Performance-Based Equitable Resource Allocation Model: The Punjab Experience
October 2013 Maternal, Newborn and Child Health Working Paper UNICEF Health Section, Program Division
unite for children
Performance-Based Equitable Resource Allocation Model: The Punjab Experience ©United Nations Children’s Fund (UNICEF), New York, 2013
Knowledge Management and Implementation Research Unit, Health Section, Program Division UNICEF 3 UN Plaza, New York, NY 10017 October 2013
This is a working document. It has been prepared to facilitate the exchange of knowledge and to stimulate discussion. The findings, interpretations and conclusions expressed in this paper are those of the authors and do not necessarily reflect the policies or views of UNICEF or of the United Nations. The text is not edited to official publication standards, and UNICEF accepts no responsibility for errors. The designations in this publication do not imply an opinion on legal status of any country or territory, or of its authorities, or the delimitation of frontiers. The editors of the series are Theresa Diaz and Alyssa Sharkey of UNICEF Program Division. For more information on the series, or to submit a working paper, please contact
[email protected], or
[email protected].
ii
MATERNAL, NEWBORN AND CHILD HEALTH
WORKING PAPER October 2013
Performance-Based Equitable Resource Allocation Model The Punjab Experience
Afeef Mahmood, Katrina Estacio, Nuzhat Rafique, Thomas O’Connell
Keywords: health system strengthening, district-level, district health system strengthening, performance, planning, financing, performance-based financing, performance-based planning, resource allocation, equity, universal health coverage Comments may be addressed by email to Afeef Mahmood:
[email protected] cc:
[email protected],
[email protected] and
[email protected]
iii
Acknowledgements We thank Mr. Fawad Hassan (Secretary of Health, Punjab), Mr. Farasat Iqbal (Project Director Punjab Health Sectors Reform Program, DoH Punjab), Dr. Anwer Janjua (Director Management Information System, DoH Punjab), Rie Hiraoka (Senior Health Specialist, Asian Development Bank), and Linda Arthur (Senior Social Sector Specialist, Asian Development Bank) for their substantial contributions to the conception and development of the Punjab Resource Allocation Model. We also thank Alyssa Sharkey (Health Specialist, Knowledge Management and Implementation Research Unit, UNICEF) and Julia Kim (Senior Health Advisor, Knowledge Management and Implementation Research Unit, UNICEF) for their invaluable input in providing appropriate revisions and comments. We thank Mary Beth Woodin for patiently editing various drafts, and keeping excellent track of the numerous comments received.
iv
Table of Contents
Acknowledgements .............................................................................................................. iv Table of Contents .................................................................................................................. v List of Acronyms ................................................................................................................... vi Executive Summary ............................................................................................................. vii Introduction .......................................................................................................................... 1 Punjab State Model for Resource Allocation .......................................................................... 5 Punjab Model: An Illustrative Example............................................................................................9
Discussion ........................................................................................................................... 14 References .......................................................................................................................... 16
v
List of Acronyms
ADB
Asian Development Bank
DHSS
District Health Systems Strengthening
GoPb
Government of Punjab
HMIS
Health Management Information System
HSRP
Health Sector Reform Programme
LQAS
Lot Quality Assurance Sampling
MDGs
Millennium Development Goals
MoU
Memorandum of Understanding
PBF
Performance-Based Financing
PBP
Performance-Based Planning
PMDGP
Punjab Millennium Development Goals Program
UHC
Universal Health Coverage
WHO
World Health Organization
vi
Executive Summary Achieving Universal Health Coverage (UHC) with equity requires addressing the needs of the most marginalized and vulnerable populations within countries. Although many countries are making progress towards the achievement of Millennium Development Goals (MDGs) 4 and 5, intra- and inter-country inequities often persist. Moving forward, strategies need to focus on reducing intra-country inequities if progress toward the MDGs is to accelerate and UHC is to be achieved. One key challenge lies in the design and implementation of context-specific performance−based financing strategies, at national and sub-national levels, that align with performance−based planning. This paper describes a performance−based model used in the Punjab province of Pakistan to promote equity in access and use of services, and improve performance as a core tenet of health planning and financing. The model was initially implemented in only a few Punjab districts in 2010, and then was followed by provincial and national rollouts. The goal of the “Performance−Based Equitable Resource Allocation Model” is to implement a financial system capable of allocating resources to districts based on local needs, while simultaneously rewarding them with additional resources for improvements in district health performance. The model focuses on three objectives: 1. Provide a system−wide equitable resource allocation mechanism; 2. Allocate rewards for good performance while addressing equity and meeting essential local resource requirements; and 3. Maintain transparency, accountability and predictability of resource allocation.
