Fetal Wastage in a Sample of Taiwanese Women

In any retrospective study, it is extremely difficult to collect complete and accurate information about pregnancy losses. The task becomes even harder if one tries to differentiate between types of pregnancy losses such as neonatal deaths, spontaneous abortions, induced abortions and stillbirths, because for this purpose one needs to know the exact period of gestation for each pregnancy. To believe that respondents are able to give such details accurately is expecting too much from them. Estimates of fetal wastage based on data collected in a retrospective survey are subject to two types of errors: (1) errors in differentiating between types of fetal wastage and (2) errors in reporting the complete number of pregnancy losses.' The magnitude of these errors is difficult to assess. Nevertheless, in this paper an attempt is made to estimate the level of pregnancy loss from the data collected retrospectively for a probability sample2 of 2,443 Taiwanese women between the ages 20 to 39 living with their husbands in Taichung City at the end of 1962. A simple technique is developed for gauging the level of underreporting of pregnancy losses. Based on estimates corrected for memory bias, variations in pregnancy losses are studied by age of woman, pregnancy order and the use of contraception during the pregnancy interval prior to the conception. For the purpose of this analysis, pregnancy losses are classified