UNIVERSAL IMMUNIZATION THROUGH IMPROVING FAMILY HEALTH SERVICES
(UI-FHS)
Immunization Costing Exercise
BASELINE REPORT
(Final Draft)
February 2014
ADDIS ABABA, ETHIOPIA
Acronyms and Abbreviations
cMYPCVD / comprehensive Multi Year Plan
Center for Vaccine Development at University of Maryland, USA
EC
EFY / Ethiopian (Julian) Calendar
Ethiopian Fiscal Year
EHNRI / Ethiopian Health and Nutrition Research Institute
EPI / Expanded Program on Immunization
FIC
FMOH
GAVI
GC
GOE / Fully Immunized Child
Federal Ministry of Health
GAVI Alliance
Gregorian Calendar
Government of Ethiopia
HC
HP
JSI / Health Center
Health Post
JSI Research & Training Institute
MS
SNNPR
SPSS / Microsoft
Southern Nations Nationalities and Peoples Region
Statistical Package for Social Sciences
UI-FHS / Universal Immunization through improving Family Health Services project (implemented by JSI)
UNICEF / United Nations International Children's Fund
WHO / World Health Organization
WoHO / Woreda Health Office
Table of Contents
Acronyms and Abbreviations
Summary
1.Introduction
1.1.Background
1.2.Objectives
2.Methodology
2.1.General Approach
2.2.Data Collection Techniques
2.3.Data Quality and Availability
2.4.Geographic Scope and Sampling
2.4.1 Geographic scope
2.4.2 Sampling technique
2.5.Data Management and Analysis
2.5.1 Data management
2.5.2 Methods of analysis
3.Results
3.1.Description of the Analysis Sample
3.2.Summary of Total Immunization Costs by Woreda and Health Facility
3.3.Disaggregated Analysis by Childhood and Tetanus Toxoid Vaccination Costs
3.4.Unit Costs per Fully Immunized Child
3.5.Unit Cost of Tetanus Toxoid (TT) Vaccination for Women
4.Comparative Analysis
4.1.Comparison of Unit Costs between the three Woredas
4.2. Comparison of Costs by Vaccine Antigens
4.3.Comparison of Vaccine and Non-vaccine Costs
4.4.Comparison of Woreda Health Budget with Immunization Expenditures
4.5.Unit Costs per Coverage: Comparison between Administrative and Survey Data
5.Discussion
References
Annexes
Summary
Immunization is one of the most cost-effective and affordable ways to control infectious diseases and should be a financial priority for national governments and donor agencies. Ethiopia launched its Expanded Program on Immunization (EPI) in the mid 1970s, and has a national EPI policy that recommends health workers immunize eligible children according to internationally recommended vaccine schedules[1]. Funding for the Ethiopian EPI comes from the national government as well as partners. As of 2003 Ethiopian Calendar (2010/11GC), the immunization coverage with fully vaccinated has reached to 74.5% (based on MOH administrative data)), with funding from the government as well as international and local partners.
The Universal Immunization through Improving Family Health Services project (UI-FHS)is a 3.5 year learning initiative funded by the Bill & Melinda Gates Foundation and implemented by JSI Research and Training Institute, Inc. The goal of UI-FHS in Ethiopia is to develop evidence in three learning woredas (districts)–Arbegona in SNNPR, Assaieta in Afarregion and Hintalo Wajerate in Tigray region, on how immunization activities can be integrated with family health approaches in a cost-effective and sustainable manner.
This is, therefore, a baseline report for the survey which aimedat estimating costs at various points in time (retrospectively) so as to create a panel data set and to evaluate changes in total and unit costs, and cost structure as interventions are implemented. There will also be mid-term and final reports that will follow data collections after interventions are implemented. The mid and final reports will also present estimates for the costs of scaling up the most successful intervention packages to other woredas and the incremental cost of adding a new vaccine to the existing schedule.
