Authors' reply



We thank Drs Bank and Jana for their interest. Difficulties in case ascertainment are a serious concern in many historical cohort studies. However, there are indications that case ascertainment in our study is satisfactory. In a previous study1 we reported that case fatality in 1967–91 ranged from 22 to 48 percent. Thus, to use the Cause of Death Registry as a supplementary source could improve ascertainment. However, only 249 additional cases of a total of 9592 cases were notified to the Cause of Death Registry but not to the Medical Birth Registry1. Furthermore, according to a perinatal database at the Department of Obstetrics and Gynaecology, University of Tromse, Norway, the occurrence of placental abruption in 1986–93 ranged from 0.7% to 1.0%, which is in agreement with data from the Medical Birth Registry (unpublished observations). Predefined criteria that have been used in earlier hospital based cohort studies have merely been based on the well known classical signs and symptoms of placental abruption and a retroplacental blood clot2. We therefore think that the gain in ascertainment using such criteria has been exaggerated.

Possible associations between low social status and occurrence of placental abruption as well as low fertility could explain the decreased fertility after placental abruption. In a recent report (unpublished observations) we have found increased risk of placental abruption in women with short education, even after adjustment for possible confounders such as maternal age and birth order. However, in a British study3 the likelihood of having a second pregnancy was considerable higher in social class IV-V than in 1– 111. If this is the case in the Norwegian birth population, adjustment for social class would most likely increase rather than reduce the association of a history of placental abruption without perinatal loss with low subsequent fertility.

Despite the rising rate of caesarean section and its possible ‘protective’ effect on placental abruption, recurrence of placental abruption did not decrease during the study period. To some extent this then may be explained by an increased risk of placental abruption in a subsequent pregnancy after caesarean section. Linked data from the Medical Birth Registry and population census reports (unpublished observations) suggest that caesarean section is associated with an increased risk of placental abruption in later pregnancies, which is consistent with earlier reports.