vii
Introduction Intra-country inequities hinder achievement of Millennium Development Goals and Universal Health Coverage. Achieving UHC with equity requires addressing the needs of the most marginalized and vulnerable populations within countries. Though many developing countries are making progress toward the Millennium Development Goals (MDGs), intra-country inequities persist, particularly in those countries that are falling short on MDG 4: reduction in child mortality and MDG 5: improvement in maternal health (UNICEF 2010b; Lozano, Wang et al. 2011). In 2008, more than 50% of all maternal deaths occurred in countries in Sub-Saharan Africa and South Asia (Hogan, Foreman et al. 2010). An equity-focused approach has been proposed as a means to accelerate progress towards MDG 4 and 5, and consequently prevent millions of maternal and child deaths (UNICEF2010a). Looking ahead, policies should move beyond the MDGs to enable significant reductions in intra-country, not just inter-country, inequities to achieve UHC (Gyorkos, Joseph et al. 2009; UNICEF 2010b; Verguet, Jassat et al. 2012) Health system strengthening is key to reducing inequities and improving health coverage, particularly among the most marginalized populations. WHO reports that overcoming health system weaknesses can lead to improvements in health outcomes (WHO 2007). However, initial health system strengthening models, focused primarily on reforming national strategies and lacked a clear linkage to operational strategies to improve district and sub-district service delivery (Travis, Bennett et al. 2004). One major challenge is the overreliance on national aggregated data to assess progress, rather than appropriately disaggregated statistics to identify patterns of inequity linked to income, gender, geographic location, and other important social determinants of health (Kim, Baum et al. 2011). Yet, granular local data is considered necessary to assess root causes of poor performance and uncover inequities at the district level (Marmot, Friel et al. 2008). Resolving intra-country inequities requires coherence between pro-poor and other equity-based planning and financing, including the concurrent creation of incentives to achieve UHC targets. Such a strategy necessitates alignment of financing with planning at sub-national and national levels, in order to reorient national health system strengthening and reforms towards achieving UHC with equity. Performance-based financing is one strategy used to support district level health system strengthening. There is a growing recognition that achieving equity-focused outcomes requires that Performance-Based Financing (PBF) is consistent with Performance-Based Planning (PBP), that is to say “linking burden with budget allocation” (Walley, Lawn et al. 2008). One approach is to use District Health Systems Strengthening (DHSS) (Bucagu, Kagubare et al. 2012; Dynes, Buffington et al. 2013; Liljestrand and Sambath 2012) as both a vehicle for delivering UHC, and as a way to craft supply-side incentives that reward greater access to quality and effective services.
1
Finance strategies must align with evidence-based district plans in order to adequately fund efforts to achieve UHC with equity. There are only a few documented attempts to align planning and financing in districts. One example, in Zambia, involved transferring control of expenditures to the district level and instituting an allocation formula based on district population size, density, and hospital beds, to disburse financial resources to districts (Bossert, Chitah et al. 2003). In Rwanda, performance-based contracts prompted district health system changes that resulted in an increase in service utilization, financial accessibility and motivation of health staff (Soeters, Habineza et al. 2006). The challenge remains to design and implement context-specific PBF that aligns with PBP well enough to enhance district health system performance and foster UHC with equity. Performance-based financing, in the context of a decentralized health system strengthening approach, aims to simultaneously achieve greater equity and improve district health system performance. Most of the literature concerning performance-based incentives focuses on improving individual health provider’s performance, and only recently, publications have addressed the impact of PBF on district level performance in low-resource settings, such as in Rwanda (Rusa 2009; Meessen, Soucat et al. 2011) and Congo (Soeters, Peerenboom et al. 2011). A resource allocation model that addresses both needs and performance has the potential to incentivize districts to address equity gaps. This paper describes such a model, currently used in the Punjab province of Pakistan, designed to bring equity and performance into the mainstream of policy planning and financing. The intention of the model is to distribute heath sector funds based on the needs of each district and to reward enhanced performance.
