[Proposed method to estimate underreporting of induced abortion in Spain]

In Spain, from 1987 to 1990 the rate of legal abortion reported to the health authorities has doubled; nevertheless, the observed geographical differences suggest to an underreporting of the number of voluntary pregnancy terminations. Based on information on several sociodemographic, economic and cultural characteristics, contraceptive use, availability of abortion services, fertility indices, and maternal and child health status, five homogenEous groups of autonomous region were identified applying factor and cluster analysis techniques. To estimate the level of underreporting, we assumed that all the regions which shape a cluster ought to have the same abortion rate that the region with the highest rate in each group. We estimate that about 18,463 abortions (33.2%) were not reported during 1990. The proposed method can be used for assessing the notification since it allows to identify geographical areas where very similar rates of legal abortion are expected.