

aus

aut

bel

can

cze

dnk

fin

fra

deugrc

hun

isl

irl

ita

jpn

kor

lux

mex

nld

nzl nor

pol

prt

svk

esp swe

che

tur

gbr

usa

72

74

76

78

80

82

84

0 1000 2000 3000 4000 5000 6000 7000 8000

Total expenditure on health per capita, US $ PPP

Life expectancy at birth, years

Source: OECD Health Data 2010.

1. Measuring health care outcomes

Life expectancy (raw and adjusted for morbidity and disabilities), specific mortality indicators (infant, premature and amenable mortality)

Volume of health care consumption

Quality of care (avoidable hospital admissions and in-hospital fatality rates)

Life expectancy at birth

Source: Health at a Glance 2009, OECD Indicators.

JapanSwitzerlandAustraliaItalyIcelandSpainFranceSwedenCanadaNorwayNew ZealandNetherlandsAustriaGermanyBelgiumIrelandFinlandUnited KingdomGreeceLuxembourgKoreaPortugalOECDDenmarkUnited StatesCzech RepublicPolandMexicoSlovak RepublicHungaryTurkey

82.681.981.481.481.281.081.081.080.780.680.280.280.180.079.879.779.579.579.579.479.479.179.178.478.177.0

75.475.074.373.373.2

405060708090

2007 1960

Years

Life expectancy at 65, women

Source: Health at a Glance 2009, OECD Indicators.

JapanFranceSwitzerlandSpainItalyAustraliaCanadaFinlandBelgiumNorwayAustriaGermanyNew ZealandSwedenIcelandKoreaNetherlandsUnited StatesLuxembourgOECDPortugalUnited KingdomIrelandGreeceDenmarkPolandCzech RepublicMexicoHungarySlovak RepublicTurkey

23.622.322.222.021.821.621.421.321.020.820.820.720.720.720.620.520.520.320.320.220.220.120.1

19.619.218.918.518.2

17.317.1

15.8

510152025

2007 1970

Years

Premature mortality, adjusted for transport accidents, suicides and assaults

Source: OECD Health Data.

0

1000

2000

3000

4000

5000

6000Years of life lost per 100 000 population < 70, 2005

Amenable mortality

Source : Gay et al. (2011), "Mortality Amenable to Health Care in 31 OECD Countries: Estimates and Methodlogical Issues", OECD Health Working Paper, No. 55.

All causes, 2007 or latest year available

0 50 100 150 200 250

EstoniaHungary

Slovak RepublicPolandMexico

Czech RepublicPortugal

United StatesChile

OECDSloveniaDenmark

KoreaUnited Kingdom

New ZealandIreland

GermanyIsrael*

GreeceFinland

LuxembourgCanada

SpainNorwayAustria

AustraliaNetherlands

SwedenJapan

ItalyIcelandFrance

Age-standardised rates per 100 000 population

Correlations between outcome measures(level and rank)

LE at birth LE at 65Total Female

Life expectancy at birth, total 1.00 0.94 ** -0.93 ** 0.96 ** -0.96 **

Life expectancy at 65, female 0.89 ** 1.00 -0.77 ** 0.91 ** -0.86 **

Adjusted PYLL, total -0.82 ** -0.64 ** 1.00 -0.90 ** 0.91 **

Health-adjusted life expectancy at birth 0.95 ** 0.85 ** -0.84 ** 1.00 -0.89 **

Amenable mortality -0.92 ** -0.82 ** 0.85 ** -0.93 ** 1.00

Amenable mortality

Health-adjusted LE

Adjusted PYLL

Source: Joumard , André & Nicq (2010), "Health Care Systems: Efficiency and Institutions", OECD Economics Department Working Paper, No. 769.

Obesity rates

Source: OECD Health Data.

0

5

10

15

20

25

30

35

Self-reported Measured

% of total population

Volume of care Hospital discharges

Source: OECD Health Data.

278 274

227

203194 191 190 189

188174 172 170 166 166 165 162 158 156

139 138 135 132 126 126

109 108 107 106 105

84

55

0

50

100

150

200

250

300Per 1 000 population, 2007

Volume of care Physician consultations

Source: OECD Health Data.

JapanCzech RepublicKoreaSlovak RepublicHungarySpainBelgiumDenmarkGermanyItalyOECDPolandAustriaIcelandAustraliaFranceLuxembourgCanadaNetherlandsTurkeyUnited KingdomNew ZealandFinlandPortugalSwitzerlandUnited StatesSwedenMexico

13.612.6

11.811.2

10.88.1

7.67.57.5

7.06.86.86.76.56.36.36.15.85.75.6

5.04.7

4.24.14.03.8

2.82.5

051015

Annual consultations per capita, 2007

Quality of careAsthma avoidable hospital admissions

Source: OECD Health Data.