2
Brief overview of Punjab’s health system In 2006, the Punjab provincial government of Pakistan initiated a Health Sector Reform Programme (HSRP) to improve the quality and coverage of services at primary care facilities towards achievement of the MDGs (Javed and Amin 2007; Government of Punjab 2009a). An analysis of Punjab healthcare expenditures revealed a steady decline in the proportion of the district budget allocated for health, from 14.49% (2002-2003) to 12.46% (2005-2006) (Mahmood 2009b). While the HSRP set the tone for policy and priority setting, addressing financial resource constraints was not a primary focus of the reform. In Pakistan, historically, the Federal Government is responsible for national health policy and planning along with prioritizing primary and secondary health care interventions. In August 2001, the Government announced a far-reaching Management and Organization Reform Agenda for reconstructing state institutions through the establishment of a democratically elected system of District Governments. This led to devolution of service delivery and health financing from the provincial level to the districts. District health system financing is now managed directly by the district assemblies. This reform proved difficult to implement as district-level governments struggled to absorb both administrative and management responsibilities. Other challenges arose due to budget limitations, varying priorities across sectors, political interference and a lack of management expertise (Javed and Amin 2007). The 18th Constitutional Amendment Bill of 2010−2011 (Government of Pakistan 2010) further devolved health sector functions to provinces, making them fully responsible for all aspects of healthcare system legislation, which were previously under the Federal Government’s domain. Planning by provincial and district health departments follows two distinct but poorly aligned types of financing mechanisms: one for the annual operational budget (i.e., salaries, drugs, supplies, diagnostics, and other operating expenditures), and another for capital development (i.e., construction of new hospitals, increasing existing service provision capacities, and in some cases, preventive vertical programs) (Mahmood 2009a). In developing provincial health plans, equity is not always considered and the distribution of social and health indicators among sub-groups and localities is poorly understood. Statistics used for planning are commonly presented as means or aggregates at the provincial level. As a result, it is challenging to develop health plans that target the specific needs of a district (Ghaffar, Kazi et al. 2000). In recent years, Punjab’s progress towards the achievement of MDGs 4 and 5 has been less than satisfactory. In 2007 – 2008, for example, the under-five mortality rate was 111 per 1,000 live births and the infant mortality rate was 77 per 1000 live births, far from the MDG target of 52 per 1000 and 48 per 1000 live births, respectively (Government of Punjab 2009). Similarly, in 2008, Punjab had an estimated maternal mortality rate of 276 per 100,000 live births against the target of 140 per 100,000 live births (Government of Pakistan 2005). Based on the rate of progress for these indicators, it was questionable whether Punjab would attain the MDG targets, unless health sector reforms were rigorously implemented (Sohani, Rafique et al. 2009).
3
The Government of Punjab (GoPb) had growing concern that the level of health sector funding was insufficient (Mahmood 2009b). Roughly 70 to 80% of funding was spent on salaries, with only 20% remaining to fund supplies and operations (Mahmood 2009b). In an effort to measure the financial resources required to improve coverage of high impact interventions and achieve MDGs 4 and 5, the GoPb defined a cost-effective set of minimum service delivery standards (Mahmood 2009b). And, at the GoPb’s request, a loan application was submitted to the Asian Development Bank (ADB) in order to meet these newly defined delivery standards. Against this backdrop, the GoPb and the ADB developed the “Punjab Millennium Development Goals Program” (PMDGP) to improve Punjab’s health service delivery through effective policy−making, investments in health system strengthening, and narrowing of financing gaps (Government of Punjab 2009b). In the midst of the health sector reforms, the GoPb realized that 36 districts were limited in their capacity to absorb new funds, in large part due to a budget allocation process that did not link funding to specific health care needs of the population. As an example, districts with large hospitals automatically receive higher budget allocations, regardless of occupancy levels. Consequently, simply providing more financial resources did not ensure that coverage, and more importantly equity of coverage, would improve. The traditional supply-side approach of allocating the majority of funds based on a district’s population size was deemed unsuitable since equity or performance are not taken into account; instead funding is based entirely on health facility needs, including gaps in supplies, infrastructure and equipment. To address these limitations, the Department of Health considered various options for allocating additional resources to the Punjab districts. After long deliberations, in 2010, the GoPb decided on an allocation mechanism that addresses base needs and provides incentives with rewards for districts that achieve measureable results toward attainment of UHC with equity. This innovative approach to financing district health systems is termed the Performance−Based Equitable Resource Allocation Model. This is a resource allocation model, that 1) recognizes that not all districts are at the same level of achievement of health targets, 2) provides financial resources to ensure that basic infrastructure and operational mechanisms are in place, and 3) rewards individual districts for health indicator improvements, relative to baseline values, rather than focus solely on one target for all districts.
4
Punjab State Model for Resource Allocation The overall concept of a “performance−based equitable resource allocation model” is to have a financial reward system that allocates resources to districts based on local needs while simultaneously rewarding them for improvements in district health performance. The model focuses on three objectives: 1) Provide a method (formula-based) for state-wide equitable resource allocations to all districts; 2) Allocate rewards to districts for good performance, while addressing equity and reflecting local needs; and 3) Maintain transparency, accountability and predictability of resource allocation. Figure 1. Punjab Model
The Performance−Based Equitable Resource Allocation Model used in Punjab (referred to as the “Punjab Model” in this paper) divides the allocation of funds to each district into Base Allocation and Performance components (Figure 1). In year one, the total amount of funds for each district is determined through the application of a “Needs Index” (Table 1). For the first year, each district receives the entire calculated amount.
