Pay-for-Performance (P4P) for Health Services in Rwanda

1
Pay-for-Performance (P4P) for Health Services
in Rwanda
A collaboration between the Rwanda Ministry of
Health, CNLS, SPH, INSP Mexico, UC Berkeley and
the World Bank

• Paulin Basinga
• Rwanda School of Public Health

2
Our team

• 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 MOH
• Claude Sekabaraga, Rwanda MOH
• Agnes Soucat, World Bank

3
The 2005 starting point

• Professionally assisted births 40
• Maternal Mortality 750 per 100,000 live births
• 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.
4
Why a pay reform?
5
P4P for Health in Rwanda

• Objectives
• Focus on maternal and child health (MDGs 4 5)
• Increase quantity and quality of health services
  provided
• Increase health worker motivation
• What?
• Financial incentives to providers
• For more quantity
• And more quality
• How?
• Contracts between government health facilities
• When?
• Piloted in 2001-2005, full scale from 2006

6
(No Transcript)
7
(No Transcript)
8
Evaluating P4P in RwandaEvaluation design
9
Conceptual framework for quality
Production Possibility Frontier
What They Do (Quality)
Productivity Gap Conditional on Ability
Actual Performance
What they know (Ability/Technology)
10
Evaluation Questions

• Did P4P improve
• the quality and quantity of maternal and child
  health services?
• the health of the population?

11
History of P4P in Rwanda

• Three pilot schemes
• Cyangugu (since 2001)
• Butare (since 2002)
• BTC (since 2005)
• National model implemented in 2006

12
Evaluation Design

• Phased roll-out at district level
• Identified districts without P4P in 2005
• Group districts into similar pairs
• based on population density, location
  livelihoods
• Randomly assign one to treatment and other to
  control
• Unit of observation is health facility

13
A few challenges

• The decentralization surprise
• MOH reallocated some districts to treatment
• A few new districts had some facilities with P4P
  must be treatment

14

• Rollout of P4P
• 2001-2005
• 2006 2008

15
Sample

• Out of 30 districts
• 12 Phase I (treatment)
• 7 Phase II (comparison)
• 165 health facilities
• All rural health centers located in 19 districts
• 2156 households in catchment areas
• Power calculations based on expected treatment
  effect on prenatal care visits, institutional
  delivery
• Panel data 2006 and 2008

16
Econometric model

• Basic difference-in-differences model specified
  as a two-way fixed effect cross-sectional
  time-series regression models.

• where
• Yijt is the outcome of interest for individual i
  living in facility js catchment
• area in year t
• PBFj,2008 1 if facility j was paid by PBF in
  2008 and 0 if otherwise
• ?j are facility fixed effects
• ?2008 1 if the year is 2008 and 0 if 2006
• Xitk are time varying individual
  characteristics
• ?ijt is a zero mean error term.

17
Evaluation design challenges

• Organizational
• Managing expectations
• The John Henry effect in practice
• Building capacity
• Time commitments
• Technical
• Small sample size (clusters at district level
  unit of operation!)
• Reconciling provider and client data

18
Facilitating factors

• Government leadership
• Integration
• Government coodination of parners

19
Baseline, health facilities
20
Baseline, utilization of maternal health services
21
Baseline, women 15-49 with birth in last 24 months
22
Evaluating P4P in RwandaEvaluation results
23
Impact on structure quality
Impact of PBF Impact of PBF Time Trend Time Trend
N Treatment 2008 (1) SD 2008 (1) SD
Availability of vaccines 155 0.703 (0.355) -0.514 (0.122)
Prenatal care service 155 0.062 (0.426) -0.659 (0.100)
Delivery service 155 0.239 (0.286) -0.484 (0.083)
24
(No Transcript)
25
(No Transcript)
26
Impact on quality of prenatal care
Impact of PBF Impact of PBF Time Trend Time Trend
N Treatment 2008 (1) SD 2008 (1) SD
Total Prenatal Quality Score (standardized score) 3683 0.157 (0.048) 0.090 (0.032)
Tetanus vaccine during prenatal visit (1) 2810 0.054 (0.023) 0.032 (0.028)
Productivity ratio 3757 0.072 (0.017) 0.036 (0.012)
Treatment 2008 Competency gt 75 (1) 3739 0.088 (0.019) NA NA
27
Impact on quality of prenatal care
28
Impact on use of prenatal care
Impact of PBF Impact of PBF Time Trend Time Trend
N Treatment 2008 (1) SD 2008 (1) SD
Number of prenatal visits received 2223 -0.028 (0.066) 0.237 (0.056)
Likelihood of first prenatal visit in first trimester 2223 0.020 (0.036) 0.163 (0.029)
Made 4 or more prenatal care visits 2223 0.009 (0.039) 0.121 (0.023)
29
Impact on use of maternal services
Impact of PBF Impact of PBF Time Trend Time Trend
N Treatment 2008 (1) SD 2008 (1) SD
Institutional delivery (1) 2108 0.077 (0.034) 0.134 (0.023)
Delivery attended by Qualified Provider (1) 2274 0.083 (0.035) 0.195 (0.028)
Use of any modern contraceptive method (1) 3121 -0.017 (0.023) 0.245 (0.033)
30
Impact on institutional delivery
31
What our results tell us

• You get what you pay for !
• Returns to effort important
• Bigger effects in things more in providers
  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 P4P
• Need to get prices right
• Evaluation feedback useful

32
Discussion

• Prenatal care entry point!
• Increase in utilization nationwide due to
• Mutuelle
• Imihigo
• HIV services
• Safe motherhood and PCIME
• Possible spill over effect to child health

33
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

34
Thank you!
35
Acknowledgments

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