[Epidemiology of fetal deaths in the Seine-Saint-Denis perinatal survey].

Objective: To analyse the epidemiological characteristics of antepartum stillbirths.

Design: A population based cases-cohort study in a limited geographical area, the department of Seine-Saint-Denis, France. Prospective collection of cases and controls between the 1989-10-01 and 1992-09-30, and retrospective collection of data.

Subjects: All antepartum stillbirths of 28 weeks gestation of greater, except lethal malformations and multiple pregnancies.

Results: 273 antepartum stillbirths were collected during the 3-year study period (52, 2% of the perinatal deaths). The causes were mainly, abruptio placenta and cordonal causes; however, unexplained antepartum stillbirth was the most important group (38, 8%). Hypotrophy was present for 49.5% of cases versus 10.1% of controls (p < 0.001). It was associated with death whatever cause or gestational age, particularly in unexplained antepartum death (54.7%). In univariate analysis, the usual risk factors including obstetrical history, socio-economic characteristics, obstetric care and country of birth were associated with death. But, in multivariate analysis, only "no pregnancy déclaration", "deficient obstetric care", "history of stillbirth", "born in Black Africa" and "born in DOM-TOM" were significantly associated with death.

Conclusion: More than one half of perinatal deaths concern antepartum stillbirth; the causes have remained unchanged for twenty years and unexplained antepartum stillbirth remains the most important group. The principal risk factor is fetal hypotrophy. The other significant risk factors after multivariate analysis are deficient obstetric care and birth in Black Africa or in the DOM-TOM. To propose prevention actions, progress is necessary in in utero diagnostic of hypotrophy and in the understanding in the risk associated with country of origin.