5
Year 1
Year 2
Calculate Needs Index
Evaluate Year 1 Performance to Calculate Performance Allocation
Define Performance Indicators
Re-calculate Needs Index
Meet Minimum Conditions
Define Performance Indicators
Distribute Base Allocation
Distribute Base Allocation & Performance Allocation
Monitor Performance
Monitor Performance
The aim is to ensure districts are able to address high priority infrastructure gaps and operational challenges. This system strengthening seeks to provide the foundation for accelerated progress in future years. In the second and subsequent years, the amount calculated for each district under the Needs Index is divided into a Base Allocation, which comprises 70% of the total amount, and a “Performance Allocation” which comprises the remaining 30% of the total amount. The amount of the Performance Allocation that a given district receives (i.e. from 0% to 30%) is based on progress achieved in the prior year toward predetermined key performance indicators. To define the Needs Index, four attributes of the health system for each district are assessed and given a weight by the State. The weights chosen reflect an equity dimension (social deprivation and mortality) as well as factoring in unit costs (actual number of facilities and rural persons). The weight determines the amount of funds distributed to a district and results in a more equitable and needs-based allocation of funds across districts. For instance, changes in the number of health facilities will have four times greater impact on the total funds a district receives than changes in the maternal and child mortality index. The Needs Index was developed in response to the previous system of distributing funds, which was not perceived as transparent or needs-driven. How funds are distributed in Punjab using the Needs Index is illustrated in Table 1. Table 1. Application of the Needs Index in Punjab State Needs Index Attributes Number of Health Facilities a Social Deprivation Index b Rural Population c Maternal and Child Mortality Index d a
b
c
d
Proportion of funds allocated 40 % 25 % 25 % 10 %
A greater number of health facilities within a district requires proportionally more funds to buy medicines, purchase equipment and perform maintenance and repairs; assuming utilization rates are similar across districts. Specific indicators used to calculate the Social Deprivation Index include adult literacy, primary school enrollment, breastfeeding practices, skilled birth attendance, modern contraceptive use, adequate water and sanitation access. The estimated rural population headcount in each district is used, rather that the percent of persons living in rural areas. District Maternal Mortality Rates and Under-Five Mortality Rates, though not weighted by district population.
Prior to receiving any funds from the state, districts are required to meet a set of minimum conditions, including a signed Memorandum of Understanding between the Punjab State government and district authorities. This sets clear district-specific targets that will be used in assessing the performance allocations. Thus, from the first year, all districts are aware of the indicators that will be used to evaluate their performance in the following years. These conditions are summarized in Table 2.
6
Table 2. Minimum conditions for transfer of funds to a district Minimum Conditions 1. Sign a Memorandum of Understanding (MoU) between provincial and district governments, whereby agreeing to: • Achieve standards or targets (as measured by performance indicators) • Report on utilization of funds • Report on key agreed upon outputs 2. Prepare a 3-year rolling plan with clearly defined areas for use of funding. 3. Update plans on a yearly basis. a
The design of these conditions can vary depending on the specific requirements of each health program within a given district (e.g. immunization, HIV/AIDS, TB control, antenatal care, etc.)
In year two of implementation, the Punjab Model shifts from the initial 100% Base Allocation of the first year, to a 70% Base Allocation and the opportunity to earn up to 30% based upon performance. The Base Allocation, still calculated using the Needs Index, is reduced to 70% of the total funds a district is eligible to receive, and the remaining 30% is disbursed based on the Performance Allocation (Figure 2). Figure 2. Punjab Model: Base and performance allocations
Resource Pool 50 million
Year 1 All funds distributed as Base Allocation
Year 2 Funds distributed as Base and Performance Allocation
Base Allocation 100%
Base Allocation 70% (based on Needs Index)
Performance Allocation 30% (based on performance