United States

Korea

Finland

United Kingdom

New Zealand

Poland

Japan

Austria

Ireland

Belgium

OECD

Spain

France

Denmark

Norway

Iceland

Switzerland

Netherlands

Sweden

Germany

Canada

Italy

120

97

92

75

73

62

58

54

52

52

51

44

43

43

42

34

32

26

25

21

18

17

020406080100120140160

Age-sex standardised admission rates per 100 000 population aged 15 and over, 2007

Quality of care - Congestive heart failure avoidable hospital admission

Source: OECD Health Data.

Poland

United States

Germany

Austria

Italy

Finland

Sweden

France

OECD

Spain

New Zealand

Iceland

Ireland

Norway

Portugal

Netherlands

Belgium

Denmark

Switzerland

Canada

Japan

United Kingdom

Korea

474

441

352

331

308

306

289

276

234

234

206

202

192

188

176

171

169

165

155

146

134

117

110

0200400600

Age-sex standardised admission rates per 100 000 population aged 15 and over, 2007

Quality of care Ischemic stroke

Source: OECD Health Data.

In-hospital case-fatality rates within 30 days after admission 2007

2.3

2.4

3.1

3.2

3.3

3.7

3.7

3.8

3.9

4.2

5.0

5.6

5.9

6.2

6.3

6.5

6.6

7.5

7.6

9.0

0 5 10

Iceland

Korea

Denmark

Finland

Norway

Italy

Austria

Germany

Sweden

United States

OECD

Luxembourg

Netherlands

Czech Republic

New Zealand

Spain

Ireland

Slovak Republic

Canada

United Kingdom

Age-sex standardised rates per 100 patients

2. Measuring health care inputs

Spending on health care

Number of physicians

Remuneration and prices

Health care spending2008

Source: OECD Health Data 2010.

0

1000

2000

3000

4000

5000

6000

7000

8000

Public expenditure Private expenditureSpending per capita, US $ PPP

Health care spending (% of GDP)2008

Source: OECD Health Data 2010.

0

2

4

6

8

10

12

14

16

Public expenditure Private expenditure% GDP

Practising physiciansper 1000 population, 2007

Source: Health at a Glance 2009, OECD Indicators.

GreeceBelgiumNetherlandsNorwaySwitzerlandAustriaIcelandItalySpainSwedenCzech RepublicPortugalGermanyFranceDenmarkOECDSlovak RepublicIrelandFinlandLuxembourgAustraliaHungaryUnited KingdomUnited StatesNew ZealandPolandCanadaJapanMexicoKoreaTurkey

5.44.03.93.93.93.83.73.73.73.63.63.53.5

3.43.23.13.13.03.02.92.82.8

2.52.42.32.22.22.1

2.01.7

1.5

0246

Remuneration of general practitioners (GPs)2006 (2003 for the US)

Source: OECD Health Data.

Self-employed Salaried

0

20

40

60

80

100

120

140

160

180US $ PPP, thousands

Remuneration of specialists2006 (2003 for the US)

Source: OECD Health Data.

Self-employed Salaried

0

50

100

150

200

250

300

350US $ PPP, thousands

Comparative price levels for hospital servicesAverage of countries in the sample = 100

Source: F. Koechlin, L. Lorenzoni, P. Schreyer, Comparing Price Levels of Hospital Services Across Countries – Results of pilot study, OECD Health Working Paper No. 53 (2010).

163

140

123 121114 113

98

85

62 59 57

90

120

104112

121

101

118

83

103

7973

142

95

115

99

113

118108

69

82 81 81

0

20

40

60

80

100

120

140

160

180

Price levels for inpatient hospital services

Price levels for GDP

Per capita GDP level

Health care prices and volumes

Source: OECD Health Data.

-2-101234

Spending per capita

Doctors

Nurses

Students

MRIs

Scanners Hospital beds

Rem. nurses

Rem. GPs

Rem. spec.

HC prices

Canada United States OECD average

3. Deriving efficiency indicators

Identify health status determinants

Implement 2 methods (panel regressions and DEA) and various robustness checks

Complement the overall efficiency index with other performance indicators

Health care resources

Lifestyle factors: diet, alcohol & tobacco consumption

Socio-economic environment: income and education

Pollution

Health care status determinants

Panel regressions – Model specification (log form)

itHCR

iit

itDIET

itDRINK

itSMOK

ititGDPitEDUitAIRPOL

Panel regressions Contribution of main explanatory variables

to cross-country differences in life expectancy

Determinants

Spending Education Tobacco Alcohol Diet Pollution GDPCountry-specific effect

United States -0.5 2.9 0.5 0.0 0.0 0.0 -0.6 0.6 -4.0 Germany 0.6 0.8 0.4 -0.1 -0.1 0.0 0.5 0.1 -1.0 France 1.3 0.9 -0.2 0.0 -0.3 0.0 0.4 0.2 0.4 United Kingdom 0.5 -0.1 0.4 0.1 -0.2 0.0 0.1 0.2 0.0 Canada 1.8 0.9 0.4 0.1 0.1 0.0 -0.8 0.3 0.9 Czech Republic -2.7 -1.8 0.5 -0.1 -0.3 -0.1 0.0 -0.6 -0.3 Korea -0.6 -2.4 0.1 0.0 0.0 0.1 0.3 -0.4 1.7

