The negative effect of progesterone levels on the day of hCG administration (Phcg) on live-birth rate (LBR) is not restricted to values above 1.5 ng/ml in IVF/ICSI cycles

Abstract Research question Previous researches suggest a decreased probability of success in stimulated IVF cycles with premature rise of progesterone (PE) on HCG day (Phcg), but this remains a controversial issue. Design Retrospective, observational, single-centre cohort study conducted on last recent 5447 IVF/ICSI cycles performed on 2192 patients during 2009-2015 period, with conventional controlled ovarian stimulation (COS). Based on this large data base, we developed a non-linear mixed prognosis model of Livebirth (LB) incorporating Phcg as a predictor. Results In addition of known predictors (age, BMI, AMH, type of infertility), P hcg was associated to a linear effect (OR=.78 per Log(PE) ng/ml , 95%CI [.611-.997], p=.047) combined with a strong quadratic effect (OR=.585 per Log 2 (PE) ng/ml [.444,.775], p hcg. for nOO≤3, but the effect of P hcg remains modest , . For higher values of nOO, LB rapidly increases until 20 oocytes and more, however LB is more sensitive to P hcg . Higher LB prognoses occur for the optimal P hcg , but strongly reduced for both lower or higher P hcg values. Conclusions Our model provides evidence of an important negative impact of P hcg at both lower and higher values, independent of oocytes number, thus defining a proper range for Fresh Embryo Transfer or Freeze-All strategy. In poor responses, PE may be neglected, thus avoiding unnecessary cancellations or embryo freezing. Conversely, in the higher responses, the negative effect of PE appears more pronounced, suggesting that freeze-all policy should be applied more widely.