The Performance Allocation component is computed by measuring the district’s improvements along pre-defined performance indicators, as specified in the MoU (Table 3). The indicators should be defined with caution since they are used to measure progress and calculate the amount of resources awarded. The intention of the performance-based component is to incentivize districts to attain improved performance, regardless of baseline conditions. The model recognizes the difficulty and expense for districts with higher baselines to move towards their targets. As an example, it is easier to move from 20 to 40% coverage, than to move from 70 to
7
75% coverage (Hailu and Tsukada 2011). It is expected that the Performance Allocation for each district will steadily increase over time, as key indicators improve. Table 3. Performance Indicators Performance Indicators 1. Proportion of pregnant women registered for antenatal care 2. Proportion of pregnant women receiving antenatal visits(s) 3. Proportion of women delivered by Skilled Birth Attendants 4. Proportion of women who delivered at a health facility 5. Proportion of fully immunized children between 18 and 30 months of age 6. Tetanus Toxoid II coverage in pregnant women 7. Population coverage by Community Midwives 8. Human Resources Index (availability of essential Maternal Neonatal and Child Health staff) 9. Drugs Index (availability of essential drugs at facilities) 10. Equipment Index (availability of essential Maternal Neonatal and Child Health related equipment at facilities)
8
Punjab Model: An Illustrative Example This section presents an illustrative example (using dummy data) of how to calculate Base and Performance Allocations using the Punjab Model. In a given province 'Z', there are five districts: A, B, C, D, and E. Government funds available for allocation to the districts over a 3-year period are as follows: Year 1 = $ 5 million Year 2 = $ 7 million Year 3 = $ 8 million In order to distribute the funds equitably among the five districts, the government applies the Punjab Model as described below. Year 1: Disbursement of Funds via Base Allocation First year funding is solely Base Allocation and each district’s allocation is computed in a threestep process (Table 4 and Table 5). Step 1 Define Needs Index attributes and assign each a percentage of the total funding (Needs Index Allocations). (In this example, mortality is left out to simplify the calculations, though in practice, maternal and child mortality also are used to calculate the Needs Index) Step 2 Calculate funding for each Needs Index attribute as follows: Needs Index Funding = Total Province Annual Budget x Needs Index Allocation Table 4. Steps 1 & 2: Needs Index Funding Calculations Needs Index Attributes Number of health facilities Social deprivation index Rural population
Needs Index Allocation 35 % 35 % 30 %
Needs Index Funding $5M x 35 % = $1.75M $5M x 35 % = $1.75M $5M x 33 % = $1.50M
Step 3 For every attribute, calculate the amount allocated for each district. District Allocation (per attribute) = (District measure/Province Total) x Needs Index Funding Total District Base Allocation = Sum of District Allocations for Each Attribute The following example illustrates the calculations to determine the amount of funds allocated to health facilities in District A (Table 5): District Allocation for Health Facilities in A = (20 district health facilities /62 province health facilities) x $1.75 = $0.56M Total District A Base Allocation = $0.56 + $0.46 + $0.48 = $1.5M
9
In this example, District A receives a total of 1.5M apportioned as follows: $560,000 for health facilities, $460,000 for Social Deprivation and $480,000 for Rural Population. Table 5 illustrates calculations for all Needs Index attributes, for all districts. The Total Allocation column indicates how the 5M total funds are dispersed as Base Allocations across the five districts in Year 1. Table 5. Step 3: District-Level Base Allocation Calculations for Year One District Measure No. Health Facilities
District Allocation Health Facilities*
District Measure Social Deprivation
District Allocation Social Deprivation*
District Measure Rural Population*
District Allocation Rural Population*
Total Allocation*
A
20
$0.56
60%
$0.46
10
$0.48
$1.50
B
12
$0.34
30%
$0.23
6
$0.29
$0.86
C
8
$0.23
70%
$0.53
4
$0.19
$0.95
D
6
$0.17
50%
$0.38
3
$0.15
$0.70
E
16
$0.45
20%
$0.15
8
$0.39
$0.99
Province Total
62
$1.75
230%
$1.75
31
$1.50
$5.00
District
* Millions
