Objectives: To evaluate whether preimplantation genetic testing for aneuploidy (PGT-A) can improve pregnancy and neonatal outcomes for patients with limited good-quality embryos. Design: Retrospective cohort study. Setting: University hospital. Population: A total of 1,553 patients who intended to pursue PGT-A for the first time but obtained only two or less good-quality embryos on day 3 after oocyte retrieval were divided into two groups: 997 in the PGT-A group and 556 in the drop-out group of withdrawing PGT-A due to poor embryological outcome. Results: After adjusting for potential confounding factors, PGT-A group exhibited significantly lower cumulative rates of biochemical pregnancy (19.96% vs. 30.22%, P-adj < 0.001), clinical pregnancy (17.55% vs. 23.38%, P-adj < 0.001) and live birth (14.14% vs. 16.19%, P-adj = 0.005) per oocyte retrieval and longer median time to pregnancy and live birth compared with drop-out group. However, significant increases in rates of biochemical pregnancy (72.16% vs. 35.50%, P-adj < 0.001), clinical pregnancy (61.86% vs. 26.98%, P-adj < 0.001), and live birth (48.45 vs. 18.26%, P-adj < 0.001) per transfer were found in the PGT-A group. No significant differences were observed in cumulative miscarriage and ectopic pregnancy rates, number of ETs needed per live birth and neonatal outcomes. Conclusion: PGT-A failed to improve cumulative live birth rate or shorten time to pregnancy, but optimized pregnancy outcomes per transfer for patients with limited good-quality embryos. Keywords: preimplantation genetic testing for aneuploidy, cumulative live birth rate, live birth rate per transfer, neonatal outcomes

Objective To explore whether a previous late miscarriage(LM) has a prognostic impact on the subsequent pregnancy outcomes in in-vitro fertilization(IVF) women. Design Retrospective cohort study. Setting Reproductive medicine centre at a tertiary hospital Population 1072 infertile women who had a LM following first embryo transfer. Methods Women were grouped by the causes of LM, 458 women with unexplained factor(unLM), 146 women with fetal factor(feLM), 412 women with cervical factor(ceLM), 56 women with trauma factor(trLM). Subgroup analysis and binary logistic regression were performed to evaluate the associations between LMs with different causes and subsequent pregnancy outcomes. Main outcome measures The frequency of live birth and miscarriage following the subsequent embryo transfer after LM. Results Compared with general IVF population, the early miscarriage rate was significantly higher in unLM group(8.28% vs. 13.47%,P=0.003, rate ratio(RR)=1.725,95% confidence interval(CI)1.204-2.470). Further, women with a unLM or ceLM had a dramatically elevated risk of recurrent LM(for unLM:4.24% vs. 9.43%,P=0.000,RR=2.348,95%CI 1.527-3.610; for ceLM: 4.24% vs. 15.53%,P=0.000,RR=4.147,95%CI 2.836-6.064) and a consequently reduced frequency of live birth(for unLM: 49.96% vs. 43.01%,P=0.005, RR=0.756,95%CI 0.622-0.918; for ceLM: 49.96% vs. 38.59%,P=0.000,RR=0.629,95%CI 0.512-0.774). Conclusions Only one previous LM resulted from unexplained factor or cervical factor was significantly associated with a higher risk of miscarriage and a lower live birth rate after the subsequent embryo transfer. Keywords Late Miscarriage/Live Birth/Early Miscarriage/Cervical Incompetence/In Vitro Fertilization Tweetable abstract Women who had a LM for unexplained factor or cervical factor following the first embryo transfer had a significantly poor pregnancy prognosis after the subsequent transfer.