In addition to the general objectives of evaluating family plan-ning and maternal/child health care program performance, examining the family planning needs of the male and female population, and describing fertility levels, the 1989 Haiti National Contra-ceptive Prevalence Survey had several more specific objectives, as follows:
a. To explain the large discrepancy between the number of contraceptives, particularly condoms, reportedly distributed, and the estimated number used according to past surveys and service statistics. Previous surveys (of females only) had indicated that about 1 percent or 14,000 Haitian couples used condoms as a method of preventing pregnancy. This finding persisted in spite of the seemingly incompatible data showing that as many as 15 million or more condoms have been issued from warehouses in Haiti in some years, which would be sufficient to supply 10 percent or 150,000 couples. The 1989 Haiti National Contraceptive Prevalence Survey sought to resolve this discrepancy by including an independent sample of males and using a newly designed condom module questioning both males and females about condom use, the number of condoms distributed to individuals, the number used, and the number remaining unused. Equally important were questions to ascertain the extent to which Haitians are using condoms outside of marital unions and/or as a means of preventing transmission of HIV and other STD infections rather than pregnancy prevention. Condoms are currently the most effective means of preventing sexual transmission of HIV short of abstinence, but as in most developing countries, the amount of condom use for this purpose in Haiti is unknown.
b. To obtain current data on contraceptive prevalence, method mix, sources of contraception, fertility, and levels of unplanned pregnancy. The 1989 Haiti National Contraceptive Prevalence Survey also ascertained whether there have been any changes in these areas and defined target groups for family planning activities in both the public and private sectors.
c. To investigate barriers to increased family planning use and identify programmatic factors that are important in improving acceptance and continuation of contraception. To achieve this objective, the survey collected information on perceived and actual problems with obtaining supplies, access to services, and the methods themselves.
d. To examine male roles in family planning decision making and male attitudes about family planning. It is often argued that male attitudes constitute a major impediment to both the adoption of family planning methods and to a reduction of family size in much of the developing world. In Haiti, contraceptive use remains low while fertility remains high, and male attitudes are sometimes cited as a reason for this lack of change. The 1989 Haiti National Contraceptive Prevalence Survey, the first to include males, collected data on the male's role in the couple's decision on whether to use a method and what kind of method to use, as well as how male and female attitudes compare regarding desired family size and the use of family planning. These data will show whether program activities might be modified to take male roles and attitudes into account.
e. To examine the proximate determinants of fertility including both modern and traditional contraception, breastfeeding, amenorrhea and patterns of union and cohabitation. Analysis of the 1989 Haiti National Contraceptive Prevalence Survey should cast further light on the difference between the 1983 estimates of total fertility of 5.5 births per women and the 1987 survey which found total fertility to be substantially higher at 6.4 births per woman.
f. To examine certain sexuality issues, particularly as these issues relate to family planning, condom use, and HIV transmis- sion. The 1989 Haiti National Contraceptive Prevalence Survey included questions on such topics as coital frequency and numbers of sexual partners for currently sexually active persons and, for 15-24 year-olds, a module on the age at which sexual activity began and early use of contraception.
Kind of Data
Sample survey data [ssd]
The 1989 Haiti National Contraceptive Prevalence Survey was a nationwide population-based household survey. The population to be surveyed was divided into 3 domains: Metropolitan Port-au-Prince, other urban areas and rural areas.
Producers and sponsors
Authoring entity/Primary investigators
Child Health Institute (CHI)
Division of Reproductive Health, Centers for Disease Control
International Planned Parenthood Federation/Haiti
United States Agency for International Development
The 1989 Haiti National Contraceptive Prevalence Survey was a nationwide population-based household survey. The population to be surveyed was divided into 3 domains: Metropolitan Port-au-Prince, other urban areas and rural areas. Urban areas contain only about one-fourth of Haiti's population. Port-au-Prince and other urban areas were over-sampled so that they include approximately one-half of the sample households so estimates in urban areas would have greater precision. Rural areas were correspondingly undersampled. Thus, all total estimates for Haiti require weights to reflect the true population distribution in each domain (stratum) while unweighted numbers of cases are shown in tables.
The sampling strategy consisted of a two-stage cluster design to select respondents for the survey. The first stage consisted of the selection of independent samples of census enumeration districts (SDEs in French) within each domain. These SDEs were sub-sampled from SDEs selected for a larger 1987 survey of household expenditures (selection of SDEs for that survey was done with probability proportional to size). There were 28 SDEs selected in Port-au-Prince, 22 in other urban areas and 44 in rural areas--a total of 94 SDEs.
The second stage consisted of the random selection of households in each cluster: 46 households in clusters in Port-au-Prince and other urban areas and 54 households in clusters in rural areas. Half the households in each cluster were designated as "male" households and half as "female" households. Male interviewers interviewed all males between the ages of 15 and 59 who resided within selected male households, while female interviewers interviewed all females between the ages of 15 and 49 years in female households.
