The Demographic and Health Surveys (DHS) program has included several questions on sexual intercourse in its model questionnaires that have been used in more than 25 surveys in Latin America, Africa, and Asia over the last five years. This article assesses the quality of the data on sexual intercourse for 12 DHS surveys and shows how these data may be useful for understanding contraceptive use dynamics and for organizing the management of family planning programs. The data show that there is considerable variation among countries in exposure to pregnancy prior to first marriage. Within marriage, the level of coital frequency varies with duration of marriage, fertility intentions, and type of contraceptive method used. Finally, in all countries there is some overlap between contraceptive use and sexual abstinence. This information can be useful in family planning programs for targeting particular populations, for assisting women to choose a method, and for assessing the effect of contraceptive use on fertility.
PIP:

Researchers took much care in designing questionnaires used for the Demographic and Health Surveys to minimize bias. For example, they included a question on the duration of time since last intercourse. Therefore, nonresponse rates for questions dealing with sexual intercourse were low (1-8%) in Latin America and Sub-Saharan Africa. Even though a substantial number of women began sexual relations at least 1 year before 1st union, few had regular relations before 1st union. Premarital sexual relations occurred more often in Sub-Saharan African than latin American and Trinidad and Tobago. For example, the highest rate for 15-17 year olds was in Uganda (32%) and lowest in Trinidad and Tobago (1%). for 22-24 year olds, it varied from 90% in Ghana to 6% in Trinidad and Tobago. Trends in the age at which women 1st marry did not necessarily parallel trends in the age at 1st coitus. Family planning programs should therefore inform never married women about risk of pregnancy even if coitus is sporadic. They should also use coitus dependent methods, e.g., condom or vaginal methods. Women were married and sexually active for an average of 15 years during their reproductive lives and between .5-2 years unmarried and sexually active. 85% of married women had had intercourse within the 3 months before the survey. Nevertheless 5-38% women who used contraceptives did not have sexual intercourse. Therefore reduction in fertility cannot be credited to only contraceptive use. The monthly frequency of coitus fell with rising marriage duration with some variation in the decline across countries. Women who wanted more children had intercourse more frequently than those who did not want more children and those who were undecided. Further women who used coitus independent methods had sexual intercourse more often than those who used coitus dependent methods.