Overview of hospitalizations in women undergoing oocyte retrieval for ART in the French national health data system.

Study question What is the incidence rate of complications in women undergoing ART procedures compared to the period prior to their first oocyte retrieval? Summary answer The study shows a significant increase in the post-ART incidence rate of some complications but a low overall rate of occurrence relative to the total number of oocyte retrievals. What is known already ART, widely used in Europe, accounts for 3.3% of births in France. The various studies of ART complications are fairly reassuring, showing relatively low overall complication rates but only few studies have used exhaustive national registers. Study design, size, duration The cohort for this study was identified from the comprehensive French national hospital-discharge database and includes women under 50 years with a first oocyte retrieval (T0) in 2012-2017, classified in three population subgroups according to the indication for oocyte retrieval: infertility (IF), oocyte donation (OD), and fertility preservation (FP). This study includes 156 916 women whose first oocyte retrieval occurred in 2012-2017 and 542 775 hospitalizations in 2010-2019 (excluding first retrieval). Participants/materials, setting, methods Hospitalizations for complications or others events (oocyte retrieval, delivery, pregnancy loss, and death in the hospital) during the 2 years before (control period) and after their first oocyte retrieval (post-oocyte retrieval period) were compared and expressed per 10 000 person-months (pm). Main results and the role of chance In the IF subgroup, incidence rates were significantly higher after (vs before) retrieval for hospitalized ovarian hyperstimulation syndrome (OHSS) (162 vs 6/10 000 pm), adnexal torsion (14 vs 3), venous thrombosis (8 vs 1), arterial thrombosis (3 vs 1), trauma (2 vs 1), and significantly lower for infections (61 vs 87). The higher incidences of OHSS, adnexal torsion and venous thrombosis could only partially be explained by the occurrence of pregnancy.In the FP subgroup, incidence increased significantly after (vs before) retrieval for hospitalized OHSS (55 vs 0), venous thrombosis (59 vs 4), and infections (176 vs 56). For the OD subgroup, hospitalized OHSS (116 vs 0) and bleeding (24 vs 0) were significantly higher after (vs before) retrieval. Limitations, reasons for caution The French national health data system, despite all its advantages, present some limitations such as the risk of coding errors. The unavailability of some personal information and the absence of consideration of risk factors prevented us from adjusting the risk. Finally, only complications resulting in hospitalization were analyzed which probably leads to their underestimation. Wider implications of the findings The use of medico-administrative bases will be a valuable tool in public health and will furnish a better overview of the complications. Further studies are needed to complete this analysis. Adding information on drugs would help to better define T0 and less severe complications. Study funding/competing interest(s) N/A. Trial registration number N/A.