1988 Family Health Survey Final English Language Report Revised December 1989

Publication date: 1989

EDITOR'S NOTE In 1988, the Asociación Demográfica Salvadoreña (ADS), the International Planned Parenthood Federation affiliate in El Salvador, with technical assistance from the Division of Reproductive Health, Centers for Disease Control, conducted a Family Health Survey (FESAL88) to examine fertility, contraceptive behavior, use of maternal and child health services, immunization levels and prevalence of diarrhea among children less than five years of age, and infant mortality. A final English-language report and a final Spanish-language report were released in February and March, respectively, 1989. In September 1989, a seminar/workshop was held in San Salvador to disseminate the major findings of the survey to officials from government organizations interested in reproductive health issues. During the secondary analysis of the survey data set that was conducted in preparation for the seminar/workshop, it was discovered that incorrect weighting factors had been used in the original fertility and infant mortality analysis. This error particularly affected the findings on infant mortality. However, once the weighting factors were corrected, the findings revealed the same tendencies and differentials as were originally presented in the final reports. We decided to reprint the English-language report in its entirety rather than release an errata, which could easily be separated from the main report. Thus, this report replaces the one released in February 1989. We regret if our oversight may have caused you any inconvenience. However, the release of this report reaffirms CDC's commitment to excellence in survey research. Richard S. Monteith Division of Reproductive Health Centers for Disease Control December 1989 FAMILY HEALTH SURVEY EL SALVADOR 1988 Final English Language Report Revised December 1989 ASOCIACIÓN DEMOGRÁFICASALVADOREÑA Luis Sidney Castro E., Survey Director Jose Mario Caceres H., Survey Coordinator Jean Carlos Salguero, Field Coordinator With the Technical Collaboration of: Richard S. Monteith, M.P.H. Howard I. Goldberg, Ph.D. Steve A. Kinchen, B .S. I .C .S . U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES PUBLIC HEALTH SERVICE Centers for Disease Control Center for Chronic Disease Prevention and Health Promotion Division of Reproductive Health Atlanta, Georgia 30333 1988 FAMILY HEALTH SURVEY (FESAL88) Table of Contents Page Tables Contributors to the Survey 1 Summary 3 I. Introduction 11 II. Survey Methodology 13 2-1 III. Fertility 17 3-1 to 3-12 IV. Knowledge of Contraceptive Methods 25 4-1 to 4-3 V. Current Use of Contraception 27 5-1 to 5-16 VI. Source of Contraception 37 6-1 to 6-4 VII. Reasons for Nonuse of Contraception and Desire to Use in the Future 41 7-1 to 7-8 VIII. Sterilization and the Demand for Sterilization 45 8-1 to 8-10 IX. Characteristics of Women in Need of Family Planning Services 51 9-1 to 9-3 X. Preferences in Family Planning Services 53 10-1 to 10-9 XI. Sexual Experience and Contraceptive Use: Women 15-24 Years of Age 55 11-1 to 11-15 XII. Use of Maternal and Child Health Services 61 12-1 to 12-14 XIII. Immunization Levels 71 13-1 to 13-12 XIV. Prevalence of Diarrhea and its Treatment 75 14-1 to 14-4 XV. Infant and Child Mortality 79 15-1 to 15-2 References 83 CONTRIBUTORS TO THE SURVEY This English language report was prepared by Richard S. Monteith, Howard Goldberg, and Raul Romaguera, and reviewed by Leo Morris, Program Evaluation Branch, Division of Reproductive Health, Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control, Atlanta, Georgia 30333, U.S.A. The Spanish language report has been published by the Asociación Demográfica Salvadoreña (ADS), San Salvador, El Salvador. We would like to acknowledge the valuable contributions made to the various phases of the survey by the following people: ASOCIACIÓN DEMOGRÁFICA SALVADOREÑA (ADS) Marco Antonio Batres, Executive Director David Araya, Director of Planning Luis Sidney Castro, Chief, Dept. of Evaluation and Investigations Jose Mario Caceres, Survey Director Juan Carlos Salguero, Field Coordinator Rolando Godoy, Consultant Jose Alfredo Rivera, Survey Administrator Rosa de Romero, Data Entry/Editing Cecilia Campos, Data Entry/Editing Nora de Barrios, Secretary CENTERS FOR DISEASE CONTROL DIVISION OF REPRODUCTIVE HEALTH Steve Kinchen, Programmer, Research and Statistics Branch Charles Warren, Demographer, Program Evaluation Branch Evelyn Finch, Programmer, Research and Statistics Branch Elizabeth Fitch, Secretary, Research and Statistics Branch Lynda Royer, Secretary, Program Evaluation Branch Evelyn Alexander, Secretary, Program Evaluation Branch Glenda Thornton, Secretary, Program Evaluation Branch USAID/EL SALVADOR Kevin Armstrong, Population Officer Raul Guillermo Toledo, Assistant Population Officer -3- SUMMARY Introduction The 1988 Family Health Survey (FESAL88) is the fourth national survey since 1975 to examine fertility and contraceptive behavior in El Salvador, and the second survey since 1985 (ADS y IRD, 1987) to examine the use of maternal and child health services and immunization levels for children less than 5 years of age. During fieldwork, which was conducted from May 2 to July 24, 1988, 3,579 women 15-44 years of age were interviewed. Data entry and editing using microcomputers occurred concurrently with fieldwork. A preliminary report was released in September, 1988 (ADS y CDC, 1988), and the El Salvador Demographic Association (ADS) prepared the final Spanish-language report (Asociación Demográfica Salvadoreña, 1989). The Centers for Disease Control has prepared this English-language report for non-Spanish speakers, which corresponds to the Spanish version, but does not include some tables that were primarily of local interest. Fertility Between 1978 (Morris et al., 1979) and 1985 the total fertility rate (TFR) in El Salvador declined from 6.3 to 4.5 births per woman. However, there was no statistically significant change in fertility between 1985 and 1988. This holds for the country as a whole and by place of residence. Given an estimated crude death rate of 7 per 1,000 population and crude birth rate of 35 per 1,000 (Population Reference Bureau, 1988), the rate of natural increase in El Salvador is about 2.8 percent. If fertility and death rates were to remain constant, El Salvador's 1988 population of 5.4 million would double in about 25 years. The survey data suggest that future declines in fertility will largely depend on an increase in contraceptive use in rural areas. -4- Breastfeeding In 1988, a slightly higher proportion of women reported having breastfed their last child (95 percent) than women in 1985 (92 percent). Mean duration of breastfeeding in 1988 was 17 months, one month longer than the duration reported in 1985. The greatest increase in duration of breastfeeding from 1985 to 1988 occurred among the most highly educated women. Contraceptive Use Overall, 47 percent of married women aged 15-44 were found to be using contraceptive methods at the time of the survey. The most prevalent method was female sterilization, which accounted for 63 percent of all use of contraceptive methods. In contrast, the use of reversible methods in El Salvador was very low. The metropolitan area of San Salvador had the highest contraceptive prevalence rate with slightly less than two-thirds of married women reporting current use, compared with 56 percent of married women living in other urban areas, and only 34 percent of married women living in rural El Salvador. Contraceptive use also varied by age, number of living children, education, and work status. Use was lowest among younger women, women with few or no living children, women with little or no formal education, and women not in the labor force. The median age and median number of living children at time of first use was 28.3 years of age and 2.6 living children. On average, women are married eight years before they initiate the use of contraception. It is evident that married women in El Salvador have strong pressures to bear children once they are married and little emphasis is given to spacing children. Contraceptive use increased by almost 26 percentage points between 1975 and 1988. However, virtually all of the increase occurred by 1985. Between 1985 and 1988, there was no significant increase in contraceptive use. Seventy-eight percent of the increase in prevalence since 1975 was in the use of female sterilization. -5- Although the use of female sterilization is high in El Salvador, the survey data suggest that female sterilization has had little effect on lowering fertility in recent years. On average, married women have 3.5 children at time of sterilization. A decline in fertility is not likely to occur in El Salvador as long as couples continue to rely almost exclusively on sterilization to limit family size and the use of reversible methods to space births or delay the start of childbearing remains extremely low. Source of Contraception The most important suppliers of contraception in El Salvador in 1988 were the Ministry of Health (57 percent of all users of contraception), the El Salvador Demographic Association (13 percent), and the Social Security Institute (11 percent). In all three residential locations, the Ministry of Health was the principal supplier of contraception. The Demographic Association was the second ranking supplier of contraceptives in rural areas. In general, each provider of contraception in the country was serving about the same percentage of total users in 1988 as it was in 1985. Reasons for Nonuse of Contraception and Desire to Use in the Future The most important reasons for not using contraception were reasons related to pregnancy, fecundity, and lack of sexual activity, which were given by 73 percent of nonusers. The major reasons for nonuse not related to pregnancy and sexual activity were fear of side effects of contraception (8 percent), spouse opposition to the use of contraception (5 percent), and religious reasons (3 percent). Only half of fecund nonusers wanted to use a method now or in the future. But only 6 percent of nonusers expressed an interest in using a method immediately. Of the women who desired to use now or in the future, 89 percent knew where to obtain contraception, with female sterilization being the method of choice. -6- Women in Need of Family Planning Services Overall, slightly less than one woman in five and 30 percent of married women were found to be at risk of an unplanned pregnancy. Sixty-eight percent of women in need of services live in rural El Salvador, 90 percent are married or are living in consensual unions, 61 percent have 3 or fewer years of schooling, and almost 84 percent are not gainfully employed. In addition, about 53 percent have 3 or more living children. In conclusion, the survey data indicate that the family planning program of El Salvador should be oriented toward high parity, married, less-educated, non-working women and rural women. However, the potential demand for family planning services is greater than suggested by these data, because the measure of need used here only reflects women who did not desire to become pregnant. Many breastfeeding and pregnant women not currently considered in need of services require, or will shortly require, family planning services. Preferences in Family Planning Services Although the public sector family planning program in El Salvador currently does not provide services on weekends, 28 percent of users of reversible methods and nonusers who desire to use now or in the future preferred a weekend day to receive family planning services. This proportion was 41 percent among working women. Eighty-two percent of the women stated that they preferred to receive family planning services from a female physician, and 75 percent said that it would be acceptable to them to receive these services from a nurse practicioner than from a physician. Only 38 percent of the women stated that they had interest in receiving family planning services from trained, nonmedical personnel living in the community. Young Adult Sexual Behavior A special module for 15-24 year-old respondents was added to the 1988 survey to obtain information on premarital sexual experience -7- and the use of contraception at first sexual experience. Overall, 23 percent reported premarital sexual intercourse, with 16 percent of 15-19 year-olds and 32 percent of 20-24 year-olds reporting premarital sexual intercourse. The mean age at first premarital intercourse was 16.4. Only 5 percent of these young adults used contraception at the time of their first premarital intercourse, the lowest use level reported in the Latin American region (Morris, 1987). Forty percent of nonusers at first premarital intercourse reported that they did not expect to have intercourse at that time as their reason for nonuse. An additional 21 percent stated that they had no knowledge of contraception, and 16 percent stated that they had wanted to become pregnant. Eighty-seven percent of all women aged 15-44 agreed with the statement that there should be special family planning services for young adults. Use of Maternal and Child Health Services A higher percentage of women with at least one live birth within 5 years of interview stated that they utilized well-baby care services (76 percent) than prenatal services (67 percent) or postpartum services (40 percent) associated with their most recent birth. The proportion of women receiving these services was highest for women living in metropolitan San Salvador, among women with 10 or more years of schooling, and among women who delivered in private or Social Security hospitals. Only 34 percent of the women said they had used all three services. Conversely, 14 percent reported that they used no MCH services. Fifty-two percent of the women reported that their last child was delivered in a medical facility, ranging from 88 percent among women living in the metropolitan area to 34 percent for rural women. In rural areas, midwives delivered 54 percent of infants, while an additional 9 percent were attended by family members. Seventeen percent of all last hospital deliveries were Cesarean. The percentage of women whose last hospital delivery was Cesarean was -8- highest for urban women, women with 10 or more years of schooling, and low parity women. The highest proportion of Cesarean deliveries occurred in private hospitals. Immunization Coverage Levels of primary immunization ranged from 61 percent for polio and DPT to 77 percent for BCG and 81 percent for measles. Only one vaccine had a higher level of complete immunization in 1988 than in 1985 --BCG. Both polio and DPT decreased by 10 percentage points from 1985 to 1988, while measles decreased by about 3 percentage points. Most of the decrease in levels of primary immunization against polio and DPT occurred in the metropolitan area of San Salvador. The Ministry of Health was the principal source of vaccinations in El Salvador. * Prevalence of Diarrhea and Its Treatment Twenty-nine percent of children less than 5 years of age were reported to have had diarrhea during the two weeks prior to interview. Rural children and children less than 2 years of age were more likely to have had diarrhea than urban and older children. Of the children who reportedly had diarrhea, 48 percent were reported to have had blood and/or mucous in their stool, a measure of severity. Of the children who were treated for their recent episode of diarrhea, only 13 percent were treated with oral rehydration solutions. Nearly 81 percent were treated with various popular pharmaceutical products, whose effectiveness in treating most diarrhea is questionable. Infant and Child Mortality A child born in El Salvador in the five years before the survey had a .055 probability of dying before reaching his or her first -9- birthday. That is, about 55 of every 1,000 children born alive died during infancy. The probability of dying between birth and the fifth birthday was .075. Infant mortality probabilities according to the 1985 and 1988 surveys suggest that a rapid decline in infant mortality has occurred over a short period of time. Overall, the probability of dying in infancy fell from .065 to .055. Most of the decline in mortality occurred in rural areas, where it fell from .081 to .061. For other urban areas, mortality declined from .056 to .051., while in the San Salvador metropolitan area, mortality declined from .046 to .041. -11- THE 1988 FAMILY HEALTH SURVEY (FESAL88) I. INTRODUCTION The 1988 Family Health Survey, popularly known as FESAL88, was conducted principally to provide estimates of fertility, prevalence of use of contraception, and source of contraception in El Salvador since a similar survey was conducted in 1985 (ADS y IRD, 1987). The 1988 survey was also designed to cover a wide range of maternal and child health topics to measure program impact. These topics include the use of prenatal, postpartum, and well-baby services, breastfeeding prevalence and duration, levels of immunization, and the prevalence of diarrhea among children less than 5 years of age and the treatment they received, if any. In addition, for the first time in El Salvador, a special module for women aged 15-24 was included to obtain information on sexual experience and use of contraception at first sexual experience. The objectives of the 1988 survey included the following: 1. Obtain fertility estimates at the national level. 2. Determine levels of knowledge and current use of contraceptives for a variety of social and demographic background factors; also determine the source where users obtain contraception. 3. Determine reasons for nonuse of contraception, and estimate the percentage of women who are in need of family services. 4. Examine the sexual experience, use of contraception, and general knowledge concerning reproductive events among females 15-24 years of age. -12- 5. Examine the use of maternal and child health services and immunization levels for children less than 5 years of age; also examine the prevalence of diarrhea among these children. 6. Estimate levels of infant and childhood mortality. The Asociación Demográfica Salvadoreña (ADS), the International Planned Parenthood Federation (IPPF) affiliate in El Salvador, with technical assistance from the Centers for Disease, Control, conducted the 1988 survey. Interviewer training was completed in April 1988 and fieldwork was conducted from May 2 to July 24, 1988. Data entry and editing using microcomputers occurred concurrently with field work. In September, ADS issued a preliminary report (ADS y CDC, 1988). CDC prepared this English language report for non-Spanish speakers, which corresponds to the Spanish version, but does not include some tables that were primarily of local interest. The survey was supported by the United States Agency for International Development (AID). -13- II. SURVEY METHODOLOGY The 1988 survey was an area probability survey with three stages of selection. The sampling frame used in the 1985 survey (ADS y IRD, 1987), which was updated during the period February 1987-March 1988, served as the sampling frame for this survey. In the first stage, a systematic sample with a random start was utilized to select census sectors with probability proportional to the number of households in each sector. Within chosen sectors, clusters of households were randomly selected for interview. The third stage consisted of selecting for interview one woman 15-44 years of age, regardless of marital status, from among all women in this age group living in the selected households. Three strata or "domains" were sampled independently-- the metropolitan area of San Salvador (referred to as AMSS in the tables of this report), other urban areas, and rural areas. In the metropolitan area, 1,500 households were selected for interview. In order to have a geographically self-weighted sample, 1,650 and 2,310 households were selected in the "other urban" and "rural" strata, respectively. For security reasons, it was not possible to conduct interviews in three selected rural segments, representing 90 households. No substitutions were made for these segments. Thus, during fieldwork, which was conducted May 2-July 24, 1988, a total of 5,370 households were visited. Although the sample was geographically self-weighted for areas of the country that were considered accessible or non-contested at the time the sample was selected, 10 percent of the households in the country were excluded from the sampling frame for security reasons. Approximately 19 percent of households in rural areas and 4 percent of households in "other urban" areas were not considered safe for interviewing. Thus, for country-wide estimates, weighting factors were applied to each of these strata to compensate for the inaccessible households. The use of weighting factors assumes that the characteristics of the population that was living in the inaccessible areas were similar to those of the population that was living in neighboring but accessible areas. -14- Only one woman aged 15-44 per household was selected for interview. Each respondent's probability of selection was inversely proportional to the number of eligible women in the household. Thus, weighting factors were also applied to compensate for this unequal probability of selection. In the tables that follow, proportions and means are based on the weighted number of observations, but the unweighted numbers of cases are shown. Weighting of the 1985 survey data was also necessary in order to make trend analyses from one survey to the other. Thus, weighting factors were applied to the 1985 data set to compensate for the households that were inaccessible when that survey was conducted. It was estimated that 30 percent of rural households and 13 percent of "other urban" households were not accessible in that year. The final report on the 1985 survey did not make adjustments for inaccessible areas. Thus, 1985 data presented in this report differ slightly from that published in the 1985 report. For the country as a whole, in the 1988 survey, the variable "current use of contraception" for married women 15-44 years of age has an estimated sampling error of plus or minus 3.2 percent within a 95 percent confidence interval, including the estimated design effect. Based on the unweighted numbers of cases, the sampling errors for metropolitan San Salvador, other urban areas, and rural areas were 5.6 percent, 5.4 percent, and 4.7 percent, respectively, with a confidence interval of 95 percent. As shown in Table 2-1, 69 percent of the 5,460 households in the sample included, or may have included, at least one woman aged 15_44. The proportion of households in which women eligible to be respondents were identified was essentially the same in each of the three strata. The bottom panel of Table 2-1 shows that complete interviews were obtained in 95 percent of those households that had or may have had eligible respondents, for a total of 3,579 interviews. Interview completion rates ranged from 90 percent in metropolitan San Salvador to slightly less than 97 percent for the remainder of the country. Overall, the refusal rate was 1.3 percent, with the refusal rate highest in the metropolitan area (3.4 percent). -15- Finally, data entry, including range and consistency checks, was conducted concurrently with fieldwork using microcomputers and specially designed software. If a questionnaire was found to contain errors that could not be corrected on the spot, the questionnaire was returned the following day to the field for corrections. During subsequent and more detailed data checks following fieldwork, less than 5 percent of the questionnaires were identified as containing errors. As a result, preliminary tabulations were available within 15 days of completion of fieldwork and a preliminary report was available within 40 days. -17- III. FERTILITY In this chapter, estimates of fertility rates for women interviewed in the 1988 survey are presented. These estimates are compared with estimates from the 1985 survey (ADS y IRD, 1987). In addition, breastfeeding, age at first marriage, and contraceptive use and their effects on fertility are discussed. Fertility Rates Table 3-1 shows the mean number of children ever-born according to the 1985 and 1988 surveys, by age of the respondents and residence. Focusing first on the 1988 survey data, as expected, mean parity was substantially higher at every age in rural areas than in urban areas. For women in age group 40-44 (an estimate of completed fertility), women living in rural El Salvador had on average 6.7 children compared to 4.7 children for women living in other urban areas and 3.9 for women living in metropolitan San Salvador. Overall, according to the 1988 survey results, 40-44 year old women in El Salvador have an average of 5.3 children. As shown in Table 3-2, mean parity is negatively associated with educational attainment within every age group. Among women aged 40-44 years, mean parity ranged from 2.2 children among women with 10 or more years of schooling to 6.5 children among women who did not attend school. Comparing mean parity of women in the 1988 survey with that of women in the 1985 survey (Table 3-1), we observe, overall, a slight decrease in the mean number of children born alive, from 2.5 in 1985 to 2.2 in 1988. Most of the decrease appears to have occurred among rural women and women with no formal education (Tables 3-1 and 3-2). The relationships mentioned above between residence and education with fertility were present in both surveys. The total fertility rate (TFR) was calculated for the births that occurred in the five years prior to interview. As shown in Table 3-3 and Figure 3-1, there was no statistically significant change in fertility in El Salvador between 1985 and 1988. This holds for the country as a whole and by place of residence. Based on these results, we conclude that fertility did not change between the 1985 and 1988 surveys. The following is a discussion of the major determinants of fertility: breastfeeding, nuptiality, and contraceptive use. Abortion, which also directly affects fertility, is not discussed in this section because reliable abortion histories, which are difficult to obtain in cross-sectional surveys, were not obtained in the 1988 survey. Breastfeeding Breastfeeding is an important determinant of fertility since its duration and intensity affect the length of the postpartum anovulatory period (McCann et al., 1981). Overall, 95 percent of -19- women who reported a live birth within 24 months of interview said they breastfed that child (Table 3-4). A slightly higher proportion of women living outside of the San Salvador metropolitan area (95 to 96 percent) reported breastfeeding their last child than women living in the metropolitan area (93 percent). In 1988, a slightly higher proportion of women reported having breastfed their last child (95 percent) than respondents in the 1985 survey (92 percent). Although increases are observed for each stratum, most of the increase occurred in other urban areas. Overall, mean duration of breastfeeding in 1988 was approximately 17 months, one month longer than the duration reported in 1985 (Table 3-5). In 1985 and 1988, rural women, on average, breastfed their children longer than their urban counterparts. In addition, duration of breastfeeding was negatively associated with education (Figure 3-2) which is similar to patterns found elsewhere in Latin America (Anderson et al., 1983). With the exception of women with 1 to 3 years of schooling, increases in duration of breastfeeding from 1985 to 1988 are noted for every educational category shown. The greatest increase in duration of breastfeeding from 1985 to 1988 occurred among the most highly educated women. In addition to duration of breastfeeding, intensity of breastfeeding also affects the length of the postpartum anovulatory period. For this reason (and because of the effect of infant feeding patterns on health), women in the 1988 survey were asked when supplemental feeding of their children began. Table 3-6 shows that the introduction of supplemental feeding varies by residence. Overall, children were 10 months of age, on average, when they were introduced to non-maternal milk and about 5 months old when they were first given solids. It is noteworthy that women living in the metropolitan area introduced other milk to their children almost 7 months before rural women and 5 months before women living in other urban areas. In addition, the introduction of solid foods begins earlier in urban areas than in rural El Salvador. By the time children are one year of age, 59 percent have been introduced to other milk and almost 100 percent are on solid food (Data not shown). Current recommendations are that children be started on solid foods between 4 and 6 months of age (MSPAS, 1986). Clearly, many infants are receiving solids earlier than is advisable in El Salvador, especially in urban areas. Nuptiality A second intermediate or "proximate" determinant of fertility is the degree to which the reproductive age population is sexually active. In a population in which most childbearing takes place within unions, this can be approximated by nuptiality or the proportion of time women .spend in formal or informal unions. Table 3-7 shows the proportion of the female population reporting that they had ever been married or in a consensual union, according to age and residence. Overall, 71 percent had ever been in union. The -21- proportion increases with age up to 30-34, where it reaches the mid-90's. Apparently, about 4 percent of women never marry. The percentage ever-in-union is considerably higher in rural than in urban areas except at the oldest ages examined. Table 3-8 shows that, like the ever-married results, greater percentages are currently married in rural areas. The overall percent currently in union reaches its peak at age 30-34. The percent of rural women who are currently married was 62 compared to 54 percent among women living in other urban areas and 49 percent among women living in the metropolitan area. Table 3-9 shows the singulate mean age at marriage (SMAM), a summary measure of the age at first marriage among women who have married by age 45, according to the 1985 and 1988 surveys. Average age at first marriage was 20.3 years of age, with a difference of 2 years between urban and rural areas. The figures in this table suggest that the age at entry into union may have increased slightly over the 3-year period, from 19.8 to 20.3. All of the increase appears to have taken place in the metropolitan area of San Salvador and other urban areas. The SMAM in rural areas remained unchanged at 19.1 years. Contraceptive Use Contraceptive use will be discussed in detail in Chapter V. of this report. However, since contraceptive use is an important determinant of fertility, summary findings of the 1988 survey will be presented here. Overall, 47.1 percent of married women aged 15-44 were found to be using contraceptive methods at the time of the survey. Women living in the metropolitan area of San Salvador had the highest prevalence rate, with slightly less than two-thirds of married women reporting current contraceptive use compared with 56 percent of married women living in other urban areas and only 34 percent of married women living in rural El Salvador. The most prevalent method was female sterilization, which accounted for 63 percent of all use of contraceptive methods. -22- A comparison of the results of the 1988 survey with those of the 1985 survey (ADS y IRD, 1987) shows that contraceptive use among married women changed negligibly between 1985 and 1988. The trends in these factors can be combined to make a general statement concerning fertility in El Salvador. Between 1985 and 1988, fertility did not change significantly (the difference in TFR between 1985 and 1988 is not significant). The lack of change in fertility was likely due to the lack of significant change in contraceptive use, breastfeeding practices, or age at first marriage between the two surveys. Results from this study show that contraceptive use is probably the most important factor affecting fertility in El Salvador. The survey data suggest that future declines in fertility will largely depend on an increase in contraceptive use in rural areas. As Chapter V. will show, future program efforts should be especially directed toward women in rural areas and encouraging young couples, regardless of place of residence, to space their children more effectively by using temporary methods of contraception. Planning Status of Last Pregnancy and Current Pregnancy Intention Future fertility of the women examined in this report may be influenced by the planning status of their last pregnancy and their desire for future births. All ever-pregnant women were asked two questions about the planning status of their last pregnancy: "Was the last pregnancy desired?" If not, "Did you want no more children, or did you want some, but later?" On the basis of these questions, each woman's last pregnancy was classified as either "planned", "mistimed", "unwanted", or "unknown". Planned pregnancies were defined as those that were desired; mistimed pregnancies were classified as those that were desired, but at some time in the future; and the unwanted pregnancies were those not desired, even at a future time. Using this scheme, the mistimed and unwanted pregnancies can be combined as an estimate of unplanned pregnancies. -23- Based on these definitions, 80 percent of respondents' most recent pregnancies were reported as planned, 13 percent as mistimed, and 7 percent as unwanted (Table 3-10). Less than 1 percent of pregnancies could not be classified. Thus, 20 percent of recent pregnancies were unplanned. The proportion of pregnancies that were reported to have been planned did not vary appreciably by place of residence. For women who reported that their last pregnancy was unplanned, about twice as many women in each stratum said their pregnancies were mistimed rather than unwanted. The proportion of pregnancies that were reported as unwanted increased with both age and number of living children. In addition, there was a negative association between unwanted pregnancies and education; women with no formal education were 4 times as likely as women with 10 or more years of schooling to report their last pregnancy as unwanted. Table 3-11 shows current pregnancy intentions of married women. Overall, three-fourths stated that they did not desire a pregnancy at the time of the survey, while 12 percent desired to become pregnant at that time. Another 12 percent were currently pregnant. The proportion of women currently pregnant of desiring pregnancy increased as place of residence became less urban, but decreased with age and number of living children. It is evident that married women with no children have strong pressures to have a child; 86 percent are pregnant or desire a pregnancy. Another factor related to current pregnancy is the proportion of women who want no more children. All fecund women who had at least one living child were asked if they wanted any more children (Table 3-12). Overall, more than one-third of the women said they did not want any more children, with the percentage highest in metropolitan San Salvador (42 percent) and lowest in the rural area (32 percent). As one would expect, the proportion of women who did not -24- want any more children increased with age and family size. It should be noted that urban women, particularly those living in the metropolitan area, want to terminate childbearing earlier in life and at lower parities than rural women. What these women were doing to control their fertility at the time of the survey is discussed in Chapter V. -25- IV. KNOWLEDGE OF CONTRACEPTIVE METHODS All women aged 15-44 were asked if they had ever heard of the different contraceptive methods available in El Salvador. Specifically, they were asked, "Have you ever heard of (Method 'X')?". Then local and popular names of each method were read by the interviewer. The findings of the survey show that knowledge of contraception in El Salvador is virtually universal. Nearly 99 percent of women 15-44 years of age have knowledge' of at least one method of contraception (excluding withdrawal). Although there was virtually no difference between women living in metropolitan San Salvador and in other urban areas (99 percent), knowledge of at least one method was slightly lower among women living in rural areas (97 percent). Table 4-1 shows wide variation in knowledge of individual methods. For all women aged 15-44 the most widely known methods of -26- contraception are female sterilization (95 percent), oral contraceptives (94 percent), and condoms (90 percent). The least well known methods included rhythm (46 percent), withdrawal (15 percent), and Billings (11 percent). Although the ranking of methods is consistent within each residential group, knowledge of individual methods decreases as place of residence becomes less urban. Table 4-2 shows that knowledge of modern contraceptive methods (excluding vaginal and natural methods) is consistently high (above 75 percent) for all women except teenagers. Although women aged 15-19 have less knowledge of individual methods, 83 percent and 86 percent reported knowledge of oral contraceptives and female sterilization, respectively. Nevertheless, knowledge of reversible methods such as condoms, IUDs, and injectables is relatively low among women in this age group. In general, the least well known methods in every age group are vaginal methods, natural methods, and withdrawal. Knowledge of the more effective methods of contraception did not vary significantly by educational attainment (data not shown). Over the three year period that elapsed between the 1985 and 1988 surveys, knowledge of contraceptive methods increased substantially for all methods except withdrawal (Table 4-3 and Figure 4-1). The most dramatic increase in knowledge of contraception occurred in rural El Salvador. Despite this increase, in 1988 the percentage of rural women reporting knowledge of contraception remains lower than that reported by their urban counterparts. Finally, there was no significant change in the ranking of methods from the 1985 survey to the 1988 survey. -27- V. CURRENT USE OF CONTRACEPTION This chapter covers contraceptive use according to variables typically associated with use: residence, age, number of living children, education, and work status. With the exception of the first table, the results presented here focus on the level of use found among women currently married or living in consensual unions who are 15-44 years of age. Overall, 31 percent of all women aged 15-44 were found to be using contraceptive methods at the time of the survey (Table 5-1 and Figure 5-1). The prevalence of contraceptive use varied from 47 percent among married women to less than 1 percent among single women. The most prevalent method used by ever married women was female sterilization followed by oral contraceptives, the IUD, condoms, and rhythm and the Billings method (Figure 5-2). Among formerly married women, 89 percent of contraceptive use was attributed to the use of female sterilization. -28- As shown in Table 5-2, metropolitan San Salvador had the highest contraceptive prevalence rate among the residential categories with slightly less than two-thirds of married women reporting current use, compared with 56 percent of married women living in other urban areas and only 34 percent of married women living in rural El Salvador. Thus, place of residence is a very important factor influencing the use of contraception, with urban women exhibiting higher levels of use than rural women. Table 5-2 also shows that the most prevalent method used by married women in El Salvador is female sterilization, which accounts for 63 percent of all contraceptive use. The comparable percentages by residence are 57 percent for metropolitan San Salvador, 60 percent for other urban areas, and 70 percent for rural areas. In contrast, the overall use of reversible methods is relatively low, ranging from a high of almost 8 percent for oral contraceptives to less than 1 percent for vaginal methods. With the exception of oral contraceptives, the use of reversible methods in rural areas is almost nil. Contraceptive use varies by age, reaching a peak among married women 30-39 years of age (Table 5-3 and Figure 5-3). Use was lowest for married women 15-19 (17 percent), but more than doubles to 37 percent among 20-24 year olds. Note that by the time women are 25 years of age or older, more than 50 percent are using some form of contraception. Oral contraceptives is the most used method among married women 15-24 years of age, but after age 25 the most prevalent method is female sterilization. A further illustration of the increasing prevalence of female sterilization as women age is that this method now accounts for 23 percent of total use among women 20-24 years of age, 75 percent among women 30-34, and 90 percent among women aged 40-44. -30- Data on contraceptive use according to number of living children is shown in Table 5-4 and Figure 5-4. Use is relatively low among married women with no living children, only 10 percent, but increases rapidly after women have their first and second child. The use of female sterilization begins to become important after the second child (28 percent) and nearly doubles to a prevalence rate of around 50 percent among women with 3 or 4 children. In contrast, 31 percent of married women with one child, 26 percent with two children, and about 11 percent with 3 to 4 living children reported using reversible methods. In Table 5-5, we see that women with 10 or more years of formal education are twice as likely to use contraception as women with no formal education. This is due entirely to their greater use of reversible methods. For example, women with 10 or more years of education were four times more likely to use oral contraceptives than women with no formal education. It should be noted that the use of every reversible method except withdrawal and vaginal methods -31- is positively related to education. There is very little difference in the use of female sterilization among women with 1 to 9 years of schooling; however, this method is least used by women with no formal education and by those with 10 or more years of schooling. The lower use of female sterilization among women with 10 or more years of schooling is explained, in part, by their relatively young age; younger women tend to be better educated. Whether a woman is in the labor force is an important factor in the use of contraception, as shown in Table 5-6. Almost two-thirds of working women were found to be using contraception compared to 41 percent of nonworking women. The differential in use is predominantly accounted for by differences in the proportion of women in each group that are sterilized: Forty-two percent of working women reported tubal ligation as their method, compared to only 25 percent of nonworking women. Differences in contraceptive prevalence according to residence persist despite controlling for age, number of living children, and work status individually (Table 5-7). Women living in metropolitan San Salvador generally have the highest use levels and rural women the lowest. It is notable that women with no formal education in metropolitan San Salvador are similar to better educated women in rural areas with respect to use. In contrast, women with no living children in metropolitan San Salvador were 20 times more likely to report contraceptive use than rural women with no living children. Within residential categories, patterns of use according to age, number of living children, and work status are similar to those found for El Salvador as a whole. However, except in metropolitan San Salvador, differences in prevalence according to education almost disappear when residence is controlled, indicating that outside of the metropolitan area education itself is not an important predictor of contraceptive use. However, the positive association between contraceptive use and education is maintained when age, number of living children, and work status are controlled, as shown in Table 5-8. That is, women with higher levels of education are more likely to use contraception than women with lower -32- levels of education. Note that contraceptive use among women with one or no living children is much higher for women with 10 or more years of schooling, suggesting that the most educated women are more likely to recognize the importance of spacing their pregnancies than less educated women. The bottom panel of Table 5-8 and Figure 5-5 show that, with the exception of working women who did not attend school (44 percent), contraceptive use is high with about two-thirds of all other working women using contraception. Similarly, with the exception of women with 10 or more years of schooling, contraceptive use is significantly higher among working women, compared with women who are not working. -33- In Chapter III. of this report, planning status of last pregnancy, current pregnancy intentions, and desire to have more children were discussed. Tables 5-9 thru 5-11 summarize contraceptive use according to these variables. As Table 5-9 shows, there is little difference in use between women who stated that their last pregnancy was planned, mistimed, or unwanted. One would expect a higher level of use by women who stated that their last pregnancy was unwanted. However, nonusers of contraception following an unwanted birth tend to be older and of higher parity, which suggests ,that they may have always had difficulty in regulating their fertility. However, of those women who reported contraceptive use, a high proportion were using sterilization as their current method. Of the fecund, married women who stated that they did not currently desire a pregnancy, 58 percent were using contraceptive methods at the time of the survey (Table 5-10). As for all married women, female sterilization was the method of choice followed by oral contraceptives, condoms and the rhythm/Billings methods. More than three-fourths of such women living in metropolitan San Salvador were using a method compared to 68 percent of women living in other urban areas and only 43 percent of rural women. As Table 5-11 shows, 72 percent of women who stated that they had all of the children that they wanted were using some form of contraception. However, if current users of sterilization are not considered, the percentage using reversible methods is only 13 percent, ranging from 14 and 19 percent among the two urban strata to only 7 percent among rural women. The most used reversible method was oral contraceptives (5 percent) followed by condoms (2 percent) and the rhythm/Billings methods (2 percent). As is true in much of the developing world, contraceptive use is low among young, married women with few children (Morris, 1988). Table 5-12, which was generated by a life table procedure, shows that the median age and median number of living children of ever users of contraception (current users and nonusers who have used in the past) at time of first use of contraception is 28.3 years of age and 2.6 living children. Median age at first use varies by residence, -34- ranging from 25 years of age for women living in the metropolitan area to 32 years of age for women living in rural areas. Median age at first use was also found to be inversely associated with years of education; women with no formal education are about 34 years of age when they first initiate the use of contraception, compared to 26 years of age among women with 10 or more years of schooling. Median number of living children at time of first use varies substantially according to residence and education level of the respondent. Rural women have about three times as many children when they first use contraception than women living in metropolitan San Salvador and 1.8 times as many children as women who live in other urban areas. Similarly, women with no formal education reported having nearly five times the number of children at first use than women who had gone on to complete 10 or more years of schooling. The average age at marriage is 20.3 years in El Salvador, indicating that, on average, women are married eight years before they initiate the use of contraception. -v All current users of contraception were asked whether they were using in order to space or to limit their births. Nearly 72 percent responded that they were contracepting to limit their childbearing while the remaining 28 percent stated they wanted to space their pregnancies (Table 5-13). A higher proportion of rural women reported that they were contracepting to limit their childbearing (77 percent) than women living in other urban areas (70 percent) and women living in the metropolitan area (66 percent). This finding is consistent with the proportion found in each stratum that reported sterilization as their method (see the discussion for Table 5-2). As expected, the percentage of current users wanting to limit childbearing increased with age of the respondent and number of living children: Two-thirds of married users 25-29 years of age and 89 percent of users with 3 living children want to have no more children. It is noteworthy that many of the current users of less effective methods (withdrawal and natural methods) were using in order to limit childbearing. Finally, using the data presented in Table 5-13, we can make a profile of the women that are using contraception to space their pregnancies. As expected, these women tend to be young (15-24 years of age), have attained a higher level of education, and have fewer than two living children. Comparing the results of the 1988 survey with those of the 1975, 1978 (Morris et al., 1979), and 1985 surveys, we observe that contraceptive use among married women aged 15-44 increased by almost 26 percentage points between 1975 and 1988, an average of two percentage points per year (Table 5-14 and Figure 5-6). However, virtually all of the increase occurred by 1985. Seventy-eight percent of the increase in prevalence between the 1975 and 1988 surveys was in the use of female sterilization. Between 1975 and 1978, prevalence increased by 13 percentage points, or by 4.3 percentage points per year. Although use increased by another 12 percentage points between 1978 and 1985, the annual rate of increase was only 2.4 percentage points. However, comparing the results of the 1985 survey with those of the 1988 survey, we observe only a slight increase in use from 1985 to 1988, an increase which is within sampling error. Thus, between 1985 and 1988 there was no significant increase in contraceptive use in El Salvador. This -35- -36- finding is further illustrated in Table 5-15. Within each of the three residential strata, changes in contraceptive prevalence were extremely small and were not statistically significant. Finally, the data appearing in Table 5-16 and Figure 5-7 place contraceptive prevalence in El Salvador in a geographic perspective by comparing the 1988 El Salvador results with recent survey results from other countries in the region. Contraceptive prevalence in Costa Rica is almost 1.5 times the prevalence in El Salvador. Guatemala has the lowest prevalence of countries recently surveyed in the region, due primarily to its large indigenous population. With the exception of Costa Rica and Honduras, by far the most prevalent method in each of the countries is female sterilization. In Costa Rica and Honduras, oral contraceptives are the most used method. Compared to other countries in the region, El Salvador is second to Panama in prevalence of female sterilization, while it has the second lowest use rates of reversible methods; only Guatemala has lower use rates. -37- VI. SOURCE OF CONTRACEPTION This chapter discusses sources of contraception and time to reach those sources. In addition, source of contraception in 1988 is compared with that in 1985. As shown in Table 6-1 and Figure 6-1, the most important suppliers of contraception in El Salvador in 1988 were the three principal institutions of El Salvador's National Family Planning Program: the Ministry of Health (57 percent); the Asociación Demográfica Salvadoreña (ADS), the International Planned Parenthood Federation affiliate in El Salvador (13 percent); and the Social Security Institute (11 percent). In all three residential locations, the Ministry of Health (MOH) was the principal supplier of contraception. In San Salvador and other urban areas the Social Security Institute (ISSS) was the supplier for 23 and 10 percent, respectively, of currently married users of contraception. It is noteworthy that ADS was the second ranking supplier of contraceptives in rural areas, which is explained in part by its promotional campaigns and its community-based distribution program. Table 6-1 also shows that, in general, each provider of contraception in the country was serving about the same percentage of total users in 1988 as it was in 1985. However, a slight decrease is shown for both the MOH and ISSS and a slight increase is shown for pharmacies and private physicians. None of these changes are statistically significant. An analysis of source of contraception by method (Table 6-2) shows that the MOH was the principal source in both 1985 and 1988 for all methods shown (sterilization, orals, and the IUD) except condoms, which were obtained primarily in pharmacies in both years. However, between 1985 and 1988 the pharmacies' share of condom users decreased while that of the MOH increased, coincidentally the same amount. While pharmacies and ADS were the second and third principal suppliers of oral contraceptives in 1988, it should be noted that 42 percent of users who obtained their orals from pharmacies in that year identified their brand of orals as being Perla, which is distributed through ADS's social marketing program. In 1988, ADS's share of IUD users was twice that in 1985. As expected, the estimated time required to reach a source of contraception for women currently using contraception was less for women living in urban areas than for women living in rural areas (Table 6-3). Almost seven of every ten rural users live at least 30 minutes from their source of contraception. As shown in Table 6-4, average time-to-source for current users living in rural areas was more than twice that for users living in metropolitan San Salvador (74 minutes versus 33 minutes). In general, nonusers who know of a source of contraception live, on the average, 9 minutes closer to a source (44 minutes) than do users (53 minutes). However, the higher -38- -39- average time-to-source for all users is influenced by average time- to-source for users of sterilization (64 minutes). This finding suggests that nonpermanent methods are more accessible in El Salvador than sterilization services, which are primarily concentrated in the metropolitan area. -41- VII. REASONS FOR NONUSE OF CONTRACEPTION AND DESIRE TO USE IN THE FUTURE In this survey, nonusers were asked the reasons why they were not currently using contraception. This chapter covers the reasons given by respondents and the relationship between the characteristics of nonusers and desire to use contraception now or in the future. Also, for women who want to use a method, the preferred source and method are discussed. Women not currently using contraception represent most of the potential for expanding family planning program coverage in El Salvador. Of married women not currently using contraception at the time of the survey, 73 percent were not using any method for reasons related to pregnancy, subfecundity, or lack of sexual activity (Table 7-1). This includes 30 percent who were postpartum or breastfeeding, many of whom were probably at risk of pregnancy. Thus, 27 percent of nonusers could be considered potential candidates for adopting contraception at the time of the survey. This percentage represents an estimated 83,000 married women aged 15-44. However, it should be noted that a sizeable proportion of the 73 percent of nonusers who are not immediate candidates for use (nearly 70 percent, or approximately 160,000 women) are potential high priority target women following their current pregnancy and/or postpartum period. Among women who gave reasons unrelated to pregnancy, the major reasons for nonuse were fear of side effects of contraception (8 percent), spouse opposition to the use of contraception (5 percent), and religious reasons (3 percent). Examining reasons of nonuse by education, we observe in Table 7-2 that the proportion of women reporting a reason for nonuse related to pregnancy, subfecundity, or lack of sexual activity increases as the formal education of the women increases. Thus, the proportion of women who could be considered potential candidates for family planning at the time of this survey is inversely related to the level of education. Among lower educated women, the major reason for nonuse was fear of side effects. -42- Former users of contraception were asked why they discontinued use of their most recent method of contraception. Overall, 28 percent of the married nonusers aged 15-44 had ever used a contraceptive method. As Table 7-3 shows, the principal method used by these women was oral contraceptives (56 percent) followed by rhythm (11 percent), the IUD (9 percent), and injectables (9 percent). Almost 37 percent of the women stated that side effects from their method led them to discontinue use, while an additional 36 percent discontinued because they desired a pregnancy (Table 7-4). The proportion citing side effects was highest among former users of oral contraceptives and IUDs (about 48 percent). Almost 18 percent of former users of condoms stated that they discontinued the use of condoms because of spousal opposition. For women whose last method was withdrawal, vaginal or natural methods, nearly 13 percent terminated the use of these methods because they did not think they were effective. The data in Table 7-5 suggest that some women in El Salvador are misinformed about the safety of certain contraceptive methods as 56 percent and 48 percent of all nonusers, respectively, agreed with the statements read to them by the interviewer that the use of orals (or the IUD) is harmful to the woman. This finding suggests that additional information and education programs may be needed to dispel adverse rumors about selected methods of contraception. Table 7-6 shows that only half of fecund nonusers want to use a method now or in the future. In fact, only 6 percent of nonusers expressed an interest in using a method immediately. This proportion was highest among nonusers living in metropolitan San Salvador (69 percent) and lowest among rural nonusers (43 percent). It should be noted that interest in contracepting decreases as age and number of living children increase, and is directly associated with educational attainment. We may conclude, then, that many older, high parity women, especially those residing in rural areas, many of whom are at higher risk for maternal mortality, will continue to remain exposed to the risk of pregnancy. This has major implications for the need for educational programs for these women. -43- The right-hand panel of Table 7-6 shows that a high proportion of nonusers who desire to use a method have knowledge of a source of contraceptives (89 percent). The only groups for which knowledge of source was substantially below the overall total were teenagers and women without schooling. Table 7-7 shows the method of choice and the source mentioned by nonusers desiring to use a method. The most frequently desired methods were female sterilization (28 percent), oral contraceptives (21 percent), and injectables (19 percent). Injectables are not readily available in El Salvador. It should be noted that female sterilization was the method of choice among nonusers living in all three strata. Slightly more than 17 percent of nonusers did not cite a method of choice, but responded "Don't know". Twenty-two percent of rural women gave this response, indicating their uncertainty about the suitability of different methods for meeting their own personal circumstances. Ministry of Health facilities were cited as a potential source by more than two-thirds of nonusers, and was a commonly mentioned source among rural nonusers. ADS was cited as a potential source by 13 percent of nonusers and was the second most popular source in metropolitan San Salvador. It should be noted that pharmacies were mentioned as an important source of contraception by nonusers living in other urban areas. As expected, rural nonusers who desire to use contraception live further from a known source of contraceptives than their urban counterparts (Table 7-8). Interestingly, women living in other urban areas reported easier access to contraception than women living in metropolitan San Salvador. -45- VIII. STERILIZATION AND THE DEMAND FOR STERILIZATION As discussed earlier, female sterilization was the most prevalent method in El Salvador in 1988 with 30 percent of currently married women 15-44 reporting that they were surgically sterilized. (Only seventeen cases of vasectomy were reported. Thus, the following discussion applies to women with tubal ligations). Because the survey results show that the use of reversible methods was relatively low, but that a significant proportion of women wanted to limit childbearing, this chapter provides additional data on sterilization and the demand for this procedure. Specifically, this chapter discusses the characteristics of women with tubal ligations, timing of sterilization, post-operative regret, interest in and information concerning sterilization among women who want to limit their family size, reasons for lack of interest in sterilization, and reasons for failure to follow through among interested and informed women. Profile of Sterilized Women The top panel of Table 8-1 shows that a somewhat disproportionate number of sterilized women live in urban areas. Twenty-eight percent and 32 percent of all sterilized women in El Salvador live in metropolitan San Salvador and other urban areas, respectively, although these women represent only 23 percent and 29 percent of the total married survey population. In contrast, 40 percent of all married women that are sterilized live in rural areas, although rural women represent 47 percent of the survey population. The average age of sterilized women was about 27 years at the time they were sterilized and 33 years at the time of the survey. These averages compare with 29 years for all married women 15-44 in the survey. Similarly, 41 percent of sterilized women had four or more living children at the time of the survey compared with about 30 percent of the married survey population. Comparing mean number of living children, we see that sterilized women have more children (3.5) on average than all married respondents (2.9). These findings -46- suggest that sterilized women are self-selected because of their relatively high fertility that occurred early in their reproductive life—3.5 living children versus 2.9 for the general married population. With regard to education, sterilized women appear to be very similar to the currently married population in general. As shown in the bottom panel of Table 8-1, two-thirds of surgically sterilized, married women aged 15-44 have been sterilized since 1980 and over 90 percent have been sterilized in the past 12 years. Sterilization rates appear to have remained fairly constant in recent years. As shown in Table 8-2, of the women who have been sterilized, 59 percent had their tubal ligation during the month of their last delivery. We interpret this finding to mean that the majority were sterilized while hospitalized for delivery. This is confirmed by the data appearing in Table 8-3 where it is shown that slightly more than two-thirds of sterilized women who had their last live birth since 1983 were sterilized during delivery (45 percent) or during the week after delivery (23 percent), for a total of 68 percent at the time of their last delivery. Table 8-3 also shows that women living in urban areas were more likely to have a tubal ligation concurrent with their delivery than women living in rural El Salvador. The bottom panel of this table also shows that 38 percent of sterilized women with 10 or more years of formal education were sterilized following a cesarean birth, suggesting that the dual procedure might have been performed at the convenience of the couples. An analysis of data not shown reveals that, overall, tubal ligation was the first method to be used by 48 percent of sterilized women. For women living in the metropolitan area this proportion decreases to 25 percent, but increases to 46 percent for women living in other urban areas, and up to 66 percent among women living in rural El Salvador. This finding is consistent with the relatively low use of reversible methods discussed in Chapter 5. Thirty-six percent of sterilized women reported oral contraceptives as their first method, 6 percent the IUD, 4 percent condoms, and 2 percent injectables. The remainder reported using rhythm, withdrawal, and vaginal methods. -47- Satisfaction with the Decision All women who had a tubal ligation were asked if they were satisfied with having had the operation. Table 8-4 shows that 94 percent of married women 15-44 were satisfied with their decision to have the operation. The proportion satisfied with the procedure did not vary by residence or education, although levels of satisfaction were lower among women 25-29 years of age and among women with two living children. However, even among those groups, satisfaction was 90 percent or higher. Of the married women who expressed dissatisfaction with having had the operation, 49 percent reported that "their children had grown and they wanted to have a baby" (data not shown). Another 21 percent reported that they were "living in a new union", implying that they wanted a child with their new partner. An additional 13 percent cited side effects from the operation as their reason for being dissatisfied, 6 percent felt that they had "done something morally wrong", and 4 percent stated that a child had died. Overall, the level of satisfaction with having had the operation is high in El Salvador, with only 6 percent of married women 15-44 expressing regret. In 1984, in Panama, 7 percent of sterilized women expressed regret (Warren et al., 1987) while in 1987 nearly 13 percent of sterilized women living in Paraguay expressed regret for having had the operation (Monteith et al., 1988). Demand for Sterilization All fecund women who had at least one living child were asked if they wanted any more children (see Table 3-12). Overall, 34 percent of the women said that they did not want any more children, with the percentage highest in metropolitan San Salvador (42 percent) and lowest in other urban areas and in rural El Salvador (about 33 percent each). -48- As shown in Table 8-5, 63 percent of women cited "economic problems" as their main reason for not wanting any more children, while an additional 27 percent stated that they already had their "desired number of children". A higher percentage of women with 2 to 3 children cited economic problems as their reason for not wanting more children than women of higher parities. Despite their expressed desire for no more children, only 29 percent of these women reported that they were using contraception at the time of the survey (Table 8-6). Use varied by educational attainment with only 14 percent of women with no formal education reporting contraceptive use compared to 61 percent of women with 10 or more years of schooling. Oral contraceptives were the most prevalent method used among the users wanting no more children followed by the rhythm/Billings methods and condoms. All women who did not want any more children were then asked whether they were interested in having a tubal ligation. As shown in Table 8-7, 41 percent said that they were interested in sterilization, with the percentage highest among women living in other urban areas (58 percent) and lowest among rural women (34 percent). Older women, especially women 40-44 years of age, were less likely to be interested in sterilization than younger women. Interest in tubal ligation was directly associated with years of education and inversely associated with number of living children. It is noteworthy that about 58 percent of women with one or two children who did not want more children were interested in sterilization. There were no clear trends in interest in sterilization associated with work status and current contraceptive use. Of women who did not want any more children and were interested in sterilization, 94 percent had knowledge of availability of services or information concerning these services (data not shown). Knowledge of where to obtain sterilization information and/or services was slightly lower for women 15-24 years of age, women with no formal education, and women with 1-3 living children. -49- Overall, 60 percent of women cited the Ministry of Health as the institution where they would obtain sterilization information and/or services (Table 8-8). ADS and the Social Security Institute were important sources in the metropolitan area while the Ministry of Health was the dominant source in rural areas. All women who said that they did not want any more children, were interested in sterilization, and knew a source of sterilization services were asked why they had not sought surgical contraception. Interestingly, about 8 percent of the women stated that they had an appointment to be surgically sterilized (Table 8-9). However, twenty-one percent of these women said that their spouse's opposition to the operation was their main reason for never being sterilized. This was a particularly important reason for women living in other urban areas (27 percent). Fourteen percent of such women said they were waiting until after their next pregnancy to be sterilized. A possible interpretation of this response is that these women may believe that tubal ligation is only possible immediately postpartum, and that they are unaware of the availability of interval sterilization. An additional 14 percent of women stated that they did not have time to obtain the operation, which indicates that these women may be unaware of the availability of interval sterilizations on an out-patient basis, or that services are not readily accessible, especially for women in rural areas where 21 percent gave this reason. Eight percent and 7 percent, respectively, of the women stated that they feared the operation or its side effects, while an additional 7 percent stated that they needed more information about the operation. In sum, the reasons discussed above account for 72 percent of all the reasons given. Many of these "barriers" to sterilization could be reduced through a vigorous and effective educational program. Finally, women who did not want any more children and said they were not interested in surgical contraception were asked the reason for their lack of interest (Table 8-10). Overall, 25 percent of these women stated fear of the operation as their reason, while an additional 20 percent stated they feared side effects from the -50- operation. Almost 17 percent cited their spouse's opposition to the operation and 11 percent said they would get little benefit from being sterilized, i.e., approaching menopause. Religious reasons accounted for only 9 percent of the reasons given. -51- IX. CHARACTERISTICS OF WOMEN IN NEED OF FAMILY PLANNING SERVICES The survey data indicate that certain segments of the population have greater need of family planning services than others. A woman was characterized as "in need of services" (or "at risk of an unplanned pregnancy") if she was not currently pregnant, stated that she did not desire to become pregnant, and she was not using any method of contraception for reasons not related to pregnancy, subfecundity, or sexual inactivity. Thus, the Women defined here as "in need of services" are noncontracepting, fecund, sexually active women (regardless of marital status), who were not currently pregnant and did not desire to become pregnant at the time of the interview. The percentage of women with unmet need for contraception calculated according to this definition varied by the characteristics of the women, as shown in Table 9-1. Overall, 18 percent of women were found to be in need of family planning services or at risk of an unplanned pregnancy. This represents an estimated 194,000 women 15-44 years of age in El Salvador. The percentage in need of services is greater among rural women (28 percent) than among women living in other urban areas (12 percent) and women living in the metropolitan area (9 percent). The proportion in need of services is highest among married (in union) women, women with less education, and among non-working women. In fact, 30 percent of married women are at risk of an unplanned pregnancy (40 percent in rural areas. The risk of an unplanned pregnancy increases with number of living children. It is noteworthy that 30 percent of all women with one living child and 42 percent of rural women with one child are in need of family planning services. The percentages in Table 9-1 indicate the segments of the population in which the relative need for family planning services is greatest. In order to derive program goals, the women defined as being in need, i.e., the numerators in Table 9-1, have been distributed across the categories of women, as shown in Table 9-2. We observe in this table that 68 percent of women in need live in rural El Salvador and 90 percent are married. It is interesting to -52- note that 18 percent of women in need are 15-19 years of age, which may indicate that the development of services for young adults may be appropriate at this time. Sixty-one percent of women in need have three or fewer years of schooling and almost 84 percent are not in the labor force. In addition, about 53 percent have three or more living children. Thus, the data indicate that the family planning program of El Salvador should be oriented toward high parity, married, less-educated, non-working women and rural women. The data presented in Table 9-3 show that the women most in need of family planning services are those women who are least likely to have used contraception any time in the past, but more importantly, women who are least likely to have expressed a desire during the interview to use contraception now or in the future. Overall, only 23 percent of women in need of family planning services have ever used contraception, while only 45 percent desire to use contraception now or expect to sometime in the future. Thus, these women are generally inexperienced in the use of contraception and, even though they stated that they did not desire a pregnancy, are not motivated to seek the means to control their fertility. The implication of these findings for the national family planning program of El Salvador is obvious. -53- X. PREFERENCES IN FAMILY PLANNING SERVICES It is noteworthy that an overwhelming majority (86 percent) of women 15-44 years of age agreed with the following statement: "The government should offer family planning services" (Table 10-1). Women living in rural areas, women with less education, and nonusers of contraception were less likely to be in agreement with the statement, but were still overwhelmingly in favor of government provision of services. Accessibility to family planning services and how those services are provided are factors in whether current and potential users of contraception will continue or elect to utilize them. Thus, all current users of contraception, except those whose method is sterilization, and nonusers who desire to use contraception now or expect to in the future, were asked a series of questions on their preferences regarding family planning services. The remainder of this chapter discusses their responses. As shown in Table 10-2, any day of the week would be convenient to receive family planning services for 31 percent of the respondents. Overall, a weekend day was preferred by almost 28 percent of the women, but this proportion increased to 35 percent for women living in the metropolitan area and up to 41 percent among working women. In general, the public sector family planning program does not currently provide services on weekends. The preferences of nonusers did not differ significantly from those of current users of contraceptives. Almost 70 percent of the women stated that the most convenient time of the day to receive services would be between 8 and 10 in the morning and between 2 and 4 in the afternoon (Table 10-3). Preferences for these two time periods did not vary significantly by the variables shown, although rural women, nonusers, and non-working women tended to favor the morning hours. In general, services provided by the public sector program are not readily available after 1 PM. About three percent of the women preferred to receive services after 4 in the afternoon. -54- Anonymity appears to be a factor in selecting a family planning clinic for some women. As shown in Table 10-4, one-fourth of the women stated that they preferred to go to a far away clinic where they might not be known. This proportion increases to almost 30 percent for rural women, nonusers, and women aged 15-19 (Table 10-5). In contrast, over 50 percent of the women stated that they would go to a nearby clinic, although they might be known there. About 17 percent had no preference. Eighty-two percent of the women stated that, if they had a choice, they preferred to receive family planning services from a female physician, while only 5 percent preferred male physicians (Table 10-6). Preference for female physicians was greatest among women living in rural areas, among nonusers, and among women 15-19 years of age (Table 10-7). About 12 percent stated that they had no preference. Similar to their preference for female physicians, 75 percent of the women said that it would be acceptable to receive family planning services from a nurse practitioner rather than from a physician (Table 10-8). This proportion increases to 86 percent among rural women and to about 90 percent among women with 3 years or fewer of formal education. Sixty-four percent of women living in the metropolitan area and 59 percent of women with 10 or more years of schooling would accept services from a nurse practitioner. Finally, interest in receiving family planning services from community-based distributors was relatively low (Table 10-9). Thirty-eight percent of the women stated that they had interest in receiving services from trained, nonmedical personnel living in the community. Interest in community-based distribution programs was negatively associated with educational attainment. -55- XI. SEXUAL EXPERIENCE AND CONTRACEPTIVE USE: WOMEN 15-24 YEARS OF AGE Because there has been much recent interest in teenage fertility in Latin America, including early initiation of childbearing and unintended pregnancies, a special module for 15-24 year old respondents was added to the 1988 Family Health Survey. This module was designed to obtain information on attitudes, sexual experience, and use of contraception at first sexual experience. The module was administered to 1,518 women 15-24 years of age (Table 11-1). Thirty percent of the teenagers had been or were married, while 68 percent of the 20-24 year olds had been ever-married. Overall, about 54 percent of the women had less than 7 years of schooling and 60 percent had no living children. Twenty-two percent of the 15-19 year olds reported having one or more living children. According to the 1988 survey data, the fertility rate for young women 15-19 years of age in El Salvador was 138 births per 1,000 women (Table 11-2), the highest rate found for any country in the Latin America and Caribbean region (United Nations, 1987). The rate for women aged 20-24 of 246 births per 1,000 was not the highest in the region, but ranks among the top five rates. In Table 11-3, we estimate the percentage of first births that were premaritally conceived among women ever-in-union by comparing the date of first birth with the date of first union. Overall, by this estimation, 12.5 percent of ever-in-union women aged 15-24 had a premarital conception. In metropolitan San Salvador, 22 percent of women aged 15-24 reported a premarital conception, compared to 17 percent among women living in other urban areas and only 6 percent among rural women. The low premarital conception rate for rural women is due to the fact that the first sexual experience for these women often marks the beginning of a consensual union. The proportion reporting premarital conceptions increases with education, and is higher among women who married for the first time when they were in their twenties than among women who married as teenagers. -56- Overall, women aged 15-24 with at least one pregnancy reported that nearly 18 percent of their last pregnancies were unintended (Table 11-4). However, this figure doubles from 11 percent to 22 percent from first to second pregnancies. Only 16 percent of unmarried women with one pregnancy reported that their pregnancy was unintended, compared to about 9 percent of married women with one pregnancy, indicating that there is a strong acceptance of children early in marriage. Forty-nine percent of the respondents reported that they were sexually experienced (data not shown). A higher percentage of rural women (57 percent) than urban women (42 percent) reported being sexually experienced. As expected, the probability of being sexually experienced increased with age. Overall, 32 percent of women 15-19 years of age reported having had sexual experience compared to 70 percent of 20-24 year-old women. For both age groups, rural women were more likely to be sexually experienced than their urban counterparts. This is due to their earlier age at marriage and higher proportion that are married. Of the women reporting sexual experience, 53 percent were married and 47 percent were unmarried at time of first intercourse. Of all women 15-24 years of age who were interviewed, 23 percent reported that their first sexual experience was premarital (Table 11-5). The proportion reporting premarital sexual intercourse was higher in the metropolitan area of San Salvador (27 percent) than in other urban areas (22 percent) and in rural El Salvador (22 percent). Twice as many 20-24 year-olds (32 percent) reported premarital experience than 15-19 year-olds (16 percent). Overall, no clear pattern of premarital experience emerges according to years of education. However, the probability that the first sexual experience was premarital was inversely associated with education for women aged 15-19. Overall, age at first sexual intercourse did not vary significantly according to marital status at time of first sexual experience --about 71 percent of unmarried and married women were -57- sexually experienced before their 18th birthday (Tables 11-6 and 11-7). In addition, there was no significant variation in the proportion reporting sexual experience when age is controlled for by residence for those whose first relation was marital. However, a higher proportion of rural women with premarital experience (almost 30 percent) reported that their first intercourse was before their 15th birthday than rural women who reported that their first partner was their spouse (20 percent). Also, a significantly higher proportion of these rural women with premarital sexual experience had their first relation before 15 years of age compared with their urban counterparts. Women with no formal education who reported that their first sexual experience was premarital were more likely to be sexually experienced by age 18 (89 percent) than their counterparts who reported that their first experience was marital (80 percent). For both women with premarital and marital experience, the first sexual experience occurred earlier for those with less than 7 years of education than for women with more years of schooling. In general, average age of the women and of their partners at first sexual experience was 16 and 22 years of age, respectively. Overall, only 4 percent of women used contraception at the time of their first intercourse (Table 11-8). The use of contraception increased with age at first intercourse, from 1 percent for women who were less than 15 years of age at the time of their first sexual experience to almost 15 percent for women who were 20-24 years of age. For women whose first sexual experience was with their spouse, only 3 percent used contraception, compared to 5 percent by women whose first sexual experience was premarital. The latter figure is the lowest use rate at first premarital sexual experience reported in the Latin American region (Morris, 1987). The methods principally used by married women were oral contraceptives, withdrawal, condoms, and rhythm, which represented 91 percent of total use. For unmarried women, condoms, rhythm, withdrawal and injectables accounted for 88 percent of total use. Thus, the vast majority of young women at first intercourse were at risk of becoming pregnant because they did not use contraception or used less effective methods such as withdrawal and rhythm. -58- Knowledge of risk of pregnancy was also investigated when all women 15-24 were asked a question about when it is most probable for a woman to become pregnant during the menstrual cycle (Table 11-9). Only 15 percent had correct knowledge of the fertile period. Knowledge of the fertile period increased with education of the respondent, but peaked at only 26 percent for women with 10 or more years of schooling. Nearly half of nonusers of contraception at time of their first sexual experience who said that their first sexual partner was their spouse did not use a method because they wanted to become pregnant (Table 11-10). About 45 percent of women who were less than 18 years of age at time of their first intercourse gave this reason for their nonuse as did about 58 percent of women 18 to 24 years of age. Slightly more than 20 percent of the nonusers reported that they did not use a method because they didn't know about contraception at the time. This reason was particularly important for women who were teenagers when they had their first sexual experience and for women with no formal education (data not shown). An additional 17 percent reported that their reason for nonuse was that their partner had opposed the use of contraception. Forty percent of nonusers of contraception at first premarital intercourse reported that they didn't use because they were not expecting to have intercourse at that time (Table 11-11). Another 21 percent reported that their reason for nonuse was because they didn't know about contraception. "No knowledge of contraception" was an important reason for those less than 18 years of age and was inversely related to age. Overall, an additional 16 percent stated that they desired a pregnancy; interestingly, about 15 percent of women who were less than 18 years of age at their first premarital experience gave this reason. An additional 13 percent said their partner had opposed the use of contraception. Attitudes on when it is appropriate to initiate sexual relations varied by timing of first sexual experience (Table 11-12). Overall, 82 percent of the respondents were opposed to couples having premarital sex. Women with no sexual experience were more -59- conservative on this issue than women with sexual experience. The most liberal were those whose first sexual experience was premarital. On the other hand, 70 percent of those whose first sexual experience was premarital said that they did not agree with premarital sex, a classic difference between attitude and behavior. Ten percent of the women stated that it is all right for a couple to have premarital sex if the couple was planning to marry. Fifty-eight percent of all respondents 15-24 years of age who have had sexual experience were currently sexually active, reporting intercourse during the month prior to interview (Table 11-13). Slightly more than one-fourth of single women reported sexual activity. Of the women sexually active, 35 percent reported using contraception. Thus, the use of contraception increases considerably from time of first intercourse, premarital or marital, to current sexual activity. The majority of users used oral contraceptives (38 percent), female sterilization (20 percent) and condoms (14 percent) during the last month (data not shown). Finally, about 87 percent of women 15-24 years of age agreed with the statement that there should be special family planning services for young adults (Table 11-14). Interestingly, a greater proportion of older women than women 15-24 felt that these services should be provided. In addition, 90 percent of women 15-44 agreed with the statement that sex education should be taught in secondary schools, but only two-thirds agreed that it should be taught in primary schools (Table 11-15). The responses of 15-24 year-olds on the issue of sex education corresponded highly with the responses of older women. -61- XII. USE OF MATERNAL AND CHILD HEALTH SERVICES This chapter covers the use of maternal and child health (MCH) services by married women aged 15-44 who had a live birth within five years of interview. Factors influencing the use of prenatal, postpartum, and well-baby services are examined as well as the sources of these services with respect to various socioeconomic factors. In addition, location and type of last delivery (Cesarean versus vaginal) are examined. One section discusses the characteristics of women who reported a history of spontaneous and induced abortion, and whether they had any complications following their most recent abortion that required medical attention. Finally, the use of MCH services is assessed in terms of its association with family planning. Prenatal Services Married women aged 15-44 who had at least one live birth within five years of interview were asked if they had a prenatal examination during their most recent pregnancy. Slightly more than two-thirds of the women responded that they had received such an examination (Table 12-1). The percentage receiving prenatal care was higher for women living in metropolitan San Salvador (84 percent) than for women living in other urban areas (77 percent) and women living in rural El Salvador (57 percent). Table 12-1 also shows that the use of prenatal services is positively associated with educational attainment, ranging from a utilization rate of 50 percent among women with no formal education to 93 percent among women with 10 or more years of schooling. In addition, location of delivery is a correlate of the use of prenatal care. From data not shown, women who delivered in private or Social Security hospitals, indicating higher socioeconomic status, were more likely to have sought prenatal care than women who delivered elsewhere. Overall, three of every four women received their prenatal care from the Ministry of Health (Table 12-1). Almost 13 percent received their care from a private clinic, while and additional 11 percent -62- reported that the Social Security Institute was the source of their prenatal care. The vast majority of women living in rural areas (93 percent) and women with no formal education (94 percent) sought their prenatal care from the Ministry of Health. For women living in metropolitan San Salvador, the Social Security Institute and private physicians were important sources of prenatal care, accounting for 51 percent of the care. Timing of the first prenatal visit is important in that the earlier this visit is made the earlier problems can be detected and managed. Of the women receiving prenatal care, 70 percent reported that they received their first checkup during the first three months of pregnancy. Women living in the metropolitan area and in other urban areas were similar with regard to the timing of their first prenatal visit; they tended to receive their first checkup relatively early. Rural women, however, tended to have their first exam relatively late during pregnancy. Similarly, the likelihood of receiving an early, first checkup is directly associated with education. Women with 10 or more years of schooling were 1.5 times more likely to have made their first prenatal visit during the first trimester of pregnancy than women with no formal education. As the bottom panel of Table 12-1 shows, nearly two-thirds of the women who received prenatal care had five or more examinations during their pregnancy. Women who live in urban areas and women with more education were more likely to have made five or more prenatal visits than rural women and women with less education. It should be noted that the survey did not attempt to assess the quality of the care these women received. Thus, our analysis of the number of prenatal visits does not imply that women with more visits received better care than women with fewer visits. It can be stated that women who initiated their prenatal care early tended to make more visits than women who initiated their care later. This suggests that the probability of detecting problems among women in the former group, because of their increased "exposure" to the health care system, would be greater than among women who made later and, thus, fewer prenatal visits. -63- Vaccination Against Tetanus Women receiving tetanus injections during pregnancy protect their newborns from neonatal tetanus. Whether the respondent received a tetanus injection during her last pregnancy is examined in Table 12-2. Overall, 69 percent of the women received tetanus injections, with the percentage in urban areas higher than in rural areas. Of the women who received their prenatal care from a Ministry of Health facility, almost 92 percent reported that they received at least one injection, while 62 percent of women who sought prenatal care from a private physician reported receiving a tetanus injection. This may reflect a lack of attention to this preventive health care component by private physicians since only 45 percent of the women who did not receive a tetanus injection from their private physician reported having received an injection(s) in the 10 years prior to their pregnancy. Overall, 38 percent of women who were not vaccinated against tetanus during their last pregnancy had been vaccinated in the last 10 years. Thus, it can be assumed that 81 percent of married women aged 15-44 were protected against tetanus at the time of their last pregnancy. Table 12-3 shows that, of the women who received tetanus vaccine during their last pregnancy, 40 percent received one dose and 56 percent received two doses. The data reveal that 18 percent of women who received one dose were pregnant for the first time; for these women the World Health Organization recommends two doses (World Health Organization, 1988), if they had not already been immunized sometime prior to their pregnancy. Interestingly, almost 4 percent reported that they received 3 doses, one dose more than is recommended by the World Health Organization. Place and Type of Last Live Birth Table 12-4 shows place of last live birth for women who had a live birth within 5 years of interview. In general, 52 percent of the women reported that their last child was delivered in a medical facility (Ministry of Health, Social Security, or private hospital). This percentage is 88 percent for women living in -64- metropolitan San Salvador, but only 34 percent for rural women. In rural areas, midwives delivered 54 percent of the births, while an additional 13 percent were attended by apparently untrained people, principally family members. The high proportion of women living in rural areas who are delivered by a midwife indicates that midwives could become key agents in promoting family planning and other health services, especially among rural women. Women whose last delivery was in a hospital were 'asked if their most recent birth was a vaginal or a Cesarean delivery (Table 12-5). Of all last deliveries occurring in a hospital, 17 percent were Cesarean. The Cesarean rate for all births, regardless of place of delivery, was 10 percent. The percentage of women whose last hospital delivery was Cesarean was highest for urban women, women with 10 or more years of schooling, women aged 30 and over, and low parity women. In addition, a much higher proportion of deliveries in private hospitals than in Ministry of Health and Social Security hospitals were Cesarean. The data appearing in Table 12-5 also imply that women of higher socioeconomic status are more likely to have Cesarean deliveries than women of lower socioeconomic status because they pay for their care in whole or in part. History of Spontaneous and Induced Abortion Not every pregnancy terminates in a live birth. Thus, all respondents were asked whether they had ever had an abortion, either spontaneous or induced. If they had, they were then asked how many they had undergone and whether their last or only abortion was spontaneous or induced. As shown in Table 12-6, slightly more than 14 percent of all women reported they had had at least one spontaneous or induced abortion. Almost 16 percent of all women living in rural areas reported an abortion compared to about 13 percent of women living in urban areas. The proportion of women reporting an abortion increases with age and parity and is higher for ever-married women. However, note that 51 percent of women with no living children reported having had at least one spontaneous or induced abortion. -65- As in almost all retrospective surveys, women in this survey probably underreported their abortion experience. While slightly more than 14 percent of women reported having an abortion, spontaneous or induced, at some time in their lives, the number of abortions reported was only 5 percent of all reported pregnancies. This is relatively low compared to estimates of pregnancies terminating in spontaneous abortion alone, which have been found to range from 10 percent to 25 percent in prospective studies (Anderson, 1979). The percentage of pregnancies reported as ending in abortion was about 5 percent in each stratum. Therefore, place of residence does not make any difference with respect to underreporting spontaneous abortions. As shown in Table 12-7, almost 20 percent of women with a history of abortion reported two or more abortions. Only one percent said that their last abortion was induced. All women with a history of abortion were also asked if they had complications requiring medical attention following their most recent abortion. Almost 50 percent reported that they had received such attention, with 44 percent reporting that they were hospitalized (Table 12-8). Hospitalization does not vary significantly by residence, suggesting that urban and rural women have about equal access to hospitalization in the event that complications occur following an abortion. Overall, of the women who sought medical attention following their most recent abortion, Ministry of Health facilities were the primary source of treatment (Table 12-9). Postpartum Care Only 40 percent of women who had at least one live birth in the last 5 years reported receiving a postpartum checkup following their last birth (Table 12-10). As with other MCH services, this proportion was higher among urban women (50 percent to 56 percent) than among rural women (30 percent). In addition, the likelihood of receiving a postpartum checkup was positively associated with educational attainment. -66- Utilization of postpartum care was almost 28 percent lower than utilization of prenatal care. The relatively low utilization of postpartum care by women compared to their relatively high utilization of prenatal services suggests that they view the latter service as more important than the former service. This finding also suggests that postpartum services should not be the principal forum to promote family planning, especially in rural areas, because of the low utilization rates of this service. Overall, 69 percent of those women receiving postpartum care utilized a Ministry of Health facility. While the Ministry of Health was the principal provider of postpartum care in rural areas, the Social Security Institute and private physicians were the primary sources of this care in the metropolitan area. Private physicians were also an important source of postpartum care for women with 10 or more years of schooling. As shown in the bottom panel of Table 12-10, about 83 percent of those women receiving postpartum care received their first checkup during the first month after delivery or earlier. Timing of the first checkup did not vary much by residence or education, although women living in the metropolitan area and women with 10 or more years of schooling were more likely than their counterparts to receive their checkup during the first month after delivery. Since postpartum checkups are generally scheduled 4-6 weeks following delivery, women who reported receiving postpartum care less than 1 month after delivery may actually have received medical attention for complications. Well-Baby Care Women with at least one live birth within 5 years of interview were also asked if they had taken their last infant for a well-baby checkup. As shown in Table 12-11, 76 percent of infants received well-baby care, indicating that this service is more important to women than prenatal and postpartum care (Figure 12-1). However, the differential in the use of postpartum and well-baby care may be due to the fact that health providers in El Salvador do not always integrate these two services. The top panel of Table 12-11 also shows that urban women are more likely than rural women to utilize well-baby care. In addition, the probability of using this service is positively associated with educational attainment. Source of well-baby care generally corresponds to the table on source of prenatal care, although a higher percentage of women utilized Ministry of Health facilities for well-baby care than they did for prenatal care. Of the infants that received well-baby care, 71 percent received this care during the first two months of life, prior to the time of their first immunizations. Of the children who received well-baby care before they were one month old, 35 percent were reported to have been sick (data not shown). Utilization of All Three Services Table 12-12 presents the utilization of all three MCH service we have examined—prenatal, postpartum, and well-baby care. Only 34 percent of women who had at least one live birth within five years -67- -68- of interview said that they had used all three services in conjunction with their most recent birth. In contrast, 14 percent reported that they used no MCH services. A higher percentage of women living in the metropolitan area used all three services (52 percent) compared with women living in other urban areas (45 percent) and rural women (22 percent). Note that almost 20 percent of rural women reported that they did not utilize any MCH service at all. The right-hand panel of Table 12-12 also shows that women who last delivered in a private or Social Security hospital were from about 2 to 4 times more likely to use all three MCH services than women who delivered elsewhere. Since 1983, there has been a decline in the proportion of women utilizing all three MCH services (Table 12-13). Among women whose last live birth occurred in 1983, slightly more than 46 percent reported that they used all three services. In contrast, only one-third of women whose last live birth was in 1987 reported using these services. The decline is due to a decrease in the use of postpartum services and an increase in the proportion of women who utilize only prenatal and well-baby services. Use of MCH Services and Contraception Finally, women who received all three MCH services were more likely to report that they were currently contracepting than women who received only some or none of these services. As shown in Table 12-14, almost two-thirds of the women who received all three services were contracepting at the time of the survey compared to only 15 percent who used no maternal and child health service. With the exception of women receiving well-baby care only, female sterilization was the most commonly used method among these women (data not shown). -69- We cannot say that the use of MCH services influences the use of contraception, or vice versa. However, the fact that contraceptive use among nulliparous women is very low in El Salvador and increases with parity, suggests that women's first exposure to family planning may indeed be in an MCH setting, which can be effectively used to promote the use of contraception. On the other hand, increase in parity may have simply been the key factor in the decision to use contraception. -71- XIII. IMMUNIZATION LEVELS We also evaluated the immunization coverage of children less than 5 years of age. Data were recorded for only those children born to the women interviewed, although other children less than 5 years of age might have been living in the household. This step was taken as a measure to increase the accuracy and reliability of the immunization data. Questions were asked on the number of doses of vaccine received against tuberculosis (BCG), poliomyelitis, diphtheria-tetanus-pertussis (DPT), and measles for each child. In addition, questions were asked about whether the vaccinated children had a vaccination certificate and where the vaccinations were obtained. The World Health Organization (WHO) recommends that primary immunizations should be completed before the first birthday (OPS, 1983). The number of doses considered primary immunization recommended by WHO and adopted by the El Salvador Ministry of Health (MSPAS, 1988) are: Three doses each of Polio and DPT vaccines and one dose each of measles and BCG vaccine. Table 13-1 shows that levels of coverage range from 61 percent to 81 percent for all four diseases. Levels are similar for polio and DPT, that is, 61 percent of children less than 5 years of age have complete immunization. Levels of complete immunization are higher for BCG (77 percent) and measles (81 percent), which may be due to the one-dose regimen required for these vaccines. Table 13-1 also shows that immunization levels vary by place of residence. While children living in urban areas tend to have higher levels of complete BCG and measles immunization than rural children, levels of complete polio and DPT immunization do not vary significantly by residence. Only one vaccine had a higher level of complete immunization in 1988 than in 1985-- BCG (Table 13-2 and Figure 13-1). Both polio and DPT decreased by 10 percentage points between the surveys, while measles decreased by about 3 percentage points. Most of the decrease in levels of primary immunization against polio and DPT occurred in the metropolitan area of San Salvador. Table 13-3 suggests that the decreases for polio and DPT, as confirmed by vaccination certificates, were real and not due to response error. Tables 13-4 thru 13-7 show that the majority of complete immunizations occur by one year of age. By this time, 60 percent of the children have received primary