This baseline report presents the descriptive and comparative cost estimates on the total and unit costs of interventions in immunization at the levels of health centers (HCs) and health posts (HPs) within the three study woredas. The major findings of the report include:
-The weighted total annual immunization costs were estimated to be USD 48,782.86 for Arbegona (population =161,321), USD 24,126.59 for Assaieta (population = 54,749) and USD 96,231.19 for Hintalo Wajerate (population = 176,527)woredas.
-As costs of vaccines are the major items within the cost estimates, accounting for 65% of the total costs in Arbegona, 51% in Assaieta and 84% in Hintalo Wajerate), the total costs excluding vaccines were: USD 17,086.13 in Arbegona, USD 11,756.71 in Assaieta, and USD 15,285.68 Hintalo Wajerate (these reflect costs that woredas would assume in their budget, as cost of vaccines is paid from national budget).
-Second to vaccine costs, expenditures on personnel take the higher share of costs in all the woredas (11.14% in Arbegona, 14.29%% in Assaieta, and 6% in Hintalo Wajerate).
-The weighted total annual child immunization cost is USD 45,852.94 for Arbegona, USD 21,852.67for Assaieta and USD 92,786.35for Hintalo Wajerate woredas. The corresponding figures for the cost of non-vaccine items are: USD 14,552.97 for Arbegona, USD 9,818.71 for Assaieta, and USD 13,084.81 for Hintalo Wajerate.
-The numbers of fully immunized children (FIC)obtained from MOH administrative records were: 2,702 for measles and 3095 for penta3 in Arbegona, 1,041 for measles and 1,323 for Penta3 in Assaieta, and 6,894 for measles and 3,621 for Penta3 in Hintalo Wajerate.
-The weighted total unit costs per fully immunized child (up to measles) are estimated to be USD 16.97, USD 20.99and USD 15.38 respectively in Arbegona, Assaieta and Hintalo Wajerate woredas. The respective costs for non-vaccine items are: USD 5.39 for Arbegona, USD 9.43 for Assaieta and USD 3.82 for Hintalo Wajerate. As these figures show, comparisonof the total unit costs of non-vaccine items per fully immunized child (up to measles) between the three woredas shows that costs are highest in Assaieta, next higher in in Arbegona and lowest in Hnitalo Wajerate.
-The weighted total annual tetanus toxoid (TT) immunization cost for all women who took the two doses of the vaccine (until TT2) are: USD 2,929.9247 for Arbegona, USD 2,273.92 for Assaieta, and $USD 3,444.84for Hintalo Wajerate, with unit cost per woman being USD 0.96, USD 0.88, and USD 0.36 respectively for the three woredas. The corresponding weighted total costs for non-vaccine items are: USD 2,533.16 (with unit cost of USD 0.83) for Arbegona, USD 1,938 (with unit cost of USD 0.75 for Assaieta), and USD 2,200.87 (with unit cost of USD 0.23) for Hintalo Wajerate.
-Comparison of FIC (up to measles) figures between those obtained from administrative sources and those obtained from coverage surveys show higher number of FIC in Arbegona (3,451 versus 2,702) and Assaieta (1,090 versus 1,041), while the coverage figures in Hintalo Wajerate (4,924) are lower than those obtained from administrative figures (6,894).
According to the estimates of this baseline report, woredas with relatively higher population and health care budget seem to have a relatively lower unit cost per fully immunized child because of economy of scale. In addition to the smaller size of the population, the hard working condition and the relatively higher investment in extra personnel seem to have contributed to the relatively higher unit cost observed in Assaieta woreda.
1.Introduction
1.1.Background
This is a baseline report of the immunization costing exercise conducted in three woredasin Ethiopia(Arbegona in SNNPR, Assaieta in Afar region, and Hintalo Wajerate in Tigray region). The three woredasare project sites for the Universal Immunization through improving Family Health Services Project (UI-FHS) implemented by JSI Research & Training Institute(JSI).