Life expectancy

at birth

Source: Joumard , André, Nicq & Chatal (2008), "Health Status Determinants: Lifestyle, Environment, Health Care Resources and Efficiency ", OECD Economics Department Working Paper, No. 627.

Panel regressionsYears of life not explained by the model

With health care resources measured in monetary terms

-5

-4

-3

-2

-1

0

1

2

3

Source: Joumard , André, Nicq & Chatal (2008), "Health Status Determinants: Lifestyle, Environment, Health Care Resources and Efficiency ", OECD Economics Department Working Paper, No. 627.

DEA – Defining the efficiency frontier and potential efficiency gains

70

72

74

76

78

80

82

84

0 1000 2000 3000 4000 5000 6000

Efficiency frontier

Life expectancy at birth (years)

Total health care spending per capitaO

utpu

t ine

ffici

ency

Input inefficiency

DEA – Results and sensitivity analysis(for different outcomes)

Source: Joumard , André & Nicq (2010), "Health Care Systems: Efficiency  and Institutions", OECD Economics Department Working Paper, No. 769.

Potential gains in life expectancy, years Potential gains in amenable mortality, %

DEA – Results and sensitivity analysis(for different inputs)

Potential gains in life expectancy, years

Source: Joumard , André & Nicq (2010), "Health Care Systems: Efficiency  and Institutions", OECD Economics Department Working Paper, No. 769.

Comparing efficiency indicators derived from panel regressions and DEA

Panel regression (years)

DEA (years)

aus

aut

can

cze

dnk

finfra

deu

grc

hun

isl

irl

kor

nld

nzl

nor

polswe

che tur

gbr

usa

0

1

2

3

4

5

6

7

0 1 2 3 4 5 6 7DEA (years)

Panel regression (years)

Efficiency: DEA efficiency score and other performance measures

Source: OECD Health Data.

-3

-2

-1

0

1

2DEA score

EquityAll, in-patient care

Colorectal cancer

Lung cancer

Breast cancer

AMI

Femur fracture

Occupancy

TurnoverCataractCons./doctor

Adm. costs

DTP

Measles

Influenza

Asthma

Bronchitis

Heart failure

AMIStroke

Canada OECD average Group 2Average length of stay

Avoi

dabl

e ad

miss

ions

Vacc

inati

ons

Fata

lity

rate

s

Group 2: Australia, Belgium, Canada, France

Efficiency: a closer look at administrative costs

Source: OECD Health Data.

0

2

4

6

8

10

12Per cent of total health care spending

4. Reaping efficiency gains: the impact on public spending

Main assumptions:

• Health outcomes improve as they did in the past

• Two scenarios on the spending side are compared:

1. No reform scenario – spending increases as it did in the past

2. Reform scenario – efficiency gains are exploited and finance all or part of the improvement in health status

In most countries, potential savings in public spending are large

Exploiting efficiency gains would allow to improve health outcomes further

Exploiting efficiency gains would help to contain future spending

Potential savings in public spending

Source: OECD Health Data 2009; OECD calculations.

0

1

2

3

4

5

6% 2017 GDP

A new set of OECD indicators on health care policies and institutions (see Health Care Systems: Efficiency and Policy Settings)

Cluster analysis to identify health care models

The bad news is: there is no ideal system…

The good news is… there is no ideal system ! …

… No « big bang » reform is required to improve performance. Incremental reform can yield large benefits

5. Performance and institutionsIs there an ideal health system ?

Characterising health care systems:country groups

Source: Joumard, André & Nicq (2010), "Health Care Systems: Efficiency  and Institutions " , OECD Economics Department Working Paper. No. 769.

Linking efficiency with policy settingsNo health care system clearly outperforms the others

Source: Joumard, André & Nicq (2010), "Health Care Systems: Efficiency  and Institutions", OECD Economics Department Working Paper, No. 769.

Indicators of health care spending efficiency can be built and are relatively robust

The efficiency indicators can be complemented by indicators of the quality of care and other performance indicators

There is a large potential for efficiency gains in many OECD countries

No health care model clearly outperforms others Incremental reform is the way forward

Conclusions

Thank you !