irreversible method of contraception is itself a commentary on women's judgement of the drawbacks of the available temporary methods; though it should also be borne in mind that the full range of methods is frequently not readily available. In the case of modern methods such as the pill and IUD, concerns about method safety are mentioned repeatedly, in interviews, as a major reason for non-use among women who are at risk of unintended pregnancy. Side effects, or fears about them, are also among the major reasons for discontinuing use of the more effective methods. The data gathered in most fertility surveys contain little or no information about the type and severity of side effects experienced, or perhaps merely feared. Separate investigations are needed in order to judge whether women's concerns are medically well founded, and whether provision of more complete and more accurate information would increase the use of the effective reversible methods. Many women may have received medical advice against using certain methods; pill use by older women is often discouraged, for instance. In the United States in 1982, over one-third of former pill users said a doctor had told them to stop ( Mosher and Bachrach 1988).
While barrier methods and non-supply methods (periodic abstinence, withdrawal) typically arouse no fears about safety, these methods are less convenient to use and are more likely to fall. Accidental pregnancy during contraceptive use becomes more common, in relative terms, as the proportion of contraceptive users rises. In Latin America today an appreciable proportion of births result from contraceptive failure. For example, in the two years preceding the 1986-7 surveys in Brazil, Ecuador, and Peru, 15, 10, and 18 per cent respectively of all births were conceived while the couple was using contraception ( Bongaarts and Rodriguez 1991; Rutenberg and Blanc 1991); in Colombia, the Dominican Republic, and Trinidad and Tobago 10, 16 and 22 per cent respectively of the most recent births during the five years before the surveys resulted from contraceptive failure.17
The overall level of contraceptive prevalence in Latin America--estimated at 56 per cent of couples with the woman of reproductive age--is higher than in other developing regions, with the exception of East Asia. In most countries in Latin America and the Caribbean, contraceptive prevalence is in the range of 30-60 per cent at the most recent date, with most of the more populous countries having values in the range of 45-65 per cent.
Where trends can be assessed, the overall level of contraceptive practice is usually increasing, and in many cases is increasing rapidly. Countries with annual increases in prevalence of 2 or more per cent of couples per year (between the dates shown in Table 8.4) include Colombia, Ecuador, El Salvador, Honduras, and Mexico. Several others have increases that are only slightly less rapid. Indeed,____________________