Doctor: Pierre Delorme

Title: Cause of preterm birth as a prognostic factor for children outcome

Supervisor: François Goffinet, Gilles Kayem

Doctoral school: ED 393 Epidemiology and Biomedical Information Sciences, Université Paris Cité

Date of thesis defense: 11/2020

Jury: Patrick Truffert, Marie-Victoire Senat, Florence Bretelle, Josiane Warszawsi

Thesis summary:

Gestational age at birth is the main factor used to assess the prognosis of children born prematurely. Currently, while there are several causes of preterm labor with different pathophysiological mechanisms, this cause is not used for prognostic evaluation. However, the pathophysiological mechanisms involved could directly impact the prognosis of the child. The main objective of this thesis was to establish a classification of the causes of premature delivery that is homogeneous from a pathophysiological point of view, in order to analyze the association between the cause and the outcomes of newborns. A second objective was to allow the analysis of other specific prognostic factors within the homogeneous groups thus formed. This work was carried out using data from the EPIPAGE 2 cohort which included more than 8,000 premature births in 2011. The first part was to create a classification based on pathophysiology. We have defined cause as the pathology that resulted, directly or through a complication, in premature birth. The classification included 6 classes, premature labor with intact membranes (43.8%), premature rupture of membranes (PROM) (23.9%), pregnancy related vascular disorders without associated Intra Uterine Growth Restriction (IUGR) (12.6%), with associated IUGR (10.7%), IUGR without maternal pathology (6%) and placental abruption during pregnancy without diagnosed pathology (3%). The second part focused on studying the association between the cause of preterm delivery and newborn mortality. A multi-level logistic regression with fixed effects, with the mother at the first level and the center at the second level was used to take better account of the data structure. Compared to the group of births caused by preterm labor with intact membranes, the risk of hospital mortality was 2-3 times higher for births caused by IUGR, whether or not it was associated with maternal vascular disease. The other causes, in particular PROM and pregnancy related vascular disorders without IUGR, had a risk of mortality comparable to that of preterm labor with intact membranes. The third part analyzed the prognostic role of severe umbilical Doppler abnormalities on the neurological outcome at two years of children in the 3 groups of causes linked to a pregnancy related vascular disorder. In children born due to vascular disease in pregnancy, the observation of severe umbilical Doppler abnormalities was associated with an increased prevalence of moderate to severe cerebral palsy or sensory deficits at two years of age. This classification has enabled the production of seven publications on the prognosis of the child depending on the context of the birth. The recent acquisition of data from the 5-year follow-up EPIPAGE 2 will make it possible to continue this work with the impact of the cause or specific factors within homogeneous groups of patients on the cognitive outcomes of children.


Back to top