This retrospective, multicenter, observational cohort study included 64,081 ovarian stimulation cycles and 31,027 intended frozen embryo transfer cycles from 18 IVF/ICSI centers (9 in Italy, 9 in Germany) between January 1, 2016 and December 31, 2020. The population comprised all second- and third-level ART cycles, with a median female age of 37 years (IQR 33–39) and median BMI of 22.7 kg/m². Antagonist protocols were used in 78.4% of cycles.
Main outcomes were clinical pregnancy and live birth rates. For fresh embryo transfers, the clinical pregnancy rate was 33.8% and live birth rate was 25.8%. For frozen embryo transfers, the clinical pregnancy rate was 34.2% and live birth rate was 24.3%. No comparative or adjusted analyses were reported.
Safety and tolerability data were not reported. The study's limitations include its descriptive design, which precludes causal inference, and inherent limitations related to variable completeness and a privacy-preserving data structure.
Clinicians should interpret these results as descriptive benchmarks from a large, harmonized real-world ART dataset. The findings may support hypothesis generation and local quality comparisons, but no causal conclusions can be drawn from this observational cohort.
View Original Abstract ↓
IntroductionAssisted reproductive technology (ART) practice varies widely across centers and evolves rapidly, often outpacing the generation of randomized evidence. OPERA (Observational ProjEct for a Research database for ART procedures) was developed to create a harmonized, multinational real-world database derived from routine clinical ART documentation and to provide a structured overview of contemporary clinical practice.Materials and methodsOPERA is a retrospective, multicenter database including 18 IVF/ICSI centers (9 in Italy and 9 in Germany) using a shared electronic platform (Meditex). All second- and third-level ART cycles performed between January 1, 2016 and December 31, 2020 were eligible for inclusion. Data were extracted from Meditex, harmonized into four relational domains (Baseline, Stimulation, Laboratory, and Pregnancy outcomes), and anonymized prior to aggregation. After data cleaning (duplicate removal, exclusion of unlinked frozen embryo transfer cycles, and cycles with incomplete follow-up), 64,081 ovarian stimulation cycles and 31,027 intended frozen embryo transfer cycles were included. Analyses were descriptive, and no imputation was performed.ResultsMedian female age was 37 years (IQR 33–39), and median BMI was 22.7 kg/m². Antagonist protocols were used in 78.4% of cycles. Clinical pregnancy rates were 33.8% for fresh embryo transfers and 34.2% for frozen embryo transfers, with corresponding live birth rates of 25.8% and 24.3%.DiscussionOPERA provides a large, harmonized real-world ART dataset enabling a comprehensive overview of clinical practice across multiple centers. While the descriptive design does not allow causal inference, it supports benchmarking and hypothesis generation for future analytical studies, acknowledging inherent limitations related to variable completeness and privacy-preserving data structure.