|Abortion in Brazil: a demographic approach|
||Cecatti JG, Guerra GV, Sousa MH, Menezes GM.
||Revista Brasileira de Ginecologia e Obstetríci, 2010 Mar;32(3):105-11.
To evaluate the prevalence of spontaneous and induced abortion reported by a sample of Brazilian women interviewed in the National Demographic Health Survey of 1996.
This was a secondary analysis of the Brazilian DHS-96 database, with information from interviews with a representative sample of 12,612 women about their reproductive life, focusing on the prevalence of spontaneous and induced abortion in the last five years and the associated factors for the various regions of the country and for Brazil as a whole. The sampling method was implemented with a strategy selection in two stages, one for the households and the other for women. The prevalence of spontaneous and induced abortion was estimated for Brazil and regions, and the socio-demographic characteristics of the women were analyzed as a function of the abortion's experience. A multinomial regression model analysis was used for the identification of factors independently associated with both types of abortion; their OR and respective 95% CI are reported.
The prevalence of reported spontaneous abortion was 14% and the prevalence of induced abortion was 2.4% for the country as a whole. The state with the highest prevalence of induced abortion was Rio de Janeiro with 6.5%, followed by the Northeast region with 3.1%. The places with the lowest prevalence were the state of São Paulo and the South region. Both spontaneous and induced abortion showed higher prevalences with increasing age of the women studied. Being from the urban area (OR=1.5; 95%CI=1.0-2.3), having had more than one live child (OR=2.2; 95%CI=1.5-3.2) and being non-white (OR=1.4; 95%CI=1.0-1.8) were the main risk factors for induced abortion.
The non-modifiable risk factors for induced abortion identified in this study indicate the need for improvement of educational and contraceptive actions, with priority for these specific demographic groups.