Content area
Abstract
In 1986, the Demographic and Health Surveys Project administered an experimental survey in Peru that included the first six-year calendar history of contraceptive use. This thesis is a study of individual level changes in the contraceptive behavior of Peruvian women from 1981 to 1986 as given in the calendar histories collected by the DHS project.
Nearly half of all Peruvian women who begin to use a method will stop using it within one year, and the highest risk of discontinuation is for reasons other than failures and stopping to get pregnant. Twenty-nine percent of women who begin use of any method discontinue for non-pregnancy related reasons within twelve months of beginning use of the method, and 65% of these women immediately adopt another method. The focus of this thesis is on discontinuation of method use and the underlying patterns of contraceptive behavior reflected in discontinuation, including method switching. The patterns of contraceptive use following discontinuation are investigated using multiple increment decrement life tables to examine women's contraceptive behavior within twelve months after they stop using a method. These patterns suggest that the accessibility of methods may have more influence over a woman's choice of method than her desire to find an effective and acceptable contraceptive.
Factors that affect the risk of discontinuation--a woman's age, parity, education, residence, duration of use, and the particular method used--as well as the behavior immediately following method discontinuation are examine in multivariate analyses of the risk of method discontinuation. It appears that the processes that cause women to discontinue method use without immediately adopting another method, and those that lead to a direct switch of methods, are very much the same. Women may stop method use or switch methods when a contraceptive they are using becomes unavailable, and they probably switch methods when preferred alternatives can be readily obtained.
The potential demographic impact of contraceptive discontinuation is examined using a model which simulates reproductive histories in populations with varying levels of contraceptive use and effectiveness, as well as method discontinuation and resumption of use. This exercise indicates that method discontinuation may have a substantial impact on the level of unwanted fertility.