Year 2: Disbursement of Funds via Base and Performance Allocations During the second year (and subsequent years) of implementation, districts receive funding from both Base and Performance Allocations. For year 2 $7 million is available for distribution among the five districts: in this instance, the government selects 85% ($5.95 million) to be transferred as the Base Allocation, with the remaining 15% ($1.05 million) set as the Performance Allocation. Base Allocations are determined exactly as described above. However, each district’s Performance Allocation is computed using the following 3-step process. Step 1 Determine allocation to district groups based on the percentage of Performance Indicators improved (Table 6): • Based on improvements in indicators, classify districts into groups according to performance level. • Compute Performance Allocation on a group basis, with higher performing districts receiving higher allocations • Determine Incentive Bonus for high performing group(s). • Calculate Performance Allocation for each group by summing Incentive Bonus (if applicable) and Group Proportion. Computations are as follows: Group Proportion = Number of Districts in Performance Group / Total Number of Districts Group Share = (Group Proportion x Incentive Bonus) + Group Proportion Group Performance Allocation = Group Share x Total Allocated for Performance in Province 10
Table 6. Performance Classification and Group Performance Allocations Performance Group
% of Indicators Improved
Number of Districts
Group Proportion
Incentive Bonus
Group Share
Group Performance Allocation*
Very Good
90% or more
2 (A, E)
40%
20%
48%
$0.50
Good
61% - 89%
3 (B, C, D)
60%
None
52%
$0.55
Average
41% - 60%
−
−
−
−
Acceptable
31% - 40%
−
−
−
−
Poor
30% and less
− 100%
100%
$1.05
Total
5 districts
*Millions
The following example illustrates application of the computations for the Very Good Performance Group (Districts A and E). Group Proportion = 2 Districts in Very Good Group / 5 Districts = 40% Group Share = (40% Group Proportion x 20% Incentive Bonus) + 40% = 48% Group Performance Allocation = 48% x 1.05M = $0.50 In this case, 48% of the funds allocated for performance in the province are awarded to the Very Good districts (A and E), and the remaining 52% is distributed to the Good districts (B, C, D) Step 2 Once the amount for each performance group is calculated, use the results to calculate the proportionate share for each district within a performance group (Table 7). Proportionate Share for District = End-line – Baseline / Sum of Differences within Performance Group Example District A: (60 End-line) – (40 Baseline) / 25 = 0.8 Proportionate Share District A Table 7. Proportionate Share for Each District Baseline Coverage*
End-line Coverage*
Difference
Proportionate Share for District
District A
40
60
20
0.8
District E
80
85
5
0.2
Performance Group Very Good
Sum of Differences within Performance Level Group
25
Good District B
65
70
5
0.19
District C
30
50
20
0.74
District D
50
52
2
0.07
Sum of Differences within Performance Level Group
27
* In this simplified example, an average coverage rate is used. In practice, the difference between base and end−line for each indicator would be used to determine the proportionate share of the performance allocation for each district.
11
Steps 1 and 2 only take into account improvements in performance indicators, irrespective of baseline conditions. As mentioned above, the model recognizes the difficulty for districts with higher baselines to move towards their targets. To account for these challenges, a Scale-Based Score was derived to assign weights depending on a district’s baseline values (Table 8). Steps 3 and 4 are included to adjust performance allocations to reflect differences in baseline conditions. Table 8. Scale-Based Score Baseline Score (S)
1-10% 1.1
11-20% 1.2
21-30% 1.3
31-40% 1.4
41-50% 1.5
51-60% 1.6
61-70% 1.7
71-80% 1.8
81-90% 1.9
91-100% 2
Step 3 Using the Scale-Based Score, compute the Composite Index and Performance Share for each district (Table 9). Composite Index (CI) = Proportionate Share for District + Scale-Based Score Performance Share (%) = (District CI / Sum of CI for District Group) x 100 Performance Share ($M) = Performance Share x Group Performance Allocation The following example illustrates the computations for District A, a Very Good district. Composite Index (CI) = 0.80 + 1.4 = 2.20 District A Performance Share = (2.20 / 4.20) x 100 = 52% District A Funding = 52% x $0.50 = $0.26M Table 9. Composite Index and Performance Share Performance Share Proportionate Share
Scale-Based Score
Composite Index ©
%
$M
District A
0.8
1.4
2.20
52%
$0.26
District E
0.2
1.8
2.00
48%
$0.24
Classification Very Good
Total
4.20
$0.50
Good District B
0.19
1.7
1.89
34%
$0.19
District C
0.74
1.3
2.04
37%
$0.20
District D
0.07
1.5
1.57
29%
$0.16
Total
5.50
$0.55
For this illustrative example, the total allocation (sum of base and performance allocation) received by each district during Year 2 is shown in Table 10.
