Résumé : Statistics on induced abortion incidence in Brazil are considered highly deficient. The randomized response technique (RRT) improves estimates of induced abortion by allowing women to furnish accurate information indirectly, maintaining privacy. Use of the RRT indicates that around 80% of women with induced abortions refuse to admit their experience when questioned directly. The present study explored induced abortion-related behavior among two samples of women 15-49 years from a subdistrict of Sao Paulo (Vila Madalena) as a function of demographic factors. In the first sample (n = 996), the incidence of induced abortion in the year preceding the survey (1987) was estimated through use of the RRT. In the second (n = 1004), the same information was sought through a conventional approach. A total of 275 induced abortions were reported. Analysis of variance indicated single women, women 15-19 years of age, those who have not had live births, those who have a number of children below the expected ideal, users of contraception (especially inefficient methods), and women who do not have any moral resistance to the practice of abortion are most likely to resort to induced abortion. The multivariate model identified a woman's beliefs as to the moral acceptability of abortion ; a factor that derives from the broader cultural context ; as the variable with the highest predictive value. Overall, these findings indicate a clear need for development of birth control methods that meet the needs of young women at the beginning of their reproductive lives since these women comprise the majority of abortion seekers.
Site web : http://www.scielosp.org/scielo.php?script=sci_arttext pid=S0034-89101998000100002 lng=en nrm=iso