Global Fund Household Health Coverage Survey 2008
Ethiopia, 2008
Identification
ETH_2008_GFHHCS_v01_M
Global Fund Household Health Coverage Survey 2008
First round study Household Health Coverage Survey 2008
| Name | Country code |
|---|---|
| Ethiopia | ETH |
Other Household Survey [hh/oth]
This is the first Household Health Coverage Survey conducted for the Five-Year Global Fund Evaluation, as part of the District Comprehensive Assessment activity. The evaluation efforts will involve a series of data collection efforts, including surveys, surveillance, research studies, service and administrative records. The survey started in January 2008 and completed in May 2008.
This survey was conducted in the scope of the Global Fund Five-Year Evaluation.
Purpose
(1) To provide population-based data on knowledge of TB and HIV/AIDS; sexual behavior; utilization of services related to HIV testing and maternal and child health; and coverage of IRS and bednets
(2) To provide basic information on health expenditure
Objective
To provide data to evaluate the following areas:
-Resource tracking [Inputs]
-Access/coverage/use of services [Outcomes]
-Public health impact [Impact]
-Health system strengthening [Impact]
Content
Types of indicators:
- household level demographic and socio-economic indicators such as education, wealth assets, residence
- HIV- Knowledge and Behavior, Prevention, Counseling and Testing, PMTCT
- Tuberculosis- Knowledge and Behavior
- Malaria - ITN, IRS, IPTp, Prompt and effective treatment
- Health system effects
- Financing
Sample survey data [ssd]
- Households
- Children under 5 years
- Women
Version
Version 1.1: Edited second version, some data errors have been corrected and revised on 29-05-2008.
2008-05-29
Scope
The survey covers the follwoing topics:
HOUSEHOLD:
- Household listing
- Demographic information
- Household members
- Bed-nets
- Health expenditures
- Deaths in the household
WOMEN:
- Respondent's background
- Birth history
- Antenatal and delivery care
- Immunization
- Diarrhea
- Fever and malaria
- Tuberculosis
- Marriage and sexual activity
- HIV/AIDS
Coverage
Representative at the Woreda cluster level, for selected Regions. Woredas are grouped, by region, to form "district" like groups (each with 5 woredas), which yields a total number of 160 woreda clusters. Each woreda group will have one index cluster with an ART hospital. The selected Woredas are not nationally representative.
The survey covered all de jure household members, all women 15-49 years,
Producers and sponsors
| Name | Affiliation |
|---|---|
| Ethiopian Health and Nutrition Research Institiute (EHNRI) | MoH |
| Name | Role |
|---|---|
| Macro International | Technical assistance |
| Name | Role |
|---|---|
| Global Fund | Funding |
| Name | Role |
|---|---|
| Federal Ministry of Health (FMoH), Ethiopia | Participated as members of the Task Force and Core Group |
| Federal HIV/AIDS Prevention and Control Office (FHAPCO), Ethiopia | Participated as members of the Task Force and Core Group |
| World Health Organization (WHO) | Participated as members of the Task Force and Core Group |
| President’s Emergency Plan for AIDS Relief (PEPFAR) | Participated as members of the Task Force and Core Group |
| Tulane University | Participated as members of the Task Force and Core Group |
| Italian Cooperation/HIVAIDS Donors’ Forum | Participated as members of the Task Force and Core Group |
| Central Statistics Agency (CSA), Ethiopia | Participated as members of the Task Force and Core Group |
Sampling
The following clustering sampling steps are used:
-
Woredas are grouped, by region, to form “district” like groups (each with 5 woredas), which yields a total number of 160 woreda clusters. Each woreda group will have one index cluster with an ART hospital. Four woredas closest to the index woreda are selected to form one woreda cluster together the index woreda.
-
Accordingly 35 woreda clusters are selected across the regions, using PPS procedures. The total number of groups were 35(35/175 = 22%). The samples were distributed to each of the 11 regions based on probability proportional to size. Regions with few facilities have at least 1 hospital, hence that hospital and its health facility network are completely enumerated.
-
For each woreda cluster, 3 Enumeration Areas (EAs) per woreda selected. There was 525 EAs across the country, which is somehow similar to size of the DHS survey. About 17 households are selected randomly per EA for a total of 8750 HHs. The 8750 households were divided across the 35 woreda groups proportionally. Then the total household allocated to each cluster was distributed to the woredas based on proportional to size.
