Jose G Franco, Claudia Petersen, Laura D Vagnini, Fabiana C Massaro, Bruna Petersen, Andreia Nicoletti, Juliana Ricci, Camila Zamara, Isabela M Pasotti, Renata A Pouza, Bianca C Matuella, Elisange-la V Espirito Santo, Joao B Meziara, Antonio H Oliani, Joao Batista A Oliveira
Real-world evidence does not demon-strate a consistent benefit of PGT-A, and the interpretation of effectiveness is strongly dependent on outcome denomi-nator selection. Therefore, the routine use of PGT-A should not be considered justified, and the medical literature has a duty to precisely define the subpopulations in which PGT-A might have value.
OBJECTIVE: The routine use of preimplantation genet-ic testing for aneuploidy (PGT-A) remains controversial in assisted reproductive technology, particularly regarding its effectiveness in improving live birth rates. Evidence is derived from randomized controlled trials (RCTs), non-ran-domized studies, and real-world data (RWD). Although RCTs offer strong internal validity, they often lack exter-nal validity. Non-randomized studies may suggest associ-ations, but their lack of randomization limits causal infer-ence because of potential bias. RWD reflect routine clinical practice but are heterogeneous and prone to selection bias. A pooled evaluation restricted to RWD, supporting what is called real-world evidence (RWE), has not previ-ously been performed.
METHODS: A critical methodological issue is the de-nominator used to report live birth rate (LBR): per embryo transfer (ET), per oocyte retrieval (OR), or per cycle initiat-ed (CI). We performed a systematic review of RWD studies published between January 2020 and January 2026 using PubMed, Scopus, Google Scholar, and the Latin American Network Registration System. Studies were eligible if at least one arm included ≥800 cycles and reported live birth outcomes comparing PGT-A and non-PGT-A cycles.
RESULTS: Seven studies met the inclusion criteria, com-prising 70,816 live births in PGT-A cycles and 63,557 in controls. Reported LBRs ranged from 24% to 44% in non-PGT-A cycles and from 17% to 53% in PGT-A cycles. All studies reporting LBR per ET favored PGT-A. In contrast, studies reporting LBR per CI or OR favored the control group. The direction of effect was therefore fully explained by denominator choice.
CONCLUSION: Real-world evidence does not demon-strate a consistent benefit of PGT-A, and the interpretation of effectiveness is strongly dependent on outcome denomi-nator selection. Therefore, the routine use of PGT-A should not be considered justified, and the medical literature has a duty to precisely define the subpopulations in which PGT-A might have value.