P-292 It is feasible to carry out a large-scale multi-national Assisted Reproductive Technology Real World Data (RWD) collection prospectively, encompassing collected patient and treatment cycle information

Abstract

Abstract Study question Is it feasible to carry out a large-scale multi-national Assisted Reproductive Technology Real World Data (RWD) collection prospectively, encompassing collected patient and treatment cycle information? Summary answer The creation of an international database collecting RWD is feasible. The quality and quantity of the data allow them to be used for research purposes. What is known already Real World Evidence (RWE) may help to answer clinical research questions that are relevant to a broad population of patients. For 40 years, Assisted Reproductive Technology (ART) treatment cycle data have been collected in national registries. The aggregated national data are further collated on an international level. These data are useful for describing secular trends in ART utilization and associated outcomes. However, registries often fall short in providing patient-level and treatment-cycle-level information for data collation and analysis. Furthermore, national registries may restrict access of raw data for open research, for political, infrastructural, or legal reasons. Study design, size, duration In a first phase, the OPERA database (Observational retrospective ProjEct for a Research database for ART procedures) included data from cycles recorded between January 2016 and December 2020 in eighteen clinics, providing a robust sample size for Real World Data analyses. Fondazione per la Ricerca Ospedale di Bergamo (FROM), together with expert clinicians in the ART field, promoted the OPERA project with the financial support from two pharmaceutical companies (Merck KGaA, IBSA SA). Participants/materials, setting, methods The Eighteen clinics located in Germany and Italy joining the OPERA data collection, have proven expertise in ART procedures and use MedITEX® DB. The study protocol was reviewed by the competent Ethic Committees and Data Protection Officers. Among the outcomes collected were the mean number of cycles, the pregnancy and birth rate per Embryo Transfer, and the pregnancy and birth rates as a function of female age. The first extraction was scheduled for December 2022. Main results and the role of chance Around 75,000 ART cycles were collected from the first ten participating centers. Among the around 45,000 cycles that have available fertilization information , 11,553 (25.7%) were the In-Vitro Fertilization treatments (IVF), 31,202 (69.3%) Intracytoplasmic Sperm Injection treatments (ICSI), and 2,282 (5.1%) a combination of the two techniques. Of the adopted stimulation protocols, 38.43% (10,538) of the total were with variable antagonists, 31.27% (8,574) were with a single antagonist, 24.31% (6,665) were with a long agonist, 5.51% (1,511) were with a short agonist, and 0.35% (96) were with fixed antagonists. The average number of oocytes retrieved is 11.28, with a median of 10.0. The average number of 2 Pronuclear Oocytes (2PN) obtained is 3.7, with a median of 3.0. Moreover, the average number of 8 cell embryos at day 3 is 2.0, with a median of 2.0. It was also recorded the rate of unsuccessful cycles with no oocytes retrieved (4.8%). Among the cycles collected, 2.9% of the total were from frozen oocytes. Most of the ET (40.0%) were performed 5 days after the oocyte retrieval. Limitations, reasons for caution Amassing ART cycle information does not compensate for the well-known insufficiencies of primary data quality, of information on confounders and of access to patient level data. Moreover, the technical risk of this project is the potential loss of participant confidentiality, mitigated through a full anonymization. Wider implications of the findings The OPERA project may facilitate the transition of ART surveillance from a national aggregate level to a large-scale individual patient and treatment cycle level. Furthermore, as an evolutionary path, the OPERA project may help building large and longitudinal databases, suitable for appropriate epidemiological research in ART. Trial registration number not applicable