Résumé : Abortion is one of the main causes of maternal mortality in Latin America. Nevertheless, its geographical distribution and epidemiological evolution are not known, because of the insufficiencies of the health information systems, and of the negative image of the abortion which leads to an under-declaration. The present study gives an exhaustive description of hospital statistics on abortion in five regions of Ecuador since the beginning of its registration in 1965. In order to reduce the bias of health care access variations in the period, the indicator chosen is the rate (by year and area) of abortions for 100 pregnancies and deliveries which gave place to an hospitalization. The rate increases until the seventies and then diminishes the maximum is higher in the urban areas than in the rural areas, and earlier on the coast than in the mountains.The interpretation of these phenomenons must take in count the fact that the data come from hospitals, that information is lacking about rural populations and statistics no differentiate spontaneous and induced abortions. In spite of these limitations, it seems that these results traduce a real decline of abortion, which further studies will have to confirm and explain.
Notes : Francés/français/French