Performance-Based Financing in Rwanda

Agnes Soucat, Adviser HNP Africa Region

Background (1)• Shortage of human resources for health

services• No cash resources in health facilities• Low levels of productivity and motivation

among medical personnel • Low user satisfaction & poor quality of service

lead to low use.• High levels of child and maternal mortality

2

Background (3)

• In 2005 , 4/10 births attended by a health professional.

• Infant Mortality : 86 per 1,000

• HIV : 3.1% Source: Rwanda 2005: results from the demographic and health survey. 2008. Studies in family planning, 39(2), pp. 147-152.

3

NATIONAL GOVERNMENT

LOCAL GOVERNMENT

AUTONOMOUS FACILITIES PROVIDERS

Clients / Citizens

VOICE

CLIENT POWER

Performance CONTRACTS

COMMUNITY HEALTH WORKERS

PROVIDERSCOMMUNITY HEALTH

INSURANCESMutuelles

COMMUNITYGOVERNANCE

Umushyikirano, Citizen Report

Cards, Ombusdman

PERFORMANCE BASED, CASH AND

IN KIND INVESTMENT INPUT

SUBSIDIES TRANSFERS

Strengthening accountability in the health sector in Rwanda

Rwanda has undertook major reforms to strengthen accountability of all institutional and individual actors for MDGs related results...

- Decentralisation of health services with strong governance structures based community participation.

- Imihigo: Performance contracts between President of Republic and mayor of Districts;

- PBF: Performance Based Financing;- CBHI: Community Health Insurance;- Autonomy of health facilities, including hiring and

firing of health personnel;

..through a shift of paradigm..

Decentralization

• Administrative, fiscal and financial decentralization has provided huge sums of money to local levels of government and given them much flexibility by providing them with block grants

Fiscal and Financial Decentralization

0

10,000,000,000

20,000,000,000

30,000,000,000

40,000,000,000

50,000,000,000

60,000,000,000

70,000,000,000

Year

Amou

nt in

RW

F

Transfers to Districts

CDF

Transfers to Provinces

Total health personnel in publicly funded facilities has almost doubled in 3 years …

HRH expenditures have progressively shifted to districts ...Distribution of HRH Expenditures

(in % of total)

7.69% 5.11% 3.62% 3.57% 5.19%

55.33% 57.83% 53.24% 53.32%

9.83%

36.98% 37.06% 43.14% 43.11%

84.98%

0%

20%

40%

60%

80%

100%

2003 2004 2005 2006 2007

Year

Minisante Transfer to Pub Facilities Provinces/Districts

Financing has more than tripled in four years (going from USD 7.5 to 30.3 millions, of which the PBF

has grown more than tenfold from USD 0.8 to 8.9 millions)

Rwanda: Scaling up of community health insurance

Proportion of individuals enrolled in health insurance

0102030405060708090

2002 2003 2004 2005 2006 2007 2008

%

Source: MOH Rwanda; 2005 EICV 2005

Results show Rwanda is now back on track towards the health MDGs…

• Health outcomes– Neonatal, infant and child mortality– Malaria incidence and mortality– HIV– Improved financial access– Reduction of catastrophic expenditures

• High Impact Interventions– ITNs– Family planning– Assisted Deliveries– Quality of care

Rwanda is back on track to reach the MDGs

Under five mortality trends with MDG target for 2015

0

50

100

150

200

250

1999 2001 2003 2005 2007 2009 2011 2013 2015

U5M

R p

er 1

,000

1990 level

MDG target for 2015

Observed

Trends since 1998

Trends required toreach the 2015 target

All income groups benefit but inequities still persist …

Under five mortality trends by income quintile (2005-2007)

211195

170

204

122

161149

132 141

84

0

50

100

150

200

250

Poorest Q uintile 2 Q uintile 3 Q uintile 4 Richest

U5M

R p

er 1

,000

DHS 2005 DHS 2007

Source: DHS 2005 and 2007.

At all income levels, those enrolled in

mutuelles are much more likely

to use health services.

26

4340

42

50

42

14

21

25

3133

25

0

10

20

30

40

50

60

Poorest 2 3 4 Best-off Overall

Quintiles

% o

f sic

k w

ho s

ough

t car

e fr

om m

oder

n he

alth

ser

vice

pro

vide

r

enrolled

non-enrolled

Source: Shimeles et al, 2009

Rwanda Health Insurance

Rwanda: Effect on MDGs High Impact Interventions

0102030405060708090

100

% delivered ina health

facility TOTAL

DPT3 (%) CurrentlyUsing any

modern FPmethod (%)

% U5 whoslept under anITN the past

night

%

2000

2007

Rwanda : Increase in utilization of high impact services

Proportion (%) of children under 5 years of age who have slept under a mosquito bed-net during the night preceding the survey

Increase in utilization of high impact services

Trends in assistance at delivery : Years 2000, 2005, 2007Percentage (%) of women delivered by a health professional

Table 1: Output Indicators (U’s) and Unit Payments for PBF Formula

OUTPUT INDICATORS Amount paid per unit (US$)

Visit Indicators: Number of …

1 curative care visits 0.18

2 first prenatal care visits 0.09

3 women who completed 4 prenatal care visits 0.37

4 first time family planning visits (new contraceptive users) 1.83

5 contraceptive resupply visits 0.18

6 deliveries in the facility 4.59

7 child (0 - 59 months) preventive care visits 0.18

Content of care indicators: Number of …

8 women who received tetanus vaccine during prenatal care 0.46

9 women who received malaria prophylaxis during prenatal care

0.46

10 at risk pregnancies referred to hospital for delivery 1.83

11 emergency transfers to hospital for obstetric care 4.59

12 children who completed vaccinations (child preventive care) 0.92

13 malnourished children referred for treatment 1.83

14 other emergency referrals 1.83

Quality Conceptual Framework

What They Know (Ability/Technology)

