Study population and ethical approval
We retrospectively analyzed 1476 ICSI cycles from women, aged over 40 years old, at the Reproduction Medicine Center in the Sixth Affiliated Hospital of Sun Yat-sen University, between December 2013 and August 2020. To avoid bias in the analysis, the exclusion criteria in this study were: (a) the patients who used frozen-thawed oocytes or donor oocytes for fertilization; (b) women or partners with genetic diseases requiring preimplantation genetic testing (PGT); (c) males with obvious sperm morphology abnormality or with sperms from either testicular or microsurgical epididymal sperm aspiration. All patients signed written informed consent on the anonymous use of their data. This study was approved by the Institutional Reviewer Board of Sixth Affiliated Hospital of Sun Yat-sen University (2017ZSLYEC-016S). All methods were carried out in accordance with relevant guidelines and regulations.
Protocols for controlled ovarian stimulation (COS)
Our controlled ovarian stimulation (COS) protocols in this study are the routine protocols in our center, containing GnRH agonist long protocol, GnRH agonist short protocol, GnRH antagonist protocol, luteal phase ovarian stimulation, and progestin-primed ovarian stimulation (PPOS) protocol [13]. Ovarian stimulation was carried out using recombinant follicle-stimulating hormone (r-FSH) (Gonal-F, Serono, Sweden) combined with gonadotropin-releasing hormone (GnRH) agonist (Decapeptyl; Ferring, Kiel, Germany) and antagonist (Cetrotide; Merck Serono, Darmstadt, Germany). COS protocols and daily doses of FSH injection of the patients in our research were performed according to female ages, ovarian reserve, and various reactions to ovarian stimulation in previous cycles. Follicular developments were monitored with transvaginal ultrasound scanning. The Human Chorionic Gonadotropin (hCG) (Lizhu Pharmaceutical Trading Co., China) 4000-10000IU or triptorelin (Decapeptyl, Ferring pharmaceuticals, Germany) 0.1 mg and hCG (Lizhu Pharmaceutical Trading Co., China) 1000 IU were administered for ovulation triggering when one leading follicle reached a diameter of 18 mm or two leading follicles reached 17 mm. Normally, all accessible follicles above 10 mm were aspirated guided by transvaginal ultrasound when picking up oocytes. Then oocytes were cultured in a human tubal fluid (HTF) medium (Quinn’s Advantage fertilization HTF medium; Quinn’s, SAGE, USA), in 5–6% CO2 at 37°C. The cumulus-corona-oocyte complexes obtained were incubated until the moment of ICSI.
Timing control
Exact times including HCG trigger, OPU, and ICSI were recorded by the operators to establish a databank. We identified HCG trigger to OPU interval as T1, OPU to ICSI interval as T2, and HCG trigger to ICSI interval as T3. The detailed times, from HCG to ICSI, were analyzed to evaluate the influence of time intervals on ICSI outcomes.
Outcome measure
In our research, we regarded the clinical pregnancy rate as the primary outcome and the fertilization rate as the secondary outcome. Fertilization rate was defined as the number of fertilized embryos obtained per oocyte injected multiplied by 100 for ICSI [25]. Transferrable D3 embryo was defined as no less than 4 blastomeres on day 3, at least slightly uneven size of blastomere and < 20% fragmentation. Implantation rate was defined as the number of gestational-sac(s) divided by the number of embryo(s) transferred. Biochemical pregnancy was defined as a positive serum β-HCG 15 days after transfer. Clinical pregnancy was defined as a visible embryo with a fetal heartbeat 5 weeks after transfer.
Statistical analysis
Student-t test or Mann-Whitney U test was used for continuous variables. Results were expressed as mean ± SD. Chi-square was used for categoric variables. Correlation analysis, multivariate linear regression analysis, and logistic regression analysis were used to study the effect of intervals on embryological and clinical outcomes in ICSI cycles. All the data were analyzed with SPSS 26.0 (IBM, Armonk, NY, USA). P value less than 0.05 was defined as statistical significance. In addition, the graphs were prepared with GraphPad Prism version 5 (GraphPad Software, La Jolla, CA, USA).