12
Table 10. Year 2 Allocation per District
District A B C D E Total
Base Allocation, 85% Health Social Facilities Deprivation $ 0.67 $ 0.54 $ 0.40 $ 0.27 $ 0.27 $ 0.63 $ 0.20 $ 0.46 $ 0.54 $ 0.18 $ 2.08 $ 2.08
Rural Population $ 0.58 $ 0.35 $ 0.23 $ 0.17 $ 0.46 $ 1.79
13
Total
Performance Allocation, 15%
$ 1.79 $ 1.02 $ 1.13 $ 0.83 $ 1.18 $ 5.95
$ 0.26 $ 0.19 $ 0.20 $ 0.16 $ 0.24 $ 1.05
Total Allocation per District $ 2.05 $ 1.21 $ 1.33 $ 0.99 $ 1.42 $ 7.00
Discussion The Punjab Model can be studied by countries seeking to align Performance−Based Financing (PBF) with Performance−Based Planning (PBP) at sub-national levels, in the context of using district health system strengthening as a means for moving towards equitable UHC. The benefits of the model arise from the distribution of financial resources through both Base and Performance Allocations. The Base Allocation, as determined by the needs of a district, provides an opportunity to address the district health system’s distinctive infrastructure, operational and epidemiological issues. The Performance Allocation also takes into account each district’s unique circumstances, rather than setting a uniform performance target. Using this approach, a district is rewarded for improved performance, not simply on progress towards a fixed target. The model acknowledges that each district is unique and progress toward national targets may occur at different rates and with varying degrees of difficulty. To obtain district level support and cooperation, adoption of the Punjab Model calls for a transparent process and broad-based stakeholder engagement as illustrated in Figure 3. Such an approach is highly participatory, allowing health managers and public sector administrators from all districts to be involved from the outset. The core group agrees on the performance indicators, financial allocation models, and implementation strategies. Figure 3. Advocacy and Stakeholder Engagement Develop Policy Advocacy • Engage individuals from Finance, Planning and Health (or other relevant sectors) at national and sub-national levels. Form Technical Core Group • Include national technical and financing experts • Adapt the model to country's health system, infrastructure and district-specific needs Organize Advocacy Workshops • Include district teams and those implementing at sub-national levels • Build district capacity and ensure buy-in on performance linked resource allocation Define Performance Indicators • Establish priority areas with district teams • Identify baseline data, source and means of verification
Several measures are recommended to ensure sustainability following implementation: • Government leadership and ownership at all levels • Positive competitive environment that emphasizes rewarding progress over punishing poor performance • Supportive supervision • Timely technical support • Third party evaluation to assess appropriateness of model for local context The focus on local context offers greater flexibility than typical PBF approaches, allowing for implementation in districts with varying priorities and needs. The model facilitates the
14
development of appropriate and feasible indicators to track needs, equity, and performance. In Punjab, indicators are derived from existing Health Management Information Systems (HMIS), making data collection feasible and sufficiently accurate to trigger funding decisions. The HMIS data is regularly validated through the Lot Quality Assurance Sampling (LQAS) technique. An annual household survey (third party validation) is also conducted and used to validate the data. Adoption of the Punjab Model can foster a broader policy perspective, shifting focus away from individual healthcare provider performance to improvements in district health system performance. Furthermore, while the model did not originally intend to promote competition among districts, it developed organically as districts were eager to be classified as a “Very Good” district, even though this designation did not guarantee an increase in a district’s performance share. In the case of districts that fully reached their targets, achieving high ratings became a non-monetary incentive to sustain high coverage levels. The model also aims to institutionalize incentives to increase coverage and inclusion of the unreached. With implementation of the Punjab Model, lower performing districts are unable to obtain funding by way of Performance Allocations, while those able to improve coverage (or sustain already high coverage) receive additional funding. In Punjab, this approach effectively reduced the gap between resource needs and district funding, while improving outcomes by ensuring all districts receive a minimum Base Allocation. By linking PBF to PBP at the district level, the model ensures that financing reflects measurable improvements in the level and equity of health outcomes. Additionally, it is both practical and strategic to build on existing planning and financing systems to avoid parallel processes that will likely lack sustainability (Hotchkiss, Diana et al. 2012). A sustainable post-2015 development agenda requires that health system strengthening be reoriented toward rewarding districts for health system performance that delivers UHC with equity. Future iterations of the Punjab Model can include development indicators to monitor multi-sectoral progress in other social welfare areas. Ensuring sub-national planning and financing align and focus on equity can potentially serve as a catalyst for extending services to the most marginalized and vulnerable populations.
15