What They Do: (Quality)

Production Possibility Frontier

Goal: Use Pay for Performance to

Close Productivity Gap

Ability/Technology

What They Do

Production Possibility Frontier

Productivity Gap Conditional on Ability

Actual Performance

Researcher & Policy Maker Collaboration

• Research Team– Paulin Basinga, National University of Rwanda– Paul Gertler, UC Berkeley– Jennifer Sturdy, World Bank and UC Berkeley– Christel Vermeersch, World Bank

• Policy Counterpart Team– Agnes Binagwaho, Rwanda MOH and CNLS– Louis Rusa, Rwanda Rwanda MOH– Claude Sekabaraga, Rwanda MOH– Agnes Soucat, World Bank

A collaboration between the Rwanda Ministry of Health, CNLS, and SPH, the INSP in Mexico, UC Berkeley and the

World Bank

Evaluation Questions: Did PBF…

– Increase the quantity of contracted maternal health services delivered?

– Improve the quality of contracted maternal health services provided?

Evaluation Design

• During decentralization, phased rollout at district level

• Identified districts without PBF in 2005• Group districts into “similar pairs” based

on population density & livelihoods

• Randomly assign one to treatment and other to control

• MOH reallocated some districts to treatment• With decentralization, some new districts had PBF in an area of

the new district – must be treatment

• Unit of observation is health facility26

Rollout of PBF in health centers in Rwanda, 2006 – 2008

27

Isolating the incentive effect

• PBF – Performance incentives– Additional resources

• Compensate control facilities with equal resources– Average of what treatments receive– Not linked to performance– Money allocated by the health center management

Sample: Panel 165 Facilities 2006-08

• 2145 households in catchment areas– Random sample of 14 per clinic

Log Expenditures

• Randomization balanced baseline • Follow-up balanced, so difference in follow-up

outcomes due to incentives not resources

Year Treatment Control Difference P-Value15.812 15.612 0.200(1.042) (1.007) 0.241

16.906 16.989 -0.083(0.71) (1.08) (0.14)

2006 0.418

2008 0.568

Baseline Expenditures & Staffing

Impact of PBF: Statistical methods

• Have balance at baseline on all key outcomes• Use difference in differences analysis

– Not a pure randomized experiment

• Clustered at district-year level• Facility Fixed Effects• Year dummy• Controls: age, parity, education, household size,

health insurance, land, value of assets

36.3

49.7

34.9

55.6

30.0

40.0

50.0

60.0

Baseline (2006) Follow up (2008)

Prop

ortio

n of o

f ins

tituti

onal

deliv

erie

s

Control facilities Treatment (PBF facilities)

7.3 % increasedue to PBF

35

Delivery at the health facility increased overall in Rwanda, but 7% more in PBF facilities ….

Prenatal Competency & Quality

• Provider knowledge/competency– Standardized vignette presented to provider– Compare answers to Rwandan CPG– Measure of ability/knowledge

• Process quality – Patient exit interview of clinical services provided– Clinical content of care– Provider effort

-0.10

0

-0.13

0.15

-0.15

-0.10

-0.05

0.00

0.05

0.10

0.15

0.20

Baseline (2006) Follow up (2008)

Stand

ardiz

ed Pr

enata

l effo

rt sc

ore

Control facilities Treatment (PBF facilities)

15 % Standard deviation increase due to PBF

37

In the last years, PBF has increased prenatal care quality significantly …

.2.3

.4.5

Pra

ctic

e

.3 .4 .5 .6 .7 .8Competence

Control facilities

Treatment facilities

Kernel Non parametric regression practice-competency at baseline

.2.3

.4.5

Pra

ctice

.3 .4 .5 .6 .7Competence

Control facilities

Treatment facilities

Kernel Non parametric regression practice-competency at follow up

Impact of PBF on Prenatal Care Quality

Impact of PBF on Child Health (z-scores)

Results Summary

• Balanced at baseline• Expenditures same, so isolate incentives• Impact on utilization

– Delivery & Child prevention, but not prenatal

• Impact on prenatal quality– Bigger for better doctors

• Reduced child morbidity & Taller children• Effect sizes bigger than

most other interventions

Discussion• PBF Effect seen despite many other national level

intervention: possible bigger effect in other countries

• Increase in utilization in country with national campaigns:– Mutuelle– Imihigo– HIV services– Safe motherhood and PCIME

• Possible spill over effect to child health

45

Discussions/ Policy implications

• You get what you pay for !

• Returns to effort important– Bigger effects in things more in provider’s control– Patient or community health workers for prenatal

care/Immunization• Provide incentives directly to pregnant women? (conditional

cash transfer program).• Financial incentive to community health workers • Low quality of care : additional training coupled with PBF

• Need to get prices “right”• Evaluation feedback useful

46

Limitations• The original randomized designed was

changed due to the political decentralization process: But sample well balanced!

• Trend analysis with HMIS data ongoing• No measure of all paid and some non paid

indicators : HMIS analysis• Cost effectiveness analysis

47

Acknowledgments

• Funding by:– World Bank– Government of Rwanda (PHRD grant)– Bank-Netherlands Partnership Program (BNPP)– ESRC/DFID– GDN