Maternal Mortality Survey 2001
Gambia, The, 2001
Identification
GMB_2001_MMS_v01_M
Maternal Mortality Survey 2001
| Name | Country code |
|---|---|
| Gambia, The | GMB |
Other Household Health Survey [hh/hea]
This is one time survey conducted to measure the maternal mortality ratio in 2001.
The Government of The Gambia has always been committed to the “Health for All” year 2000 and beyond Alma Ata Declaration (1978) as well as other conventions such as the 1987 Global Conference on Safe Motherhood, the 1990 Convention on the Rights of the Child and the 1994 ICPD-Cairo Plan of Action, amongst others. A unique recommendation from all these conventions was the reduction of maternal mortality by half by the year 2000 and the provision of a comprehensive reproductive health programme using the life cycle approach from birth to death.
The 1987 conference on safe motherhood brought about increased awareness in the health sector on the issue of maternal mortality following which the “sisterhood” method of estimating levels of maternal mortality was first tested in The Gambia in 1987. This field test was done by the MRC field station located in one of the rural divisions of The Gambia and a total of 90 maternal deaths were identified. The lifetime risk of maternal death was estimated to be higher than one woman in twenty (Greenwood et al.). Subsequently, this revelation by the MRC study sparked a new impetus into the “silent epidemic” of maternal mortality following which the Department of State for Health through its MCH/FP programme commissioned a national survey in 1990. The results, which were quite startling, revealed a maternal mortality level of 1,050 per 100,000 live births nationally. There were variations between urban (600 per 100,000) and rural communities with trained birth attendants (894 per 100,000), and communities without trained birth attendants (1,600 per 100,000).
Recent isolated studies on maternal mortality have suggested a general decline in those areas. However, in the absence of a viable vital registration system in The Gambia, there has been a felt need to conduct another national survey, since the 1990 survey. Furthermore, the Department of State for Health’s proposed shift from MCH/FP service provision into a broad-focussed reproductive health programme also requires the availability of current baseline information and the identification of relevant process indicators, all of which justify the need to establish current levels of mortality and use of contraceptives.
It should be noted that current national policies and programmes continue to refer to data obtained from the 1990 maternal mortality study, the 1990 Gambia contraceptive prevalence and fertility determinants survey as well as the 1993 population and housing census as baseline benchmarks both for programme intervention and implementation. This long period to some extent renders the data quite obsolete and unsuitable for many national and development purposes. A simple compromise has been that of making comprehensive demographic, health and socio-economic projections. However, one important limitation of statistical projections is the period between the time the base data were collected and the time span of the projections. The probable margin of error in making projections with reference periods of eight or more years ago could be so large to warrant the acceptance of such projections within any reasonable statistical intervals.
Since there has been no comprehensive national survey on maternal, infant and child mortality during the past 10 years, and given that it would take a number of years before the final analyses of data obtained from the forthcoming census, it was found prudent to carry out a comprehensive study that would collect information on key reproductive health indicators. Furthermore, the complexity involved in studying maternal mortality compounded by its rarity of occurrence in the general population has necessitated conduction of a specialised study. Such a study would be useful in filling in the data deficiencies and providing baseline data for programme intervention and evaluation, especially in an era of a general shift of emphasis of population programmes from vertical family planning activities in favour of a more generally accepted concept - reproductive health.
Objectives of the survey:
a) To establish current levels of maternal, peri-natal, neonatal and infant mortality rates.
b) To establish the current levels of contraceptive prevalence rates and barriers to use.
c) To elicit how the situation has improved or otherwise during the last ten years.
d) Make practical recommendations to Department of State for Health for subsequent and long-term actions required.
Sample survey data [ssd]
- Households
- Women and men (both in child-bearing age)
Scope
The 2001 Maternal Mortality Survey covered the following topics:
HOUSEHOLDS
- Household Demographics
WOMEN
- Respondent's Background
- Reproduction
- Contraception
- Marriage
- Fertility Preferences
- Maternal Mortality
Coverage
National.
The survey covered women age 15 to 49 years old and men age 18 years and over.