Immunization is one of the most effective and affordable ways to control infectious diseases, and should be a financial priority for national governments and donor agencies. If the financial commitment to immunization services is to be reduced, the subsequent increase in morbidity and mortality would result in serious social and economic consequences. Therefore, immunization clearly requires a sustained commitment and a successful immunization program should reach every eligible woman and child.
Ethiopia developed a health policy in 1993 and the Expanded Program on Immunization (EPI) was launched in 1980 with an immunization policy updated in 2007.[2][3] The national EPI policy recommends that health workers should use every opportunity to immunize eligible women and children according to the recommended schedule. The program strategies of EPI are directed at: increasing immunization coverage, reducing missed opportunities and defaulters, increasing the quality of immunization services, improving public awareness and community participation to sustain high coverage and disease control/eradication. Funding for the program comes from government as well as partners. Salary for personnel and running costs of health facilities is budgeted by the Government of Ethiopia (GOE). Costs of vaccines, cold chain equipment, transport, social mobilization and other operational costs are covered by the GAVI Alliance (GAVI), United Nations Children’s Fund (UNICEF), World Health Organization (WHO) and other development partners; GOE also covers some proportions of the costs of traditional vaccines. As of 2010 GC, the immunization coverage with penta 3, measles and fully vaccinated has reached to 84.7%, 81.5% and 74.5% respectively.[4]
As one of the development partners supporting the EPI, JSI, has recently launched the Universal Immunization through improving Family Health Services (UI-FHS) project through an award of a 3.5 year learning grant funded by the Bill & Melinda Gates Foundation. The project works in close collaboration with the Federal Ministry of Health (FMOH), the Ethiopian Health and Nutrition Research Institute (EHNRI), and the Center for Vaccine Development (CVD) at the University of Maryland (USA) as implementing partners. The goal of UI-FHS is to develop evidence for informing the decisions of the FMOH on whether and how to pursue nationwide universal child immunization in the country, integrated with family health approaches, and what it will take to do so effectively, affordably and sustainably[5]. One specific activity of the UI-FHS project is to develop and disseminate a costed scale out plan for sustainable and universal child immunization based on evidence from three learning woredas.
1.2.Objectives
The objective of the study was to estimate costs at various points in time (retrospectively) so as to create a panel data set and to evaluate changes in total and unit costs, and cost structure as interventions are implemented in three learning woredas representing three different regions and health system strength contexts. Findings will be used to determine immunization related costs that could be borne by woreda health budgets in strengthening routine immunization systems across Ethiopia.
2.Methodology
2.1.General Approach
In this costing exercise the focus is on evaluating the economic (rather than financial) costs incurred by the health sector (the government) in delivering immunization services. Economic costs differ from financial costs in that they are based on the opportunity costs of the resources allocated in the provision of services rather than only considering the actual expenditures made.[6] Furthermore, the costing exercise is limited to costs incurred by the health system in providing immunization services and is not involved in estimating patient (population) side costs.
In terms of the description of the interventions/intervention packages that will be implemented, the current costing exercise is aimed at estimating the total and unit costs of the interventions and later, to estimate the cost of scaling up the most successful intervention packages. To do this, first estimates are made for the costs of all the resources that are being employed in running a vaccination program, including basic infrastructure in this baseline report. Then, in the mid and final reports, estimation will be made on the costs of scaling up the most successful intervention packages to the already existing program[7].
A number of frameworks and guidelines[8][9][10][11][12] that have been internationally used in estimating the costs of immunization are used as appropriate for guiding the current exercise in calculating the cost per fully vaccinated child as well as the cost of a single vaccination.
2.2.Data Collection Techniques
The following data collection and extraction techniques are used for determining the unit costs of conducting routine immunization and to determine approximate costs to implement a model in the contexts of the three select woredas (one pastoralist and two agrarian) and what factors drive variations in cost:
-Review of the documents using checklist: Relevant project documents such as financial documents, work plans and reports were reviewed and compiled to identify and extract all expenditures and cost related data.