A total of 4,650 households were included in the sample. (One SDE on the island of Gonave was not visited because of inaccessibility.) It was estimated that this sample size would yield about 4,000 completed individual interviews--1,800 male and 2,200 females. The sample size was based on the minimum number of interviews needed per stratum to obtain adequately precise estimates for most of the survey topics, based on census estimates of potential respondents per household and projected response rates. Complete interviews were conducted with 1,842 males and 1,996 females.
In the 2,324 "male" households, interviews were completed in 91.1 percent of the selected dwellings. Interviewers identified males eligible respondents in 76.6 percent of the selected dwellings, while 14.5 percent of the selected dwellings did not contain eligible males. In occupied "male" households where household interviews were not obtained, the reasons were because selected individuals could not be located in 3.4 percent of selected dwellings and because of refusals in 1.6 percent of the selected dwellings. Of the rest, 3.1 percent were found to be unoccupied and 0.8 percent were classified as "other" (demolished, no longer existing, etc.).
In the 2,326 "female" households, the results were almost identical. Female household interviews were completed in 90.8 percent of the sample dwellings. As a result, interviewers found females eligible to be respondents in 78.1 percent of the selected dwellings and no one eligible in 12.7 percent of the selected dwellings. Among occupied "female" households where household interviews were not conducted obtained, the reasons were that residents could not be found in 3.3 percent of selected dwellings and because of refusals in 1.3 percent. Of the rest, 3.7 percent were found to be unoccupied and 0.9 percent were classified as "other" (demolished, no longer existing, etc.).
The completed household questionnaires identified 2,287 men as eligible for an interview. Of these, interviews were completed with 1,842, or 81 percent. The major reason that individual interviews were not completed was the absence of the respondent despite several visits to the households (17.5 percent). Only 0.7 percent refused to be interviewed.
The completed household questionnaires identified 2,301 women as eligible for interview. Of these, completed interviews were obtained for 1,996, or 87 percent. Contact was not made with 11.2 percent and 0.7 percent refused to be interviewed.
Table 2-2 of the Final Report shows the percent distribution of the total survey population by age group compared with those respondents for whom interviews were completed. The column to the left for both males and females shows the percent distribution of all household residents, not just respondents, based on listings of residents from the household questionnaire. These data were gathered on the separate household questionnaire which listed all household residents. The data in the right hand column for both males and females show the proportion each age group represents in the sampled population, and in general, decreases as age increases, reflecting recent high fertility. However, respondents 20-24 years of age of both sexes represent a smaller proportion of the population than expected, reflecting lower response rates for this hard-to-reach mobile population. Except for this one age group, completed interviews follow the same pattern seen for all household residents.
Dates of Data Collection (YYYY/MM/DD)
Mode of data collection
Type of Research Instrument
The survey instrument consisted of two parts--a short household questionnaire and a much longer respondent questionnaire. The household questionnaire was filled out for every residence visited. It included information on the household's location and type of construction, water and latrine facilities, a listing of all residents and a small amount of information on each person listed.
The respondent questionnaire was to be administered in "male" households to all males 15-59 years of age and in "female" households to all females 15-49 years of age listed on the household questionnaire.
The male questionnaire covered the following topics:
a. Socioeconomic and demographic characteristics, including age, religion, and socioeconomic status indicators;
b. A complete marriage and cohabitation history;
c. Contraceptive knowledge and use, including knowledge and past and current use of all family planning methods;
d. Condom utilization, including information on numbers of condoms obtained, the number used and on hand, and attitudes toward condoms and their use;
e. Male roles in the couple's decision regarding use of family planning methods and male attitudes concerning contraception and fertility;
f. Numbers of current sexual partners, coital frequency and, for 15-24 year old young adults, information on the initiation of sexual activity.
The female questionnaire covered all of the above topics except male roles and attitudes. In addition, it included:
a. Pregnancy and childbearing information and information on breastfeeding, postpartum amenorrhea, desired fertility, and the planning status of the last pregnancy;
b. Barriers to family planning use, including information on reasons for not using or for having stopped using contraception, accessibility of family planning services, satisfaction with services used or available, and other factors which may be hindering acceptance or continuation of methods;
c. Pregnancy termination, including information on reported induced abortions.
Data processing activities were carried out at the CHI using micro-computer operators hired for the survey. A CHI programmer supervised data entry and editing. Data entry and editing were done concurrently using software developed at the Centers for Disease Control (CDC) and modified for this survey. This software performs checks on the ranges of all variables, the consistency between variables and the "skip patterns" of the questionnaires. The data entry staff, thus, had the added responsibility of passing questionnaire problems on to the data manager. These staff members were trained with the interviewers to insure familiarity with the data collection instruments. A CDC programmer traveled to Haiti shortly after field work and data collection began for the purpose of installing the software and training the CHI staff in the use of the software, as well as testing and debugging the program that had been modified for the Haiti questionnaire.
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 country, 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.