Producers and sponsors
| Name |
|---|
| Department of State for Health |
| Name | Affiliation | Role |
|---|---|---|
| United Nations Children's Fund | UN | Technical assistance |
| World Health Organisation | UN | Technical assistance |
| United Nations Population Fund | UN | Technical assistance |
| Gambia Bureau of Statistics | Technical assistance |
| Name | Role |
|---|---|
| World Health Organisation | Financial assistance |
Sampling
A multi-stage stratified cluster sampling procedure was used for this study. The country is divided into 41 Districts and each of these districts was identified as a stratum. Stratification by districts increases the efficiency of the sample given the homogeneity of the districts. The sample size for the study was 4000 households and was based on the level of maternal mortality which was estimated at 1050 per 100,000 at the time of the study. According to WHO/UNICEF, 1997 publication on Sisterhood Method in Estimating Maternal Mortality, 4,000 households or less would be adequate for study of maternal mortality if the level of maternal mortality is at least 500 per 100,000.
Based on the Rule of Thumb, a 15 per cent sample of EAs (240) was selected for this study, which is also more than adequate for the study of other variables like contraceptive prevalence, infant mortality, fertility and its determinants. The selection of population elements were done at two stages; a representative sample of 240 Enumeration Areas (EAs) were randomly selected and allocated based on the Probability Proportional to the Size (PPS) of the district using random numbers. The EA is a cluster of settlements with an estimated population of 500 peoples.
A total of 4,000 households were then allocated to the districts with probability proportional to the size of each district. For the 240 selected EAs, a specified number of households were randomly selected for interview using a systematic sampling procedure. A complete listing of selected household members was done and all eligible male and female respondents were interviewed.
There were no discrepancies between the sample units obtained and the iniitial planned samples.
All respondents with missing age were excluded from the model. There were about 15 percent of the responses with missing information on the deaths. Imputations were made to establish whether or not they qualified to be classified as maternal deaths. For instance, those missing sex of the sibling but had correctly answered maternal death-related questions, the sex was taken to be female and therefore included in the maternal mortality model. Responses with no information on the type of maternal death, but had indicated the death as having been as a result of complications of pregnancy or child birth, were imputed to be pregnancy-related. On the other hand, responses on symptoms before death were used to impute the type of death in case it was missing.
No weights were applied.
Survey instrument
The survey tools included a compound and household schedules, female and male questionnaires. The compound and household schedules were used to collect information on local government area, health division and household number, together with residence, sex, age, education and eligibility status of the household members. The female and male questionnaires were administered to women aged 15-49 years and men aged 18 years and above respectively. The survey instruments were similar to the core modules of the Demographic and Health Survey questionnaires (Macro International), with adaptation to suit The Gambian needs. In addition a review of medical records in the three main hospitals in The Gambia (Royal Victoria Hospital, Farafenni Hospital and Bansang Hospital) was carried out in November 2001 to undertake first-hand assessment of the maternal mortality situation at the major referral facilities.
The Survey team with support and guidance of the Technical Team prepared the survey instruments by adapting the Demographic and Health survey modules. The main instruments for this study are:
- Male questionnaire which was used to obtain information from males 18 years and above;
- Female questionnaire, which obtained information from females, 15-49 years ;
- Household questionnaire contains information on Local Government Area (LGA), Districts and Household numbers.
For each person listed on the household questionnaire, relationship to head of household, age, and sex are recorded.
The female questionnaire contains the following key information:
- Respondent's background
- Reproduction
- Contraception
- Marriage
- Fertility preferences
- Maternal mortality
The male questionnaire on the other hand, contained the following information:
- Respondent's background
- Contraception
- Marriage
- Maternal mortality
Data collection
| Start | End |
|---|---|
| 2001-09-28 | 2001-11-30 |
| Name |
|---|
| Department of State for Health |
| Gambia Bureau of Statistics (GBOS) |
The fieldwork lasted for 45 days. The total number of households covered is 3989. The males and females interviewed in these households were 5038 and 5786 respectively.
The data collection exercise was conducted in teams of 5 and a supervisor headed each team. Three technicians worked with these teams to provide technical advice during fieldwork.
Data appraisal
All respondents with missing age were excluded from the model. There were about 15 percent of the responses with missing information on the deaths.
Data Access
| Is signing of a confidentiality declaration required? | Confidentiality declaration text |
|---|---|
| yes | The confidential policy of the Health department must be abided. Signing of a confidentiality declaration is needed to access the data. |
The Data and metadata is provided to you for exclusive use. The data and/or metadata may not be transferred to any other user without prior written authorization from DOSH.
Data may be accessed by academic users for non-commercial purposes.
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
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 interpretations or inferences based upon such uses.
Metadata production
DDI_GMB_2001_MMS_v02_M
| Name | Role |
|---|---|
| World Bank, Development Economics Data Group | Generation of DDI documentation |
2007-06-20
Metadata version
Version 1.1 (June 2011) Adopted from "mmrstudy" Nesstar file that was done by Musa MM Sowe.