Abstract
OBJECTIVES: This study aimed to investigate the correlation of ovarian sensitivity index (OSI) and clinical parameters in IVF treatments. METHODS: IVF data files between January 2011 and December 2020 in a single unit were included. The primary outcome measure was the correlation between the OSI and clinical pregnancy and live birth rates. A generalized linear model was employed to assess group differences while controlling for age. Correlations between the OSI and clinical parameters were analyzed using Pearson's correlation test. RESULTS: In total, 1,627 patient data were reviewed, comprising 1,160 patients who received GnRH antagonists and 467 who received GnRH agonists. There was no difference in the incidence of premature ovulation and LH surge in women receiving either GnRH antagonists or agonists. A higher number of mature oocytes and good embryos were obtained in the GnRH agonist cycles. No differences were observed in pregnancy and live birth rates between both groups. Regarding the correlation of the OSI with clinical parameters, serum anti-Müllerian hormone, cycle day 2 follicle-stimulating hormone, LH, and estradiol concentrations, numbers of larger follicles, fertilization rate, and the incidence of premature LH surge were positively correlated with the OSI. Whereas the body mass index, mature oocytes obtained, embryo transfer number, and dose of GnRH antagonists were negatively correlated with the OSI. In the GnRH antagonists group, an OSI of 225.75 significantly distinguished pregnancy from non-pregnancy (p < 0.001), with an AUC of 0.615, and an OSI of 208.62 significantly distinguished live births from non-live births (p < 0.001), with an AUC of 0.637. As for the GnRH agonist group, an OSI of 228 significantly distinguished live births from non-live births, (p =0.020) with an AUC of 0.569. CONCLUSION: We demonstrated the capability of employing OSI to distinguish the clinical pregnancy and live birth outcomes in IVF cycles.