-
Within each woreda, one urban EA was selected with the other two EAs being rural. This was yield roughly 35 urban EAs and 315 rural or a 1:9, which is similar to the overall urban-rural distribution of population.
Level of representation - Woreda level
Strategy for absent respondents - no replacement
Sample frame - Original sample frame of clusters is based on the 2007 census and household listing in selected clusters prior to systematically selecting the households.
From the whole selected Woredas, one Woreda is not covered during the survey due to inaccessability
The total number of households interviewed was 8,325, yielding a household response rate of 99 percent. A total of 8,358 eligible women were identified in these households and interviews were completed for 7,457 women, yielding a response rate of 89.2 percent.
The weights are not available
Survey instrument
Questionnaire modules were based on the standard DHS, including a HH and a woman questionnaire.
- Household modules include: Household listing of members and demographic information, Household members, Bednets, Health expenditures, Deaths in the household,
- Woman modules include: Respondent's background, Birth history, Antenatal and delivery care, Immunization, Diarrhea, Fever and malaria, Tuberculosis, Marriage and sexual activity, HIV/AIDS
Data collection
| Start | End | Cycle |
|---|---|---|
| 2008-01-01 | 2008-03-25 | one round |
| Name | Affiliation |
|---|---|
| Ethiopian Health and Nutrition Research Institiute | MoH |
Interviewing was conducted by teams of interviewers. Each interviewing team comprised of 5 female interviewers and a supervisor.
Field coordinator teams which were headed by experienced researchers selected from EHNRI were organized into 8 groups for coordinating, closely monitoring and regularly visiting field work and data collection activities.
Data processing
Data editing, processing and analysis of the DCA was carried out using Epi-Info version 3.3.2 and CSPro 3.3.
Data appraisal
Frequency checks and data cleaning during data colletion, and final data cleaning was done using Epi-Info
Data Access
| Name | Affiliation | |
|---|---|---|
| Ethiopian Health and Nutrition Research Institiute | MoH | ehnri@ethionet.et |
| Global Fund | wilson.lo@theglobalfund.org |
| Is signing of a confidentiality declaration required? | Confidentiality declaration text |
|---|---|
| yes | No personally identifiable information is available in the microdata sets. See informed consent statement in Qs. |
The Global Fund 5-year impact evaluation data is available to outside researchers as licensed data files. These files can be accessed though the public web site and require user registration. There is a two step procedure required to access the files:
- The researcher should register as a user.
- Once registered the researcher will be required to fill out a detailed license request form that will provide the Global Fund with details on how the data will be used.
Certain conditions may apply depending on the data requested (facility or household). General terms and conditions are provided in the licensing form. Specific terms that apply depend on the survey requested (Facility or Household)
Facility:
The names of the facilities have been suppressed. However, some of the files contain geographic references that could be used to re-identify the facility. If your research project will use the district level data to identify the facility, please be sure to specify the reasons for the re-identification In certain instances, country concurrence may be required. For any information required to validate the name of the facility, please contact the appropriate country authority.
Household:
Household data is recoded and no direct identifiers are available. No special conditions apply for access to the recoded data.
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including acronym and year of implementation)
- the survey reference number
- the source and date of download
Example:
Ethiopian Health and Nutruition Reseach Institute (EHNRI). Ethiopia Global Fund Household Health Coverage Survey 2008. Dataset downloaded from [website/source] on [date].
Disclaimer and copyrights
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretation or inferences based upon such uses.
(c) 2009,Ethiopian Health and Nutrition Research Institiute (EHNRI)
Contacts
| Name | Affiliation | |
|---|---|---|
| Ethiopian Health and Nutrition Research Institute | MoH | ehnri@ethionet.et |
Metadata production
DDI_WB_ETH_2008_GFHHCS_v01_M
| Name | Affiliation | Role |
|---|---|---|
| Ethiopian Health and Nutrition Research Institute | MOH | Carry out the whole survey conduction |
| International Household Survey Network | Technical Assistance |
2008-05-07
Metadata version
Version 1.1: (April 2011) Adopted from "ETH-EHNRI-GF-DFA-2008-v1.0_ddi" DDI which was produced using recoded data.