-Sample Survey:Facility surveys were conducted in sample health institutions in the three study woredas to identify: a list of interventions and inputs for those interventions in a facility, the unit replacement price of inputs; and the percent use of resources for immunization services to apportion costs of inputs.
-In-depth Interview: In-depth interviews were made with key persons (program officer/managers) in all three Woreda Health Offices (WoHOs), health facilities, donor agencies involved in the interventions as well as with officials at the Maternal and Child Health Directorate Child Health Program within the Federal Ministry of Health.
The tools for data collection were piloted in Adama (agrarian) and Fentale (pastoralist) woredas—different from the actual study woredas. Two WoHOs, two HCs and two HPs were included in the pre-testing. The pre-testing helped to review the applicability of the tools in the Ethiopian setting in order to capture any changes/developments and peculiarities within the health structure that may not be incorporated in the generic version of the tools. After the pre-test, the necessary amendments were made on the tools to make them ready for the actual data collection.
2.3.Data Quality and Availability
Data for this costing exercise are collected through review of administrative records and through interviewing health sector officials and health workers at their places of operation. Observations were also made of some of the items included in the cost estimates as well as some of the activities. However, there are limitations to the completeness of records on the items required for costing as well as to the accuracy and consistency of responses from the interviewed health workers. This has sometimes necessitated using imputed data for some of the items where information is missing.The method of imputation used was the mean imputation method where by missing values are replaced with the mean of the observed values (within a given woreda) for that variable. The variables for which such imputations were used include: wastages for BCG, PCV and measles (5-10% of observations); costs of injection materials for BCG, penta, measles, PCV and TT (up to 15%% of observations); costs office equipment and supplies (up to 5% of observations). In addition, for socio-demographic figures such as population growth, literacy and infant mortality rates, values for the respective regions were used.
Therefore, as is the case this type of cost studies in developing countries, these limitations and the overall context must be taken into consideration in the interpretation of the results.
2.4.Geographic Scope and Sampling
2.4.1 Geographic scope
The costing exercise is conducted in Arbegona woreda in SNNPR, Assaieta woreda in Afar region, and Hintalo Wajerate woreda in Tigray region. Data were collected within the three woredas at WoHOs and Primary Health Care Units (PHCUs), including health centers (HCs) and health posts (HPs). Relevant information was also collected from the project office, from other donors, and from zonal, regional and/or national levels as appropriate (for data required and could not be available from other levels).
2.4.2 Sampling technique
The baseline survey was conducted on sampled health centers and health posts from the three study woredas.This survey employed a stratified sampling approach to select health facilities to be included at different levels. The sampling frame used was the list of facilities by type woredasobtained from the three WoHOs. The main objective of the study was to estimate the unit cost of immunization services using either a HC or a HP as costing center; these health units were selected in a purposive stratified manner so as to fit within a referral network set in each woreda(implying a need to first select sample of PHCUs and then select HPs that are within the catchment area of the selected PHCU). As the numbers of HCs within the three woredasare not many (see table 1) for considering standard power calculation procedures and as there were no previous studies of this nature in the study area, taking about forty percent of the available health institutions within the three woredaswould give reasonable estimate of the parameters to be measured within each of the study areas that is also statistically justified for generalize-ability.
Therefore, two PHCUs (out of five) were selected from Arbegona, and three PHCUs (out of the available seven) were selected from Hintalo Wajerate woredas. Regarding HPs, using sampling proportionate to size, five HPs within the catchment areas of each of the selected PHCUs in Arbegona and three PHCUs within the catchment areas of each of the selected PHCUs in Hintalo Wajerate were randomly selected. For Assaieta woreda, since there is only one HC, that HC as well as all seven HPs in the woredawere selected automatically. Based on this procedure, therefore, the sampled health units are as follows: