Click for next page ( 20


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 19
Levels and Trends in Contraceptive Use This chapter describes the extent to which contraception is being prac- ticed in sub-Saharan Africa. We first review sources of data on and defini- tions of contraceptive use. We then discuss what these data show regarding levels and trends in prevalence and differentials in current use by women. Next, we look at indicators related to contraceptive use: ever use, discontinuation, contraceptive knowledge, knowledge of sources, of supply and fertility pref- erences. Finally, we briefly present evidence of use from the perspective of men. Information on contraceptive use is far from complete, given that contraceptive prevalence data based on nationwide surveys are not available for all African countries.) SOURCES OF DATA ON CONTRACEPTIVE USE There are three main sources of data on contraceptive use in Africa based on representative samples at the national level: the World Fertility Surveys (WFS), the Contraceptive Prevalence Surveys (CPS), and the De- mographic and Health Surveys (DHS). Studies to date in Africa for which However, most countries that have initiated active family planning programs have also conducted surveys; thus, if data from the existing surveys were used to obtain a regional average, it would tend to overstate the actual use of modern contraception in Africa. 19

OCR for page 19
20 FACTORS AFFECTING CONTRACEPTIVE USE data are available include nine WFS (from 1977-1982), seven CPS (1982- 1984), and 12 DHS (1986-1990) (Kendall, 1979; Rutstein et al., 19921. Six populations were included in both the WFS and the DHS programs: Cameroon, Ghana, Kenya, Nigeria, Northern Sudan, and Senegal.2 For Benin, Cole d'Ivoire, Lesotho, and Mauritania, data are available from the WFS only. Populations with only DHS data include Botswana, Burundi, Liberia, Mali, Togo, Uganda, Zimbabwe, and the state of Ondo in Nigeria. Surveys using the CPS questionnaires have been conducted in selected sites in Zaire (four urban and two rural areas in 1982-1984) and in selected regions of Somalia (1983), but they do not provide national-level estimates and thus are omitted from most of the tables in this chapter. There are also numerous small-scale studies that have been conducted in the capital cities andlor selected regions of countries. (For a review of studies of fertility and knowledge, attitudes, and practice of contraception for Africa from 1960 to 1973, see Baum et al., 1974. For the period after 1983, see Gelbard et al., 1988.) However, no data from these have been included in the tables below, because it is doubtful that the results are representative or that the data collection techniques are comparable. Details on the year of data collection and of data included in this chap- ter are listed in Table 2-1. As the table indicates, despite the three major data collection efforts, for large parts of sub-Saharan Africa neither fertility (except from occasional censuses) nor contraceptive use data are available. This situation is most notable in Central Africa; with the exception of the Cameroon study and the 1982-1984 contraceptive prevalence survey in Zaire (subnational in scope), there is very little demographic information for this region of Africa. DEFINITION OF CONTRACEPTIVE USE In the DHS studies conducted in sub-Saharan countries between 1986 and 1990, at least half the current users of contraception in 5 of the 12 countries were relying on traditional methods, such as withdrawal and rhythm. There is little literature on the effectiveness of such methods, but they are considered much less effective than modern methods; thus the convention has emerged of differentiating between modern and traditional (or modern versus all methods) in reporting contraceptive prevalence results, especially in the context of Africa. In the three main types of studies conducted to date (WFS, CPS, DHS), methods defined as "modern" include oral contraceptives, intrauterine de 2Data for Cameroon were not available during the writing of this report. Some data were available for Nigeria (from the DHS country report) and are included in relevant tables.

OCR for page 19
LEVELS AND TRENDS IN CONTRA CEPTI VE USE TABLE 2-1 Data Sources on Contraceptive Use in Sub-Saharan Africa 21 Survey Program and Date Country wFsa cpSb DHsc Others Western Africa Benin 1981- 1982 Cote d'Ivoire 1980-1981 Ghana 1979- 1980 1988 Liberia 1986 Mali 1987 Mauritania 1981 Nigeria (Ondo State) 1986- 1987 Nigeria 1981- 1982 1990 Senegal 1978 1986 Togo 1986 Eastern Africa Burundi 1987 Ethiopia 1990 Kenya 1977- 1987 1984 1988- 1989 Malawi 1984 Rwanda 1983 Sudan (northern) 1978-1979 1989-1990 Somaliad 1983 Uganda 1988-1989 Zimbabwe 1984 1988-1989 Central Africa Cameroone 1978 1991 Zaired 1982- 1984 Southern Africa Botswana 1984 1988 Lesotho 1977 South Africa 1975- 1976 1981 Swazilandf 1988 aWorld Fertility Survey. bContraceptive Prevalence Survey. CDemographic and Health Survey. dThe data for the CPS in Zaire and in Somalia are not national. In Somalia, the survey was conducted for urban areas. In Zaire, the survey covered four cities and two rural areas. eThe DHS data for Cameroon are not available as of the writing of this report. fThe data for Swaziland are not presented in the tables. As noted in this chapter, Swaziland is not discussed in this report because of its small population size, estimated at under one million people.

OCR for page 19
22 FACTORS AFFECTING CONTMCEPTIVE USE vices (IUDs), female and male sterilizations, injections, condoms, diaphragms, and spermicides. The standard questionnaires explicitly mention two types of traditional methods: periodic abstinence (also called rhythm or the cal- endar method) and withdrawal. Any other method that the respondent con- siders to be a means of preventing pregnancy is recorded as "other." Other traditional methods include douche, herbs, and "gns-gris" (amulets, charms, or spells to prevent conception). Many of the country-specific question- na~res list the traditional methods common to their populations. As discussed in Chapter 1, two postpartum practices observed in much of sub-Saharan Africa are sexual abstinence and breastfeeding. Although these practices may result in lower fertility, in Africa they are generally motivated by the desire to ensure the survival of the youngest child and allow the mother to recuperate between births. Accordingly, neither is considered a contraceptive method in this report (see Appendix A for a discussion of the rationale). However, a high proportion of women in the African WFS and DHS did respond that they were using abstinence as a contraceptive method, although many were in fact practicing postpartum abstinence. For example, much of Togo's high level of traditional contra- ceptive use is due to a large proportion of women who reported using abstinence as a method (72.5 percent of these women were also in postpar- tum abstinence). In addition, more than half of the DHS did not include abstinence as a possible contraceptive, so we could not consider the use of abstinence as a method for all countnes.3 Although postpartum abstinence and abstinence are not considered contraceptive methods in this report, pe- nodic abstinence (i.e., rhythm) is. In sum, throughout this volume, the proportion currently using contra- ceptives refers to the proportion of women currently married or in union, aged 15-49 years, who are using a contraceptive method at the time of the survey (unless otherwise stated).4 The data are based on all women fulfill- ing these criteria, regardless of whether or not they are breastfeeding or postpartum abstinent. 3For those surveys that did include abstinence as a method, the proportion of women who reported practicing abstinence are given in a note to Table 2-2. 4There is evidence that contraceptive use is rising among women who are sexually active and unmarried. This topic is addressed for adolescents in a report by the Working Group on the Social Dynamics of Adolescent Fertility (1993) of the Panel on the Population Dynamics of Sub-Sarahan Africa.

OCR for page 19
LEVELS AND TRENDS IN CONTRACEPTIVE USE PREVALENCE OF CURRENT CONTRACEPTIVE USE IN SELECTED COUNTRIES Results from Demographic and Health Surveys Conducted from 1986 to 1990 23 Table 2-2 presents contraceptive prevalence rates (CPRs) for both cur- rent use and ever use of all methods combined,5 of modern and traditional methods separately, and of specific methods listed in the 12 DHS studies conducted in African countries between 1986 and 1990 for which data are available.6 Of the 12 countries, only two show a modern contraceptive prevalence rate of more than 30 percent: Zimbabwe (36 percent modern, 43 percent all methods combined) and Botswana (32 percent modern, 33 per- cent all methods combined). Kenya shows the next highest rate with 18 percent modern use (27 percent all methods combined). No other DHS country in Africa has even 6 percent of women of reproductive age using a modern contraceptive, although in Ghana and Togo the percentage currently using any method is greater than 10. The U.S. Agency for International Development denotes populations with 0 to 7 percent modern usage as having emergent family planning programs, and clearly the majority of the countries in Table 2-2 fall into that category (Destler et al., 1990~. Data from these 12 countries also reflect regional differences. The three higher use countries (Zimbabwe, Botswana, and Kenya) are all anglophone countries in eastern or southern Africa. In the four francophone countries surveyed, three of which are in western Africa (Mali, Senegal, and Togo), less than 3 percent of the women use modern contraceptive methods. These differences may reflect a growing divergence between southern and East Africa and West Africa. For example, data from the Swaziland Family Health Survey show prevalence rates for modern methods of almost 14 percent, suggesting that this nation may be following the path of the three high-use countries (Warren et al., 1992~.7 However, there are still too few high-prevalence countries in sub-Saharan Africa to be able to attribute dif- ferentials to geographical factors or different colonial histories. Moreover, these geographical or historical distinctions may be proxies for other deter- minants of demand for children and use of contraception discussed in Chap- ters 3 to 5, including socioeconomic factors, infant and child mortality rates, levels of female education, and the strength of family planning pro grams. SThe ever-use rates are discussed later in this chapter. 6Data for Ondo State are excluded when later data for Nigeria as a whole are available. 7Swaziland is not discussed in this report because of its small population size, estimated at under one million people.

OCR for page 19
24 FA CTORS AFFECTING CONTRACEPTIVE USE TABLE 2-2 Women in Union Age 15-49 Who Currently Use or Have Ever Used a Contraceptive Method, DHS, 1986-1990 (percent) Modern Methods Any Any Vaginal Method Modern Pill IUD Injection Method Condom Current Use Botswanaa 32.5 31.7 14.8 5.6 5.4 0.0 1.3 Burundia 6.7 1.2 0.2 0.3 0.5 0.0 0.1 Ghana 12.9 5.2 1.8 0.5 0.3 0.3 0.3 Kenya 26.9 17.9 5.2 3.7 3.3 0.4 0.5 Liberia 6.4 5.5 3.3 0.6 0.3 0.2 0.0 Malia 3.2 1.3 0.9 0.1 0.1 0.1 0.0 Nigeria 6.0 3.5 1.2 0.8 0.7 0.1 0.4 Senegala 4.6 2.4 1.2 0.7 0.1 0.1 0.1 Sudan (northern) 8.6 5.5 3.9 0.7 0.1 0.0 0.1 TOgoa 12.1 3.1 0.4 0.8 0.2 0.6 0.4 Uganda 4.9 2.5 1.1 0.2 0.4 0.0 0.0 Zimbabwe 43.1 36.1 31.0 1.1 0.3 0.0 1.2 Ever Use Botswanaa 61.7 60.1 49.4 17.4 15.0 1.3 10.1 Burundia 23.8 2.3 0.8 0.6 0.9 0.1 0.3 Ghana 37.0 22.5 14.9 1.3 10 8.3 4.5 Kenya 45.0 29.0 18.0 8.4 6.7 2.1 4.3 Liberia 18.8 15.9 13.5 2.8 1.7 1.0 1.7 Malia 7.6 3.4 2.7 0.5 0.2 0.4 0.4 Nigeria 14.0 8.4 4.8 1.7 1.9 0.6 2.0 Senegala 11.9 5.7 3.5 1.5 0.4 0.8 1.4 Sudan (northern) 25.2 18.6 17.3 1.6 1.1 0.3 1.9 TOgoa 33 0 10.1 3.8 1.6 1.1 3.7 3.3 Uganda 21.5 7.0 5.0 0.5 1.3 0.2 0.7 Zimbabwe 79.0 63.0 57.1 3.4 14.5 0.3 17.0 aAlthough abstinence is included as a traditional contraceptive method in these country surveys, it is not included as a method in this report (see this chapter for rationale). For the information of the reader, rates of abstinence for these countries follow. Current use: Botswana, 0.5; Burundi, 2.0; Mali, 1.5; Senegal, 6.7; and Togo, 21.8. Ever use: Botswana, 7.9; Burundi, 14.6; Mali, 12.7; Senegal, 23.5; and Togo, 59.0. SOURCE: Rutenberg et al. (1991) and different national DIES survey reports and data tapes.

OCR for page 19
LEVELS AND TRENDS IN CONTRACEPTIVE USE Traditional Methods Sterilization Any Periodic Other Female Male Traditional Abstinence Withdrawal Method 25 4.3 0.3 0.80.2 0.3 0.3 0.1 0.0 5.54.8 0.7 0.0 1.0 0.0 7.76.2 0.9 1.6 4.7 0.0 9.07.5 0.2 1.3 1.1 0.0 0.90.6 0.1 0.2 0.1 0.0 1.91.3 0.1 0.5 0.3 0.0 2.51.4 0.5 0.6 0.2 0.0 2.20.9 0.1 1.2 0.8 0.0 3.12.2 0.3 0.6 0.6 0.0 9.06.4 2.3 0.3 0.8 0.0 2.41.6 0.3 0.4 2.3 0.2 7.00.3 5.1 1.6 4.3 0.3 9.45.1 5.8 0.6 0.1 0.0 22.618.8 8.2 0.5 1.0 0.0 25.119.7 8.3 3.0 5.0 0.1 24.220.9 3.0 3.2 1.1 0.0 6.73.3. 3.4 1.2 0.1 0.0 5.72.8 0.8 2.1 0.3 0.0 8.14.3 2.8 2.3 0.2 0.0 8.63.4 1.2 8.7 0.8 0.0 14.912.1 4.0 1.9 0.6 0.0 29.420.2 13.1 1.0 0.8 0.0 17.413.5 4.8 3.0 2.3 0.2 48.77.2 41.1 10.9

OCR for page 19
26 FACTORS AFFECTING CONTRACEPTIVE USE In the three higher use countries, the pill is the most widely used mod- ern method, followed by the IUD, injection, and female sterilization (though the rank ordering of these other methods varies by country). Of note is the fact that 5 percent of women in Kenya and 4 percent in Botswana have undergone sterilization, contradicting the frequent assumption that this method is totally unacceptable in any sub-Saharan African country (see Chapter 4 for discussion). Only two countries report any use of vasectomy: Zimba- bwe (0.2 percent) and Botswana (0.3 percent). Chapter 5 provides further discussion of the use of modern methods in specific countries. Among traditional methods, rhythm is most commonly used, except in Zimbabwe and Botswana where the main traditional method is withdrawal. In no country is the current use of traditional methods greater than 10 percent. Trends in method mix to date suggest that as countries move from low contraceptive prevalence to higher levels, the use of traditional methods gives way to an increase in oral contraception (particularly in the 10-45 percent CPR range) followed by a rapid rise in female sterilization in coun- tries above the 45 percent level (Destler et al., 1990~. Quality issues change as the mix changes, particularly because the more permanent and effective methods (the IUD, NORPLANTR, voluntary surgical contraception) require clinical settings. National contraceptive prevalence rates tend to mask the substantial regional diversity that exists within countries. These differences are shown in Table 2-3. The national-level data for each country refer to the percent- age of women married or in union currently using a modern contraceptive method. The regional data that follow indicate the number of percentage points by which each region differs from the national average. The range between regions within a given country (i.e., the number of percentage points between the lowest and highest prevalence) varies from 5 percentage points (in Burundi and Senegal) to 30 percentage points (in Zimbabwe). In fact, the countries with the greatest interregional variations are those with higher than average national levels of prevalence, which result from the high levels of use (by African standards) in selected regions. For example, two of the regions with high positive deviations Harare and Chitungwiza, Zimbabwe, with 48.0 percent modern use, and the Central province of Kenya with 30.8 percent are in countries with the highest prevalence. However, Kampala, Uganda, shows the greatest positive deviation of all the regions (15.4 percentage points above the national level) with a prevalence rate of 17.9 percent.

OCR for page 19
LEVELS AND TRENDS IN CONTRA CEPTIVE USE TABLE 2-3 Regional Variations in Modern Contraceptive Prevalence Rates, DHS, 1986-1990 (average prevalence) Country and Region/Province Prevalence Rate Burundi Imbo Mumirwra/Mugamba Plateau Lowlands Range Ghana Western Central Greater Accra Eastern Volta Ashanti Brong Ahafo Other regions Range Kenya Nairobi Central Coast Eastern Nyanza Rift Valley Western Range Liberia Since Grand Gedeh Montserrado Rest of country Range Mali Kayes, Koulikoro Sikasso, Segou Mpoti, Gao, Tombouctou Bamako Range Nigeria Northeast Northwest Southeast Southwest Range 27 (1.2) 4.9 0.4 -0.4 -1.2 5.1 5.2 -2.0 -0.3 5.4 0.6 -1.3 1.3 0.0 -4.5 9.9 (17.9) 10.0 12.9 -3.1 1.6 -7.7 0.2 -7.9 20.8 (5 5) -1.6 -2.6 4.2 -1.1 6.8 (1.3) -0.5 -0.7 -0.5 4.7 6.8 6.0 -4.0 -4.8 2.8 9.0 13.0

OCR for page 19
28 FACTORS AFFECTING CONTRACEPTIVE USE TABLE 2-3 continued Country and Region/Province Prevalence Rate Senegal West Central Northeast South Range Sudan (northern) Khartoum Northern Eastern Central Kordofan Darfur Range Togo Coast Plateau Central Kara Savana Range Uganda West Nile East Central West Southwest Kampala Range Zimbabwe Manicaland Mashonaland Central Mashonaland East Mashonaland West Matabeleland North Matabeleland South Midlands Masvingo Harare/Chitungwiza Bulawayo Range (2.4) 3.1 -1.8 -1.8 0.0 4.9 (s.5) 9.8 2.8 -3.4 -1.4 -4.1 -5.3 15.1 (3.1) 1.5 -0.7 -1.2 0.2 -2.8 4.3 (2.5) -2.5 -0.5 -0.1 o.s -1.6 15.4 17.9 (36. 1) -10.5 4.0 7.0 7.1 - 18.1 - 14.9 -0.9 -0.8 11.9 5.6 30.3 NOTE: Shown in parentheses for each country is the average prevalence. The figures for each region represent percentage point deviations from the national average. The range indicates the difference between the highest and the lowest regional rates. SOURCE: DHS national survey reports.

OCR for page 19
LEVELS AND TRENDS IN CONTRACEPTIVE USE Results of WFS, CPS, and Other Surveys, 1975-1990 29 Data on contraceptive use by method are presented in Table 2-4 for African countries participating in the WFS, CPS, and other national-level fertility or family planning surveys from 1975 to 1990. The countries that have had at least two surveys are discussed under the section on trends below. For the ten countries that have not had a subsequent DHS (Benin, Cameroon, Cote d'Ivoire, Ethiopia, Lesotho, Mauritania, Rwanda, Somalia, South Africa, and Zaire), these are the only available national-level data on contraceptive prevalence. Several findings merit comment. First, contraceptive prevalence was less than 6 percent at the time of the survey in all but two of these ten countries (the exceptions being South Africa with 48 percent in 1981 and Rwanda with 11 percent in 1983~. Although some change may have taken place since that time, the available data do not indicate any significant use of contraception (apart from South Africa, which is by no means typical of the region) among this group of countries. Second, even at these very low levels of prevalence, there was greater use of traditional than modern meth- ods in all countries except Ethiopia and South Africa. For example, Rwanda reported an overall prevalence of 11 percent, but only 1 percent of women were using a modern method. Third, among the users of modern methods, the most common method was the pill, except in Rwanda, where injection was the most common. This finding parallels the experience in other Afri- can countries, as well as numerous other developing countries in other re- gions (United Nations, 1989a). Trends in Modern Contraceptive Prevalence The analysis of trends in contraceptive prevalence in the region is lim- ited by the number of countries for which there are at least two points of observation over time. Apart from Kenya, a veritable demographic labora- tory in terms of the number and frequency of surveys conducted to date, only seven other countries have data from two time points: Botswana (1984 and 1988), Ghana (1979-1980 and 1988), Nigeria (1981-1982 and 1990), Senegal (1978 and 1986), South Africa (1975-1976 and 1981), Sudan (1978- 1979 and 1989-1990), and Zimbabwe (1984 and 1988-1989). In Zimbabwe, Botswana, and Kenya as well as in South Africa, there were dramatic increases in the use of modern methods between surveys (shown in Table 2-5), which is evidence of the growing acceptance of fam- ily planning in these countries during the 1980s. Modern prevalence rose an average of 1.2 percentage points per year in Kenya, 1.9 points per year in Zimbabwe, and 3.3 points per year in Botswana. In South Africa, it rose an average of 2.0 points per year between the two surveys.

OCR for page 19
LEVELS AND TRENDSIN CONTRACEPTIVE USE 4 1 ported awareness of different contraceptive methods. Specifically, the in- terviewer asks the respondent to name all methods of which she or he has heard ("spontaneous mentions. For each method that is not mentioned, the interviewer names the method, gives a one line description of it, and then asks if the respondent has ever heard of it ("prompted recalled. In the family planning literature, these two questions are combined into a single variable that is conventionally labelled as "knowledge," either of a specific method or of any method. We report knowledge of methods in Table 2-8. More than two-thirds of the women ever in union have heard of at least one modern method of contraception in almost all of the 12 African DHS countries; the exceptions are Burundi (63 percent) and, more notably, Mali (29 percent). Knowledge is higher in urban than in rural areas, although the gap is small or insignificant in the higher use countries where contraception is known to more than 90 percent of women, whether urban or rural. The data also show a positive and monotonic relationship between knowledge and level of education. However, Zimbabwe is remarkable, because 94 percent of women with no education have heard of at least one modern method, which suggests that family planning programs have reached all segments of the population. Knowledge of contraception is related to age in a pattern best described as an inverted U. although the differences by age are not significant in Burundi, Mali, and Togo. Generally, women aged 15-19 are less likely to have heard of a modern contraceptive method than are women aged 20-39. However, among women 40 and over, the percentage drops markedly, re- flecting perhaps the lower educational attainment and lower need for con- traception over the life cycle among this age group. Zimbabwe is again the only exception, where at least 95 percent of women in each age group know of a modern method. 1lSome might question the validity of this self-reported "knowledge," especially when the respondent can qualify for "knowing a method" simply by a nod of the head when the inter- viewer reads those methods not previously mentioned. However, the results obtained from this series of questions from numerous DHS or DHS-type surveys tend to be highly consistent with other sources of data. For example, the levels of professed knowledge are high for those methods most widely used in a given country, as shown by program service statistics. By contrast, few respondents claim to have heard of vasectomy, a method that is not widely promoted in public programs in developing countries. Also, data on knowledge of methods tend to show a consistent pattern of increase in subsequent surveys in a given country and to correlate predictably with education, urban residence, and economic status. Thus, the problem lies not with the utility of the indicator, but rather with the label "knowledge" when in fact "awareness" is a more accurate term. In this report we have retained the term "knowledge" to be consistent with the family planning literature.

OCR for page 19
42 FACTORS AFFECTING CONTRACEPTIVE USE TABLE 2-7 Women Who Have Ever Used a Modern Method (percent) Residence Education Countrya Total Urban Rural None Primary Secondary Botswana 60.2 71.2 55.1 40.8 65.5 81.7 Burundi 2.5 26.7 1.5 1.4 4.8 28.8 Ghana 23.1 33.2 18.4 11.0 30.9 52.1 Kenya 29.1 44.4 26.1 16.1 31.1 48.1 Liberia 17.8 28.9 10.8 8.4 23.0 62.2 MaliC 3 4 11.8 0.4 1.3 12.6 56.3 Ondo 9.4 13.8 6.5 3.3 10.7 23.7 Senegal 6.5 16.7 1.1 2.3 19.5 50.6 Sudan 18.1 34.9 8.2 7.0 29.1 41.8 Togo 10.6 21.0 6.2 5.2 15.6 38.8 Uganda 7.6 34.1 4.6 2.2 7.9 37.4 Zimbabwe 62.7 76.0 56.6 47.2 62.6 76.2 NOTE: The base for the table is all women age 15-49 ever in a union. aThe Nigeria DHS of 1990 is not included because the standard recode data tape was not available during the writing of this report. TABLE 2-8 Women Who Know Any Modern Method (percent) Residence Education Countrya Total Urban Rural None Primary Secondary Botswana 94.4 99.4 92.1 85.2 98.7 99.5 Burundi 63.3 92.3 62.1 60.1 75.9 94.2 Ghana 76.8 88.4 71.5 60.5 89.8 97.2 Kenya 91.3 95.4 90.6b 82.8 94.4 98.5 Liberia 69.5 78.5 63.9 61.5 85.6 96.0 MaliC 28.8 56.5 19.2 23.1 60.8 100.0 Ondo 49.4 60.0 42.3 33.2 58.2 79.2 Senegal 68.6 87.9 58.7 64.8 87.7 97.8 Sudan 71.0 91.1 59.1 54.8 91.1 98.3 Togo 81.4 92.4 76.7 75.5 91.7 98.2 Uganda 78.5 93.6 76.9b 68.0 85.2 96.3 Zimbabwe 97.9 99.0 97.4b 94 0 98.4 99.5 NOTE: The base for the table is all women age 15-49 ever in a union. aThe Nigeria DHS of 1990 is not included because the standard recode data tape was not available during the writing of this report. bThe contrast is not significant at the .05 level in a multivariate logit model.

OCR for page 19
LEVELS AND TRENDS IN CONTM CEPTI VE USE 43 Age 15-1920-2425-2930-3435-3940-4445-49 36.658.367.868.365.454.131.2b 3.81.gb2.9b3.0b1.7b2.2b2.6b 11.619.425.627.524.826.416.9 10.222.630.235.834.828.926.7 8.118.520.222.419.215.512.5 3.83.7b4.9b4.4b2.4b1.8b0.0 7.99.5b12.59.511.77.75.9 1.94.0b6.6b9.2b10.46.36.3 5.314.118.423.320.219.418.3 9.110.4b1 l.0bl3.2b9.8b10.3b7.4b 4.45.9b8.88.510.110.44.7 41.167.770.571.863.952.440.9b bThe contrast is not significant at the .05 level in a multivariate logit model. CFor Mali, logistic regression omits cases in which the woman was age 45-49. Age 15-19 20-24 25-29 30-34 35-39 40-44 45-49 92.4 96.7b96.9 95.9 95.7b 9o.9b 84.2b 61.3 61.4b71.0b 62.9b b b 50.6b 70.5 78.7.81.4 77.8 77.5 73.8 66.9b 86.9 94.594.2 93.1 92.8 85.2 82.8b 54.2 73.772.8 74.4 69.5 67.2 63.2 32.7 33.0b34.1b 30.7b 24.3b 22.0b 14.1 40.9 52.257.4 55.5 52.5 46.2 34.2 51.2 72.671.1 74.3 71.0 65.3 66.4 67.9 75.0b72.7 73.1 69.6 67.6 65.0 76.1 82.4b84.1b 83.sb 80.6b 8o.sb 74.9b 75.3 79.5b83.4 77.5 81.2 75.4b 69.5b 96.0 98.798.4 99.1 98.0 94.9b 96.5b CFor Mali, the logistic regression omits cases in which the woman had secondary schooling or more.

OCR for page 19
44 FACTORS AFFECTING CONTRACEPTIVE USE Sources of Modern Contraceptive Methods Information on knowledge of sources of contraception is one indicator of the extent to which contraceptives are accessible in a given country (although the data pertain to perceived availability, not actual availability). As shown in Table 2-9, among women aged 15-49 who know at least one method, a high proportion (78 to 100 percent) also know a source where they can obtain contraception. However, the type of source varies markedly by country, reflecting differences in the configuration of family planning service delivery in these countries. The most frequently mentioned source is a government clinic or pharmacy, except in Senegal where three-quarters of the respondents cited a private clinic or other service delivery point. However, knowledge of a second source is much more limited; only in Ghana can at least 50 percent of the respondents name a second source (data not shown). The strong influence of government programs is reflected in the data in Table 2-9. Far more women know of government than private sources in 11 of the 12 countries. One can assume that the relatively low levels of knowl- edge of private sources in Botswana, Kenya, and Zimbabwe simply reflects the extent to which family planning services are delivered through govern- ment facilities. In contrast, at least one-third of respondents knowledgeable about at least one source mention a private source (clinic or pharmacy) in the low-prevalence countries of Senegal, Ghana, Togo, and the Sudan. Given that the pill is the most widely known and used modern method in Africa, it is also of interest to examine knowledge of at least one source of the pill. Not surprisingly, the most widely known sources of the pill are similar to the most widely known sources for any method, although the specific percentages knowing each type of facility tend to be lower (data not shown). With the growing concern about AIDS, knowledge about condoms takes on a new importance. Table 2-10 provides information on women ages 15- 49 currently in a union who have knowledge of condoms. There is a re- markably wide range in awareness of condoms (or willingness to admit this knowledge to the interviewer) among countries: In Burundi, Mali, and Sudan, less than 30 percent of women who know any modern method ac- knowledged having heard of the condom, in comparison to 92 percent in Botswana. Among women with knowledge of condoms, most (67 percent or more) also know at least one source from which they believe condoms can be obtained. Government clinics or pharmacies are the most widely cited sources in eight of the eleven countries. By contrast, private pharmacies are most frequently mentioned in the remaining three countries (Ghana, Sudan, and Togo).

OCR for page 19
LEVELS AND TRENDS IN CONTRA CEPTI VE USE Fertility Preferences 45 Data on reproductive preferences are useful indicators of fertility norms and intentions. Both the WFS and DHS collected data on ideal family size and intentions to have another child. Data from the DHS on ideal family size among ever-married women show that the preferred number in sub-Saharan Africa averages just under six. The lowest ideal number is 4.7 (Kenya) and the highest is 7.1 (Senegal). The ideal number has declined between the WFS and DHS; for the three countries that conducted both surveys, Kenya showed the greatest decline (35 percent) and Ghana showed the smallest decline (9 percent) (Westoff, 1991~. Care should be taken in using these numbers as predictors of change in fertility; they are influenced by the number of children the respondent already has. The DHS also collected information on whether married women in- tended to postpone or halt childbearing. In sub-Saharan Africa, more women indicated that they wanted to delay a future birth than avoid one (Westoff, 1991~. Between the WFS and DHS, the number of women who wanted no more children increased in the three countries that conducted both surveys. Kenya showed the greatest increase (32 percent), followed by Ghana and Senegal (10 and 9 percent, respectively) (Westoff, 19914. Such data are indicative of changes in fertility in the short run (Westoff, 1990~. Westoff and Ochoa (1991) have used data on reproductive preferences to develop measures of unmet need for family planning. They define unmet need as the proportion who are exposed to the risk of conception, are not using a contraceptive method, and say they want to delay or stop childbearing. The total demand for family planning is the proportion of women with unmet need plus the women who are currently using contraception. Results from Westoff and Ochoa show that most of unmet need is for spacing rather than limiting births in sub-Saharan Africa, in sharp contrast to other regions of the world. Table 2-11 summarizes unmet need, as calculated by Westoff and Ochoa, by subgroup for eight populations in Africa. Unmet need gen- erally increases with number of children; is greater in urban areas than rural areas, except in Botswana and Kenya; and is greatest for women with pri- mary education, except in Botswana, where it is greatest among the unedu- cated, and in Burundi, Liberia, and Uganda, where it is greatest among those with secondary education. Overall, unmet need is highest in Togo (40.1 percent) and lowest in Mali (22.9 percent), suggesting that there is a substantial demand for family planning in all eight countries. Compansons of unmet need for limiting births between the DHS and the WFSi2 in Ghana 12Comparisons are noted only for unmet need for limiting births because the WFS did not collect information on spacing preferences.

OCR for page 19
46 "s o o Cat en .~ no 3 be so o o o Cut C) o V, o - o Cat Cut C) o ~ ~ ~ o o ~ ~ o . . v Cut of 4 - C> C) C) o ~: o CD C~ o ~S E~ 04 o ~ oo oo U~ e~ oo o . . . . . . . ~ o C~ o ~ o C ~_4 o ~ ^ U' ~ ~ ~o o ~ ~ o o o ~ ~ ~ ~ V - ~ ~ ~ ~ ~ e~ ~ ;~ Ce P~ P~ . ;> ~ - .- .- ~L) P" ~ ~ C: ;^ o ~ o o o V o .o C~ 3 o ~ o C~ C =C o o ~ ~ ~ ~ 0 cr 0 ~ ~ 0 ~ . . . . . . . . . . . - ~ C~ ~ oo ~ o ~ U~ ~ ~ ~ ~ _ ~ ~ _ =\ t_ ~ _ ~ ~ oo ~ ~ ~ - . . . . . . . . . . . C~ _ ~ _ o ~ ~ ~ ~ ~ - ~ C~ ~ ~ ~ o o ~ o o o o o o ~ ~o . . . . . . . . . . . o o ~ o o _ ~ o o o ~ _ ~ ~ ~ ~ a~ oo ~ oo . . . . . . . . . . . C~ ~ oo ~ ~ ~ oo ~ ~ ~o ~ ~ ~ oo ~ oo ~ ~ ~ oo oo a~ oo ~ O oo ~ ~ ~ ~ ~ . . . . . . . . . . . _ o oo oo ~ ~ ~ ~ ~ oo o~ ~ ~ C C .` 0.0 C o D o C ~0 3 o 3 ~ o C~ - .~: o o o C C) CO ,0. ~L> 0 U~ - oc o 3 - Ct ._ e~ o E~ . . E~ o z Ct Ct o C) (O C) 3 Oe~ C~ s: ~ t o o c) ~ o ~: o s: 3 ~: 3 - o U' - ._ C) C~ C~ o os ._ ._ Ct C~ s~ : - o. . ~ o s: ~: ~o o o U) C~ C~ ._ o o ~q e~ ._ ~ Z ~o . ~t_ Ct ~ ._ _! ~C~ ._ o ~: CD _ . ~D ._ Ct ;> ~t C, C) .S ~_ C40 . _ (L) o oc ao o:-~ ~ ~ 3 ~ = ~S: 04 C~ ~ - o ~ c: C;: ~ 5 C.) ~ ~ - - 0D Ct s~ ~ ;> O Ct' ~ O o ~ ~ Ct ~5 C) Z 3 c: ts p,.= Ct

OCR for page 19
47 C~ C) o V) o C) E~ Ct C) s~ o U) C~ ~4 ._ 3 o ~: o o o bC ._ 3 o C) b4 e~ P~ . . C~ o o V o Cq o U: o C~ C) Z o o ~o c: 'o o o oo ~ o ~ ~ CM ~ ~ ~ ~ . . . . . . . . . . . o ~ ~ o ~ o ~ o o , ~ o o ~ ~ ~ o o ~ o oo _ o o ~ _ ~ o o o ~ ~ o C~ V ~ CC ~ c) Ct _ s~ ~ C ;- =: oo In ~ ~ ~ 00 0 ~ O ~ ~ . . . . . . . . . . . oo oo O oo _ ~ Ir; c~) t~ Ir~ ~ C) ._ ._ - V ~:- 5: ~ CO ~ o ~ _ .= o ~ ~ ~ _ _ oo ~ ~ ~o ', s~ ~ o ._ _. _ ~ ~ ~ ~ ~ ~ ~ o o V, ~ ~ ~ ~ _ c: ~ E C ~ o 0 0 ~ ~ := V ~ C) ;~., o .O ~ ;> ~ Ct O '-v ~ o 0 ~ o 0 ~ ~ o 0 ~ ~ . . . . . . . . . . . o o o o o o _ o o o oo ~ ~ _ ~ o ~ oo oo ~ o ~ ~ oo o~ oo _ _ ~ o C~C~ o ~ ~ vo ~ ~ ~ ~ ~ C~ _. . . . . . . . . . . - 'oo o oo ~ U~ ~ ~ ~ ~ ~ o o ,,;~ ~ o 3 o o V ~ C4 ~ o C~ ~ t_ _ t_ ~ ~ ~ o ~ ~ . . . . . . . . . c~ ~ ~ - oo ~ cr ~ ~ c~ ~ ~ C~ ~ ~ ~ ~ ~ C~ ~ ~ oo c o ~o c, ct 3 v ~mmo~o~ C) CJ C~ Ct Ct C5 Ct Ct e~ ~: 0 0 O ._ :- O P~ _ Ct ~ C ~Z O V ~: U, O Z . _ 53 ~0 _ C) .= ~ ~O O ~ `: CC ._ o O C) _ ._ O ~ O ~O ._ ~ Ct ~O ~C40 ._ ._ Ce: ;^ Z Ce c: 5: . ~. 's" O .D -0 3 ~-9 ~ In ~ _ C OCR for page 19
48 FACTORS AFFECTING CONTRACEPTIVE USE and Kenya (data not shown) indicate that unmet need has increased, 80 and 129 percent respectively. RESULTS FROM MALE SURVEYS The evidence on levels and trends in contraceptive use presented thus far in this chapter comes from data gathered for women. Several DHS also questioned men. In Burundi, Ghana, and Kenya, a subsample of the hus- bands of the women interviewed were selected to respond to the question- naire for men. In Mali, a sample of men was selected from the households visited by the interviewers; thus, the sample included single men. Results from these surveys indicate that men were more likely to report using a contraceptive method than women, except in Ghana, where the reported use of any method was approximately the same. Contraceptive use among men was higher in urban areas; generally decreased with age, except in Kenya, where use increased and then decreased with age; and increased with education.~3 In all four countries, men were more likely to know of a method than women. In Burundi, Ghana, and Mali, of the men that knew any method, more than half of them had not talked with their wives about family plan- ning in the past year. In Kenya, 36 percent of the husbands had not talked with their wives. More than three-quarters of men in Burundi, Ghana, and Kenya approved of family planning; only 16 percent of men in Mali ap- proved. Except in Burundi~4, more than a third of the wives of husbands who approved of family planning either did not know their husbands' atti- tudes toward family planning or thought their husbands disapproved. In Burundi and Kenya, men and women indicated the same ideal family size, 5.5 and 4.8 respectively. In Ghana, men preferred more children than their wives, 7.6 versus 5.5. The DHS report for Mali did not report ideal family size for men. CONCLUSION Modern contraceptive use in sub-Saharan Africa remains low except in Botswana, Kenya, and Zimbabwe. In all countries, use of modern methods is associated with urban residence and greater education, perhaps indicating that future changes in the composition of the population might be associated with increased use. The knowledge of modern contraceptives is greater i3In Kenya, there was little difference in use between men with no education and those with primary education. i4Data for Burundi are not presented in the DHS report.

OCR for page 19
49 ~'e to two is: . 'it "o .s o at ~ ~o o ad ~ A 3 .- a: so o + it, ._ so Cal o ~ z C) o ~ o ~ C, -^ EM ~u ~ ~ ~ o Do ~ ~ ~ . . . . . . . . in ~ Do car car O ~ ~ ~ ~ ~ ~ ~ Cal Do ~ Go ~ ~ ~ o US . . . . . . . . Oo ~ ~ car ~ Do cry Cal ~ ~ ~ ~ ~ o Do Cal ~ ~ Cal . . . . . . . . ~ ~ ~ o o ~ o Go Cal ~ ~ ~ ~ ~ US Go Go . . . . . . . . ~ - , ~ ~ ~ ~ ~ ~ o o ~ ~ ~ ~ ~ o U~ ~ 00 0 0 00 C~ C~ . . . . . . . . C~ ~ ~ C~ ~ 0 ~ 0 ~ ~ ~ ~ 0 ~ ~ ~ . . . . . . . . ~ 0 oo ~ ~ C~ ~ ~ . . . ~ ~ U~ ~ ax ~ ~ . . . . 0\ 0 ~ ~q C~ cr ~ 0 ~ cr . . . . . . . . ~ ~ 0 C~ C~ . . . . . . . . \0 ~ ~ ~ 1- C'N CM C~ 0 ~ oo r~ 0 ox ~ ~ _ CM ~ ~ ~ ~ ,: Ct 0 ~ ..~: ~ 53 C OCR for page 19
50 FACTORS AFFECTING CONTRACEPTIVE USE TABLE 2A-1 Differentials in the Percentage of Women Who Currently Use a Modern Method Residence Education Country Total Urban Rural None Primary Secondary Botswana 31.7 40.8 27.5 18.9 32.9 50.3 Burundi 1.2 14.0 0.8 0.6 2.2 16.9 Ghana 5.2 8.1 3.9a 3.2 6.5 10.1 Kenya 17.8 25.5 16.4 9.7 19.2 29.3 Liberia 5.5 9.7 3.1 2.5 6.6 22.1 Malib 1.3 4.9 0.1 0.4 5.5 20.3 Ondo 3.8 5 3 2 Sa 1.9 3.6 9.0 Senegal 2.4 6.7 0.3 1.0 5.9 22.1 Sudan 5.5 11.3 2.2 1.9 7.9 14.9 Togo 3.1 6.5 1.7 1.7 3.9 12.2 Uganda 2.5 12.2 1.5 0.9 2.7 11.3 Zimbabwe 36.1 48.8 30.9 25.0 34.0 52.3 NOTE: The base for the table is all women currently in union. aThe contrast is not significant at the .05 level in a multivariate logit model. bFor Mali, the logistic regression omits cases in which the woman was age 45-49. than their use, and the knowledge of government sources of supply in the three higher use countries indicates the importance of government family planning programs in those countries. Estimates of unmet need indicated that there is substantial demand for additional family planning services, especially for spacing births.

OCR for page 19
LEVELS AND TRENDS IN CONTRA CEPTI VE USE Age 51 i5-19 20-24 25-29 30-34 35-39 40-44 45-49 14.5 25.2a 36.0 33.7 37.5 35.0 14.7a 0.6 o.7a l.4a 1.sa l.3a 1 oa 1 6a 2.3 3.4a 4.3a 6.9 5 pa 9 0 4.2a 6.7 11.8 16.8 22.2 22.9 21.2 17.5 2.0 4.9a 6.7 6.3 5.1 5.7 7.1 1.4 l.4a 1.6a 1.6a l.4a 0 ga 0.0 1.7 2. la 3.oa 4.6 5.4 4.2 3.5 0.5 0.8a 25a 4.8a 3.8 1 la 2~4a 2.2 42a 5 oa 7 9 5.7 7.7 4~4 2.1 1 ga 2 sa 3.7a 4.7a 42a 3.4a 1.2 1 1a 1 ga 2.6 6.0 5.2 2.5 28.3 41.5 43.8 42.9 30.8a 261a lima