References 1. Bossert, T., M. B. Chitah, et al. (2003). "Decentralization in Zambia: resource allocation and district performance." Health Policy and Planning 18(4): 357. 2. Bucagu, M., J. M. Kagubare, et al. (2012). "Impact of health systems strengthening on coverage of maternal health services in Rwanda, 2000–2010: a systematic review." Reproductive Health Matters 20(39): 50-61. 3. Collins, C. D., M. Omar, et al. (2002). "Decentralization, health care and policy process in the Punjab, Pakistan in the 1990s." The International Journal of Health Planning and Management 17(2): 123-146. 4. Dynes, M., S. T. Buffington, et al. (2013). "Strengthening maternal and newborn health in rural Ethiopia: Early results from frontline health worker community maternal and newborn health training." Midwifery 29(3): 251-9. 5. Ghaffar, A., B. M. Kazi, et al. (2000). "Health care systems in transition III. Pakistan, Part I. An overview of the health care system in Pakistan." J Public Health Med 22(1): 38-42. 6. Government of Pakistan (2010). 18th Constitutional Amandment and National Health Programmes, Governement of Pakistan with support from Technical Resource Facility. 7. Government of Pakistan, MoH. (2005). MDGs: Status, Challenge and Future Direction. Lahore. 8. Government of Punjab (2007). District-based Multiple Indicators Cluster Survey (MICS), 20072008. Lahore. 9. Government of Punjab, H. D., Project Management Unit - Punjab Health Sector Reforms Programme. (2009a). "Punjab Health Sector Reforms Programme (PHSRP), 2009-2010." Retrieved October 10, 2011, 2011, from http://phsrp.punjab.gov.pk/phsrp.asp. 10. Government of Punjab, H. D., Project Management Unit - Punjab Health Sector Reforms Programme. (2009b). "Punjab Millenium Development Goals Programme (PMDGP), 20092010." Retrieved October 10, 2011, 2011, from http://www.phsrp.punjab.gov.pk/pmdgp.asp. 11. Gyorkos, T. W., S. A. Joseph, et al. (2009). "Progress towards the Millennium Development Goals in a community of extreme poverty: local vs. national disparities in Peru." Tropical Medicine & International Health 14(6): 645-652. 12. Hailu, D. and R. Tsukada (2011). Achieving the Millennium Development Goals: A Measure of Progress (Working Paper Number 78). 13. Hogan, M. C., K. J. Foreman, et al. (2010). "Maternal mortality for 181 countries, 1980−2008: a systematic analysis of progress towards Millennium Development Goal 5." The Lancet 375(9726): 1609-1623. 14. Hotchkiss, D. R., M. L. Diana, et al. (2012). How Can Routine Health Information Systems Improve Health Systems Functioning in Low- and Middle-Income Countries? Assessing the Evidence Base. Health Information Technology in the International Context (Advances in Health Care Management). N. Menachemi and S. Singh, Emerald Group Publishing Limited. 12: 25-58. 15. Javed, T. A. and S. Amin (2007). "Health sector reforms programme in Punjab: a primary healthcare initiative." Clin Med 7(1): 19-22. 16. Kasi, A. (2001). National Health Policy 2001. The Way Forward. Ministry of Health, Islamabad. Accessed 9/16/13. http://www.unfpa.org/sowmy/resources/docs/library/R205_MOHPakistan_2001_NHP.pdf.
16
17. Kim, D., C. F. Baum, et al. (2011). "The contextual effects of social capital on health: A crossnational instrumental variable analysis." Social Science and Medicine 73(12): 1689-1697. 18. Liljestrand, J. and M. R. Sambath (2012). "Socio-economic improvements and health system strengthening of maternity care are contributing to maternal mortality reduction in Cambodia." Reproductive Health Matters 20(39): 62-72. 19. Lozano, R., H. Wang, et al. (2011). "Progress towards Millennium Development Goals 4 and 5 on maternal and child mortality: an updated systematic analysis." The Lancet 378(9797): 1139-1165. 20. Mahmood, A. (2009a). Costing of MNCH-related minimum service delivery standards, Asian Development Bank. 21. Mahmood, A. (2009b). Public Expenditure Review Health Sector - Punjab, Asian Development Bank. 22. Marmot, M., S. Friel, et al. (2008). "Closing the gap in a generation: health equity through action on the social determinants of health." The Lancet 372(9650): 1661-1669. 23. Meessen, B., A. Soucat, et al. (2011). "Performance-based financing: just a donor fad or a catalyst towards comprehensive health-care reform?" Bulletin of the World Health Organization 89(2): 153-156. 24. Rusa, L. (2009). "Performance-based financing for better quality of services in Rwandan health centres: 3-year experience."Tropical medicine and international health 14(7): 830-837. 25. Soeters, R., C. Habineza, et al. (2006). "Performance-based financing and changing the district health system: experience from Rwanda." Bulletin of the World Health Organization 84(11): 884-889. 26. Soeters, R., P. B. Peerenboom, et al. (2011). "Performance-Based Financing Experiment Improved Health Care In The Democratic Republic Of Congo." Health Affairs 30(8): 1518-1527. 27. Sohani, S., N. Rafique, et al. (2009). Punjab Health Sector Analysis: Achieving Millennium Development Goals (Draft), Asian Development Bank. 28. Travis, P., S. Bennett, et al. (2004). "Overcoming health-systems constraints to achieve the Millennium Development Goals." The Lancet 364(9437): 900-906. 29. UNICEF (2010a). Narrowing the Gaps to Meet the Goals. 30. UNICEF (2010b). Progress for Children: Achieving the MDGs with Equity. 31. Verguet, S., W. Jassat, et al. (2012). "Measles control in Sub-Saharan Africa: South Africa as a case study." Vaccine 30(9): 1594-1600. 32. Walley, J., J. E. Lawn, et al. (2008)"Primary health care: making Alma-Ata a reality." The Lancet 372(9642): 1001-1007. 33. WHO (2007). Everybody's Business: Strengthening Health Systems to Improve Health Outcomes: A Framework for Action. Accessed 9/16/13. http://www.who.int/healthsystems/strategy/everybodys_business.pdf.
17