immunization against polio and DPT, and 80 percent are immunized against tuberculosis and measles. Although some increase is seen after age one, the percentage with complete immunizations levels off. These tables also show that, in general, urban children tend to be immunized sooner than rural children. Tables 13-8 thru 13-11 and Figure 13-2 expand on these results. In general, the picture that is drawn by these tables for all vaccines except BCG is a dramatic increase in complete -72- -73- immunization levels from less than one year of age to one year of age, followed by a lesser increase from age one to age two and an even slighter increase thereafter. For BCG, about 79 percent of the children are immunized before their first birthday; thereafter, increases in coverage is almost nil. Finally, Table 13-12 shows that the Ministry of Health is the principal source of vaccinations in El Salvador. In addition to vaccinating children in its health facilities, the Ministry of Health also conducts vaccination campaigns on a periodic basis. Although coverage of these campaigns appears to be relatively low, we expect that some women who reported that their children were immunized by the Ministry of Health were actually immunized during one or more of the recent mass campaigns. -75- XIV. PREVALENCE OF DIARRHEA AND ITS TREATMENT Acute diarrhea is a clinical symptom caused by viral, bacterial, or parasitic infections as well as by chemical agents. It is characterized by loose or watery stools and sometimes vomiting and fever. The major mode of transmission is fecal contamination of water, food, or fomites. It can be spread by person to person contact, and is usually associated with inadequate food handling, inadequate personal hygiene, and poor environmental sanitation. For the purpose of this survey, diarrhea was defined as two or more loose or watery stools over a period of at least 24 hours. Therefore, it is possible that some chronic cases of diarrhea were included, thus, increasing the prevalence of diarrhea reported here. Each respondent was asked whether each of her children less than 5 years of age had had diarrhea during the two weeks prior to interview. If a child had had diarrhea, the respondent was then asked if the stool contained blood and/or mucous (a measure of severity), and what treatment the child received for that episode of diarrhea. Table 14-1 shows that 29 percent of children less than 5 years of age were reported to have had diarrhea during the two weeks prior to interview. Nineteen percent of children living in the metropolitan area recently had diarrhea compared to almost 27 percent of children living in other urban areas and 33 percent of rural children (Figure 14-1). The table also shows that children less than 2 years of age were more likely to have had diarrhea than older children. There was an association between the prevalence of diarrhea and various household characteristics, which are indicators of the sanitary conditions of the child's environment (Table 14-2). Children living in one room households, whose sanitary services were either a common latrine or none, whose kitchens did not have a refrigerator, or where food was cooked using wood or charcoal, were more likely to have had diarrhea than children living in better conditions. Although outside sources of drinking water appear to be associated with higher prevalence of diarrhea, the data were inconsistent with respect to source of drinking water. From data not shown, the probability of having had diarrhea was negatively associated with the mother's education. The severity of cases can be assessed by asking respondents if there was blood and/or mucous in the stool of those that had diarrhea. Of the children who reportedly had diarrhea in the two weeks prior to interview, nearly 48 percent were reported to have had blood and/or mucous in their stool (Table 14-3). The presence of blood and/or mucous in the stool suggests that the recent episode of diarrhea was bacterial in origin and, therefore, not self-limiting. Children living outside of the metropolitan area and children three years of age or younger were more likely to have had blood and/or mucous in their stool than other children. In addition, the presence of blood and/or mucous in stools was negatively associated with the mother's education. -76- -77- For each child who reportedly had diarrhea in the two weeks prior to interview, respondents were asked what was done about the illness. The object was to see to what extent effective treatment was being used and, conversely, what was the extent of inappropriate treatment. Overall, 87 percent of children who had had diarrhea were reported to have been treated for their recent episode of diarrhea (data not shown). The likelihood of being treated did not vary substantially by demographic or socioeconomic variables. As Table 14-4 and Figure 14-2 show, use of oral rehydration solutions (ORS), either prepared in the home from readily available ingredients or from packets, is minimal in El Salvador: only 13 percent of the respondents said that their children were treated with oral rehydration solutions made from packets (11 percent) or from ingredients found in the home (2 percent). ORS is an effective, low-cost, easy to prepare and administer means of preventing and treating dehydration due to diarrhea (Parker et al., 1980). It should be noted that a higher proportion of younger children were treated with oral rehydration solutions than older children. However, the presence of blood and/or mucous in the stool did not make any difference in whether oral rehydration solutions were used or not. Nearly 81 percent of the children were treated with various popular pharmaceutical products and another 4 percent were given hot or cold food, herbal preparations, or had food withheld. It is likely that the majority of these treatments were either ineffective or deleterious. For example, antibiotics are ineffective against viruses, the most common cause of diarrhea in El Salvador. Preparations containing anticholinergics or kaolin may provide minor symptomatic relief, but do nothing to treat dehydration, the most important complication of diarrheal diseases in El Salvador. As mentioned above, only 13 percent of children with a recent episode of diarrhea received oral therapy to treat or prevent dehydration. Differences in treatment between strata were relatively minor, except that rural women were more likely to use traditional treatments than urban women. In addition, mothers of children less than two years of age were more likely to use oral rehydration solutions than mothers of older children. Finally, of the mothers who said that their children were not treated for their recent episode of diarrhea, 61 percent said that they did not treat their children because they did not think that it was necessary, 17 percent said that they lacked the time to seek treatment, and an additional 16 percent did not treat their children because they lacked the financial means to do so (data not shown). -78- -79- XV. INFANT AHD CHILD MORTALITY Each respondent was asked whether each of her children born alive since the beginning of 1983 was still living. For each child no longer alive, age at death was obtained. Based on this information and the current ages of children still alive, life tables were constructed to estimate mortality rates for El Salvadoran children. The life tables make use of reported survival information for all children born in the previous five years to determine the probability of surviving to any age up to five years for children born during that period. Table 15-1 shows the probability of dying before reaching a given age. A child born in El Salvador in the five years before the survey had a .055 probability of dying before reaching his or her first birthday. That is, about 55 of every 1,000 children born alive died during infancy. The probability of dying between birth and the fifth birthday was .075. Mortality is clearly higher in rural areas than in urban areas. In rural areas the probabilities of dying are .061 in infancy and .084 before the fifth birthday. Mortality is also higher among children living in other urban areas compared to children living in the San Salvador metropolitan area. The probability of dying before the fifth birthday in the San Salvador metropolitan area may be underestimated since no deaths were reported there for children 1 to 4 years of age. Estimated mortality falls sharply as the education of the child's mother increases; the chance of dying is about twice as high for children of mothers with no formal education than it is for children with well educated mothers (Figure 15-1). In addition, mortality of male children only slightly exceeds that of females. The probability of a male infant dying is .057 compared with .053 for females. Further analysis of the data will be done at a later date to look into questions concerning the effect of behaviors and programs on infant and child mortality. For instance, it will be important to see how the use of prenatal services and breastfeeding affect the likelihood of a child surviving to a given age. Table 15-2 shows infant mortality probabilities according to the 1985 (ADS y IRD, 1987) and 1988 surveys. These data suggest that a rapid decline in infant mortality has occurred over a short period of time (Figure 15-2). Overall, the probability of dying in infancy fell from .065 to .055. The top panel of Table 15-2 looks at mortality according to time since birth. These figures give considerable credibility to the decline in mortality. First, in each survey, infant mortality is higher the earlier the births occurred. Second, and more importantly, the 1985 and 1988 surveys show about the same mortality levels for births occurring from mid-1983 to mid-1985. The 1985 survey found that the probability of dying for infants born during this period was .056. Correspondingly, the 1988 survey shows almost an identical probability for infants born during the same period ---.059 (data not shown). The significance of this correspondence is that both -80- -81- -82- surveys seem to have collected equally complete mortality data, thus supporting the conclusion that the decline in infant mortality is real and not a function of differential completeness in the reporting of deaths. It should be noted that most of the decline in mortality occurred in rural areas, where it fell from .081 to .061. For other urban areas, mortality declined from .056 to .051, while in the San Salvador metropolitan area, the decline was from .046 to .041. -83- REFERENCES Anderson, John E. 1979. Measurement of Abortion in World Fertility Surveys and Contraceptive Prevalence Surveys. Working Paper. Division of Reproductive Health, Centers for Disease Control, Atlanta, Georgia. Anderson, John E., Rodrigues, Walter, and Thome, Antonio Marcio Tavares. 1983. Analysis of Breastfeeding in Northeastern Brazil: Methodological and Policy Considerations. Studies in Family Planning, Vol. 14, No. 8/9 (August/September): 210-218. Asociación Demográfica Salvadoreña (ADS) y Institute for Resource Development/Westinghouse (IRD). 1987. Encuesta Nacional de Salud Familiar (FESAL-85). San Salvador, El Salvador. Asociación Demográfica Salvadoreña (ADS) y Centros de Control de Enfermedades (CDC). 1988. Encuesta Nacional de Salud Familiar (FESAL-88): Informe Preliminar. San Salvador El Salvador. McCann, M.F., Liskin, L.S., Piotrow, P.T., Rinehart, W., and Fox, G. 1981. Breastfeeding, Fertility and Family Planning. Population Reports, Series J., No. 24 (November/December): 525-575. Ministerio de Salud Publica y Asistencia Social (MSPAS), Departamento de Nutrición. 1986. Normas de Atención Nutricional. San Salvador, El Salvador. Ministerio de Salud Publica y Asistencia Social (MSPAS), Unidad de Epidemiología. 1988. Programa Nacional de Vacunación, Ano 1988. San Salvador, El Salvador. Monteith, Richard S., Carrón, Juan Maria, Warren, Charles W., Melian, Maria Mercedes, Castagnino, Dario, and Morris, Leo. 1988. Contraceptive Use and Fertility in Paraguay, 1987. Studies in Family Planning, Vol. 19, No. 5 (September/October): 284-291. Morris, Leo, Rugamas, Ricardo Castañeda, Mendoza, Angela M. de, and Taylor, Sam. 1979. Contraceptive Use and Demographic Trends in El Salvador. Studies in Family Planning, Vol. 10, No. 2 (February): 43-52. Morris, Leo. 1987. Sexual Experience and Use of Contraception Among Young Adults in Latin America. Presented at the Annual Meeting of the American Public Health Association, New Orleans, Louisiana, November, 1987. Morris, Leo. 1988. Contraceptive Use and Reported Levels of Unplanned Pregnancies in Latin America. Presented at the Latin American Studies Association, XIV International Congress, New Orleans, Louisiana, March 17-19, 1988. Organización Panamericana de la Salud (OPS), Organización Mundial de la Salud. 1983. Avances Recientes en Inmunización. Publicación Científica, No. 451. Washington, D.C. -84- Parker, Robert L., Rinehart, Ward, Piotrow, Phyllis T., and Doucette, Louise. 1980. Oral Rehydration Terapy (ORT) for Childhood Diarrhea. Population Reports, Series L, No.2 (November/December): L-43. Population Reference Bureau, Inc. 1988 World Population Data Sheet. Washington, D.C., April, 1988. United Nations. 1987. Fertility Behaviour in the Context of Development: Evidence from the World Fertility Survey. Department of International Economic and Social Affairs. Population Studies, No. 100: 24. New York. Warren, Charles W., Oberle, Mark W., Medica, Wilma, and Morris, Leo. 1987. Uso de Anticonceptivos y Fecundidad en Panama. Perspectivas Internacionales de Planificación Familiar, Numero Especial de 1987: 6-11. World Health Organization Expanded Programme on Immunization. 1988. Neonatal Tetanus: Update. World Health Organization, Geneva, Switzerland, September, 1988. Cover Table of Contents CONTRIBUTORS TO THE SURVEY SUMMARY Introduction Fertility Breastfeeding Contraceptive Use Source of Contraception Reasons for Nonuse of Contraception and Desire to Use in the Future Women in Need of Family Planning Services Preferences in Family Planning Services Young Adult Sexual Behavior Use of Maternal and Child Health Services Immunization Coverage Prevalence of Diarrhea and Its Treatment Infant and Child Mortality I. INTRODUCTION II. SURVEY METHODOLOGY III. FERTILITY Fertility Rates Breastfeeding Nuptiality Contraceptive Use Planning Status of Last Pregnancy and Current Pregnancy Intention IV. KNOWLEDGE OF CONTRACEPTIVE METHODS V. CURRENT USE OF CONTRACEPTION VI. SOURCE OF CONTRACEPTION VII. REASONS FOR NONUSE OF CONTRACEPTION AND DESIRE TO USE INTHE FUTURE VIII. STERILIZATION AND THE DEMAND FOR STERILIZATION Profile of Sterilized Women Satisfaction with the Decision Demand for Sterilization IX. CHARACTERISTICS OF WOMEN IN NEED OF FAMILY PLANNING SERVICES X. PREFERENCES IN FAMILY PLANNING SERVICES XI. SEXUAL EXPERIENCE AND CONTRACEPTIVE USE: WOMEN 15-24 YEARS OF AGE XII. USE OF MATERNAL AND CHILD HEALTH SERVICES Prenatal Services Vaccination Against Tetanus Place and Type of Last Live Birth History of Spontaneous and Induced Abortion Postpartum Care Well-Baby Care Utilization of All Three Services Use of MCH Services and Contraception XIII. IMMUNIZATION LEVELS XIV. PREVALENCE OF DIARRHEA AND ITS TREATMENT XV. INFANT AHD CHILD MORTALITY REFERENCES Tables

View the publication