An experiment using a month-by-month calendar in a family planning survey in Costa Rica

Abstract

Recent demographic surveys have incorporated a month-by-month calendar for the five-year reference period before the survey for the recording of fertility-related events (sexual unions, contraceptive use, pregnancies, and breastfeeding). In the 1986 survey of Maternal and Child Health and Family Planning in Costa Rica, approximately one-half of the 3,527 women interviewed were administered a questionnaire with traditional fertility and family planning questions; the other half were asked virtually the same questions, but the women's responses were entered in a month-by-month calendar. The assignment of questionnaire type was randomly alternated by cluster. Comparisons of the number of events (live births, pregnancy losses, and contraceptive use) showed that more events were recorded among the women in the calendar group. Significantly less erroneous superposition of events (contraceptive use in the last trimester of pregnancy and hormonal contraceptive use in the first month postpartum) was noted when the calendar was used.

PIP: There are many approaches to obtaining data on reproductive events. The World Fertility Survey uses a detailed pregnancy history questionnaire and family planning use for the last birth interval. The Demographic and Health Surveys use a live birth history and family planning method use for as many as 2 methods during each birth interval in the prior 5 years. The 1965 National Fertility Survey in the US used a chart or calendar to code respondent's reproductive status and events for each month in the reference period. This report is a comparison of data collection methods for the Costa Rica Maternal and Child Health-Family Planning Survey in 1986. The reference period for respondents' pregnancy history, family planning events, and marital and employment history was 5 years. Out of 322 clusters, the calendar method and the traditional questionnaire are alternated. Questionnaires and interviewer training were the same. The criteria for comparing data quality were established as 1) the higher number of events recorded, 2) the extent of conflicting events for contraception and pregnancy and for contraception and lactation, and 3) extent of method choice data missing from the calendar method. The results showed high response rates for both groups (95%). In the calendar questionnaire the mean number of pregnancy losses was significantly higher (by 17%) and the number of contraceptive use segments was higher (by 11%). A greater proportion of women reported with the calendar method at least 1 event for all variables. Inconsistency was reported for 10% of women in the traditional questionnaire on overlaps between pregnancies of 4 months and contraceptive use. Sometimes the overlapping continued through the entire pregnancy. Only 1% of women reported inconsistencies in reported pregnancy and contraception using the calendar method. Inconsistencies of lactation and contraceptive use occurred with 7% of traditional questionnaire reports and only 3% of calendar reports. 8.6% of women using the traditional questionnaire reported multiple method use. In the calendar method only the most effective method was indicated where there was multiple use. The recording of more events with the calendar method was confirmed by comparison conducted in Peru but not in the Dominican Republic.