Debate continues over which morphological parameter is most important in selecting blastocysts for transfer. We aimed to investigate which parameter more accurately predicts the occurrence of a live birth by designing a retrospective cohort study of 1084 fresh elective single blastocyst transfers. Primary outcome was live birth rate (LBR) and secondary outcomes were implantation, clinical pregnancy and early pregnancy loss rates. Blastocyst expansion and inner cell mass (ICM), but not trophoectoderm, were associated with LBR in the definitive multivariable regression analysis. When ICM grade dropped from A to C the likelihood of achieving a live birth was reduced by 55% (OR= 0.45, 95% CI 0.26-0.79, p = .005). These results were similar for clinical pregnancy rates. Early pregnancy loss rates of embryos with ICM grade C were more than double (38.0%) compared to those of grades A (15.95%) and B (17.17%, p = .002). The transfer of an embryo with an optimal inner cell mass reduces early pregnancy loss and increases the likelihood of a live birth. We did not find any significant association between trophectoderm and LBR in the multivariable analysis in contrast with recent studies.

Antral follicle count (AFC) is a reliable predictor of ovarian response to stimulation and its inter-cycle and inter-observer variability has been extensively studied on in vitro fertilization (IVF), mostly in highly selected populations within studies not originally designed for this purpose. In this retrospective cohort study, we assess the inter-cycle variation of AFC in a setting similar to that of the daily practice. We included only patients undergoing mild stimulation for intrauterine insemination (IUI). One hundred and forty-eight patients had two (62 patients, group A), three (49 patients, group B) or four (37 patients, group C) IUI cycles and AFC was measured on early follicular phase of each cycle by one of the members of the medical team within daily practice. Intra-class correlation coefficients were used to estimate variability. Inter-cycle variability rendered ICCs above 0.70 in all groups improving along with the number of cycles [Group A ICC 0.78 (95%CI 0.66-0.86), Group B ICC 0.87 (95%CI 0.80-0.92) and Group C ICC 0.91 (95%CI 0.85-0.95)]. Inter-observer variability showed a high degree of concordance with ICCs above 0.95. We provide the closest approximation to real inter-cycle and inter-observer AFC variability expected in routine clinical practice.

No abstract