Reproductive Health Survey 1993
Czech Republic, 1993
Identification
CZE_1993_RHS_v01_M
Reproductive Health Survey 1993
| Name | Country code |
|---|---|
| Czech Republic | CZE |
Other Household Health Survey [hh/hea]
The last large, national survey covering a wide array of demographic and reproductive health topics was the 1977 Czechoslovakia Fertility Survey (World Fertility Survey 1978).
The 1993 Czech Republic Reproductive Health Survey (CRRHS) was a nationally representative household survey of women 15 to 44 years of age, undertaken in order to explore important issues related to reproduction and women's health in that country. This survey was carried out at a time when substantial reforms in the Czech health care system were taking place or were being considered.
The reforms taking place have largely grown out of the fall of communism and the resultant changeover to a free market economy, the rise of democracy, and the breakup of the Czechoslovak Federation. Among the changes underway at the time of the survey were: Reforming the administrative structure of the health care system from one in which most responsibility for health care delivery resided at the regional level to one in which districts, municipalities, and the national Ministry of Health take over those responsibilities; privatization of the health sector, including hospitals, polyclinics, physician services, and pharmacies; and, reforming health care financing by introducing a system of health insurance, rather than having health care funding completely centralized and financed from general tax revenues. In this period of change and uncertainty it is important that policymakers and health care providers be aware of the current status of reproductive health throughout the population and the use of and need for various services related to family planning and maternal and child health.
The CRRHS was intended to serve several purposes. The last large, national survey covering a wide array of demographic and reproductive health topics was the 1977 Czechoslovakia Fertility Survey (World Fertility Survey 1978). Because no such surveys had been conducted in recent years on family planning and reproductive health in Czechoslovakia or the Czech Republic, this survey was the first opportunity in many years to update basic measures, such as contraceptive prevalence, unmet need for contraception, use of maternal and child health services, and many others. These measures constitute important information in determining a population's needs and problems. In recent years, at least two useful surveys have been carried out, but both were relatively limited in regard to both content or sample size (Kraus 1987; Uzel, Ketting, et al. 1992). The size and scope of the 1993 CRRHS allows highly detailed analysis with regard to many reproductive health questions for subgroups of the Czech population.
There are several issues in the area of reproductive health in the Czech Republic that were known to merit close examination prior to carrying out the 1993 CRRHS. Official statistics show that the rate of induced abortion and the ratio of abortions to live births has remained very high (although they appear to have declined recently). Reducing the number of abortions would be in the best interest of public health in the Czech Republic, regardless of one's views regarding the legality and morality of abortion. It appears that modern, highly reliable contraceptive methods are not as widely used as they are in most other developed countries. The failure to use such methods contributes to high rates of unintended pregnancy. Furthermore, pregnancy and unprotected intercourse among Czech women under 18 years of age appear to be quite widespread. In addition, relatively little data exist on the extent of certain behaviors that affect the health of women and their infants, such as smoking, alcohol consumption, and the prevalence of being overweight. The CRRHS was designed to provide information that would contribute to strategies for addressing these issues in the Czech Republic.
Besides information about health status and practices, the 1993 CRRHS measured attitudes and opinions about the quality and types of reproductive health services offered, restrictions on the availability of induced abortion, and problems associated with the use of modern contraceptives. Also, as AIDS emerges as an important public health problem, it is important to find out how well informed women are about the means of transmission and prevention of this disease.
Scope
The specific subject areas included were:
- Pregnancy and childbearing history
- Knowledge, past and present use of, and need for contraceptive methods
- Use of maternal and child health services
- Sexuality, contraception, and pregnancy among young adults (15-24 years)
- Women's health issues
- Knowledge and attitudes about selected issues in reproduction
- Knowledge about HIV/AIDS and its transmission
- Social, economic, and demographic characteristics
Coverage
The 993 CRRHS was designed to collect information from a nationally representative sample of women of reproductive age from throughout the Czech Republic.
The universe from which women were selected consisted of all females between the ages of 15 and 44 years living in households in the Czech Republic at the time that the survey was carried out. Although some pregnancy and childbearing takes place outside the ages 15 to 44, because of the relative rarity of these events at those ages, it was decided that it would be most efficient to limit the sample to women between 15 and 44 years of age.
Producers and sponsors
| Name |
|---|
| Factum (Czech public opinion research firm) |
| Czech Statistical Office |
| Name | Affiliation | Role |
|---|---|---|
| World Health Organization (WHO) Collaborating Center for Perinatal Medicine | Czech Institute for Maternal and Child Health Care | Design of the questionnaire and data analysis and financial accounting |
| Division of Reproductive Health | United States Centers for Disease Control and Prevention (CDC) | Assistance in survey design, questionnaire development and survey implementation and analysis of the data and report preparation |
| Name |
|---|
| United States Agency for International Development |
Sampling
Selection of survey respondents was made using a three-stage cluster sampling design. First, 166 pairs of basic census enumeration districts (ED) (332 districts in all) were systematically selected from throughout the country, with the probability of selection proportional to the population of each census enumeration district. These 332 ED constituted the primary sampling units (PSU) for the survey. The sample included ED in all but one of the Czech Republics 85 districts. The Czech Statistical Office then made listings available of all known dwelling units in each of the selected ED. These listings were taken from the 1991 Population and Housing Census. Because the selection of PSUs was done proportional to population size and no strata were oversampled, the sample was geographically selfweighting.
