[Morbidity and mortality of premature infants of less than 33 weeks]

The present study in a case-control study in a population based collection of data of all births from 22 to 32 weeks in all maternities of the Seine-Saint-Denis district for three years, 999 on 67819 total births. The study describes the site of births by level of perinatal care (level 3 maternity associated with a neonatal intensive care unit, level 2 with a neonatal pediatric ward without intensive care, and level 1 without any of these pediatric services. The proportion of births at a level 3 maternity is 6.3% for all births, and 12.9% for live birth of less than 33 weeks. The study describes the transfers of the live babies to a neonatal intensive care by external transfers from level 1 and level 2 units or internal transfers in level 2 and 3 units. All babies are transferred. The study describes the per partum and neonatal deaths up to 28 days by week of gestation duration and site of birth. The study uses the comparison between the deaths during delivery (from 28 to 32 weeks) and in the first 28 days compared to controls, using a regression analysis method. The risk of per partum and neonatal death is: OR = 8.08 [1.03-76.40] when site of birth 2 is compared to 3 and OR = 11.78 [1.33-103.77] when site of birth 1 is compared to 3. The study shows that the french obstetricians and pediatricians have not accepted the international consensus proposing that births of less than 33 weeks should take place in level 3 perinatal units, goal which is reached for 80% of those births, with less neonatal deaths and developmental handicaps.