In the second stage of selection, individual dwelling units were sampled from each ED chosen in the first stage. Because of slightly different mean numbers of 15 to 44 year-old women per household according to size of community, the number of dwelling units selected in an ED depended on the population of the community in which the ED was located, as follows:
Community Selected / Population Dwelling Units
1 - 4,999 : 43
5,000 - 19,999: 37
20,000+ : 39
One dwelling unit was randomly chosen from the each PSU. That unit and the 42, 36, or 38 units (depending on the community population) listed next constituted the sample of housing units to be visited. No substitution was made for dwellings that were unoccupied or where nobody was at home. By varying the number of units selected according to the average number of women per household we were able to hold fairly constant the average numbers of women living in selected ED across population size categories. If the selected ED had fewer households than were needed to constitute a sample sector, all dwelling units were sampled and another ED (the one with the next highest number, which was usually geographically contiguous) was added to the sample.
The third and final stage of sampling consisted of the selection of individual respondents. In households in which more than one woman between the ages of 15 and 44 lived, the household respondent was asked for a listing of all such females. One of those women was randomly selected by the interviewer to be the survey respondent. During analysis each record was weighted by the number of women of childbearing age in the household, to compensate for the fact that only one woman per household was selected with probability inverse to the number of women in the household.
A total of 12,747 dwelling units were visited in the 332 sample sectors. Women eligible for interview were identified as residing in 38.2% of these units. In 52.2% of the households visited there were no 15-44 year-old women resident. Nobody was home at 4.0% of units and 3.0% of units were uninhabited. Residents refused to be interviewed at 2.7% of households. As the size of community increased, the percent of refusals increased slightly and the percent of households with no eligible females decreased.
Of the 4,870 households identified as containing at least one 15-44 year-old female, interviews were completed in 4,497, for a completion rate of 92.3%. Of those women selected, 5.3% refused to be interviewed and 1.6% were not found at home, even after repeated visits. The completion rate fell slightly as community size grew, from 93.8% in towns under 5,000 population to 91.3% for cities with at least 20,000 people.
In regard to residential characteristics, the sample appears to have been highly representative of the 15-44 year-old population of the Czech Republic. Of the Czech Republic's nine regions, only in Prague was the difference between the 1991 Census and CRRHS proportion of the female 15-44 year-old population as much as two percentage points. Likewise, the census and survey distributions of size of place of residence were very similar.
Survey instrument
The CRRHS questionnaire covered a wide range of topics in the area of reproductive health and related areas.The specific subject areas included were:
- Pregnancy and childbearing history
- Knowledge, past and present use of, and need for contraceptive methods
- Use of maternal and child health services
- Sexuality, contraception, and pregnancy among young adults (15-24 years)
- Women's health issues
- Knowledge and attitudes about selected issues in reproduction
- Knowledge about HIV/AIDS and its transmission
- Social, economic, and demographic characteristics
Data collection
| Start | End |
|---|---|
| 1993-02 | 1993-06 |
Data collection for the CRRHS was done by about 150 female interviewers from throughout the Czech Republic. Each interviewer received intensive training in interviewing techniques, field procedures, and proper administration of the CRRHS questionnaire. Each interviewer attended one of three training courses held immediately before the beginning of field work, two in Prague and one in Brno in January and February of 1993. These sessions included practice interviews with community residents to ensure that interviewers were able to administer the questionnaire competently.
Data collection lasted from February until June of 1993. Each interviewer was assigned to visit selected households in from one to three sample sectors easily accessible from her home. Interviews were conducted at respondents' homes and generally lasted between 30 and 60 minutes. Completed questionnaires were sent by the interviewer to her supervisor, who reviewed each questionnaire and, if it was satisfactorily completed, forward it to the Czech Statistical Office.
Data processing
At the Czech Statistical Office questionnaires were reviewed once more by coordinators and sent for data entry. Questionnaire data was entered onto computers by three clerks, using SURVEY, a data entry/editing package developed at CDC. The SURVEY package is used to edit data as they are entered, by identifying data items that are outside of their allowable range or inconsistent with other information on the questionnaire.
Data appraisal
Survey estimates of percentages, proportions, means, or other measures are subject to two types of errors: sampling error and non-sampling error. Non-sampling 'errors can arise from many sources, most notably interviewer errors, inaccurate information given by respondents (intentionally or unintentionally), data processing errors, failure to administer the interview to the correct households or individuals, as well as other factors. Although efforts were made during the design, training and implementation of the 1993 CRRHS to minimize the incidence and severity of non-sampling errors, it is virtually impossible to eliminate all sampling error or to evaluate their importance statistically.
Sampling error is a measure or the variability between all possible samples of a given size that could have been selected from the population being studied (in this case, women between the ages of 15 and 44 throughout the Czech Republic) using a given sample design. Sampling error is measured in terms of the standard error for a particular statistic, which is the square root of the variance of that statistic. The standard error can be used to calculate confidence intervals around estimated statistics. The confidence interval most commonly used is the 95 percent interval, which defines the range within which the actual percentage, mean, etc. falls, in the absence of all non-sampling error. The 95 percent confidence interval for a statistic is the estimated value plus or minus 1.96 (approximately two) times the standard error of the estimate.
The standard errors of statistics estimated using a multistage sample design, such as that used in the 1993 CRRHS are more complicated to calculate and are larger than standard errors based on simple random samples. The software package SUDAAN (Research Triangle Institute, 1993) was used to compute the CRRHS standard errors with the appropriate statistical methodology. In addition to standard errors, SUDAAN was used to compute the design effect for each estimate, which is defined as the estimated ratio between the variance using the multistage cluster design actually used and the variance that would have resulted if a simple random sample had been used. Therefore, the design effect demonstrates the increase in the increased variability of an estimate (including the standard error and confidence interval) due to the use of a complex survey design.
Standard errors, design effects, and 95% confidence intervals (the limits of which are P-2SE and P+2SE) are presented in Table A.2 of the Final Report for variables considered to be of major interest. For each variable, the definition and the population used for the calculation are presented in Table A.1.
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_CZE_1993_RHS_v01_M_WBDG