Ethical approval was obtained from the Medical Ethics Committee of Weihai Maternal and Child Health Hospital (WFEY-QR-CR-825, 3 January 2017). The written cognitive and approval consents were also signed by patients. Patients undergoing routine IVF treatment (the number of COCs acquired ≥10) at the Reproductive Medical Center of Weihai Second Municipal Hospital between Jun 2017 and Aug 2019 were treated as candidates for this study. The characteristics of patients were listed in Table 1, including age, body mass index (BMI), basal sex hormone levels, duration of infertility. Women with endometriosis, poor endometrium (<8 mm diameter), premature ovarian insufficiency on the hCG trigger day or the transfer day were excluded. Samples from their husbands were also excluded if they had severe asthenospermia/oligospermia and aspermia.
The follicles of women receiving gonadotropin releasing hormone agonist (GnRH-a) long protocol were monitored by ultrasound. When 10 or more follicles had reached a mean diameter of ≥14 mm, the women were given appropriate dose of hCG to induce oocyte meiotic maturation. Cycles with more than 10 COCs retrieved were assigned in this study and all COCs were allocated equally and non-selectively to either incubator.
To avoid frequent opening/closing of the incubator door, the ‘one patient one incubator’ strategy was conducted in our study, which means COCs or embryos from one patient were cultured separately in one incubator.
Incubators and parameters monitoring
As showed in Fig. 1A and 1B, the air jacket incubator (EC9 triple gas bench-top incubator, ASTEC CO., LTD. Japan) and a conventional water jacket incubator (Penguin AQ series/APM30D, triple gas incubator, ASTEC CO., LTD. Japan) were used in our research. The specifications of different incubators were listed in Fig. 1C. A range of 12-15 repeated opening/closing processes were conducted in a single ART cycle (from oocyte collection to blastocyst transfer).
For temperature monitoring, a handheld temperature measuring equipment with a long and soft linear sensor (TES 1310 TYPE-K, China) was used to monitor the variation of temperature in a center-well organ culture dish (FALCON, 353037) with 1 ml medium covered with 1 ml mineral oil in the dishes. Briefly, according to the length of time consuming in routine embryo culture, we made a single 10-seconds door opening/closing process, after which the temperature of incubator chambers was detected. It should be noted that 5-seconds door opening/closing was enough for air jacket incubator, in which only one dish was usually placed. Considering the consistency of this study, 10-seconds opening/closing treatment was accepted for two kinds of incubators. CO2 and O2 recovering times were recorded according to the corresponding display panels.
For pH measurement, 5 ml medium was poured into a tube and equilibrated for overnight. At the second day, we tested the initial pH values (initial state) by a pH meter (PB-10 Sartorius). As temperature monitoring, after a 3 min holding on the thermostatic desk, pH values were recorded again (out for 3 min), after which the medium was put back into incubators and detected at 10 min, 30 min and 1 hour (showed ‘in for 10 min’, ‘in for 30 min’ and ‘in for 1h’ respectively).
Sperm preparation, fertilization and embryo culture
After semen liquefaction (nearly 30 minutes), density gradient centrifugation combined with swim-up was used to sort sperm with normal morphology and high motility . G-IVF (vitrolife) was used to wash sperm and120,000 motile sperm/ml was used for short-time in vitro fertilization. After 4 hours co-culture, oocyte denudation was performed using mechanical method and the remaining sperm was also removed. Depending on the presence of the second polar body, we judged if oocytes were fertilized and only these oocytes with two polar bodies were then transferred into cleavage culture medium (G1 medium, vitrollife). At day 1 (16-18 hours after fertilization), the number of pronucleus (PN) was recorded and 2PN-gametes were identified as normal fertilization, after which these gametes were transferred into new G1 medium. At day 3 and day 5, embryos were transferred into G2 medium (vitrolife) for blastocyst culture.
Embryo and blastocyst scoring
Embryos and blastocysts were graded according to the Istanbul consensus and Gardner criteria [8, 9]. Briefly, embryos (day 3) with 7-9 cells, less than 10% fragmentation by volume and symmetric blastomeres were identified the good. Blastocysts (day 5) graded 4BB or even better were identified the good, including 4BB, 4AB, 4BA and 4AA. Embryos on day 3 or blastocysts on day 5 were assessed by three experienced embryologists and the assessments were recorded individually. Although most of the results were consistent between embryologists, the lowest score (when exist) was accepted.
Embryo transfer and clinical follow-up
Using abdominal ultrasound guidance, one or two embryos (fresh or frozen-thawed embryos) were transferred to each woman. In some situations, such as the patient with ovarian hyper-stimulation syndrome (OHSS), embryos were cryopreserved and the frozen-thawed embryos were transferred later. Serum β-hCG levels were monitored on day 14 after embryo transfer, which was used to confirm biochemical pregnancy. When the gestational sac (should have heartbeat) was observed using ultrasound one month after embryo transfer, clinical pregnancy was confirmed. Considering the possibility of failure in one cycle, we calculated the successful rate of every transfer cycles to compare the clinical outcomes in two groups. For example, if one patient was conducted two times of frozen-thawed embryo transfer (FET) (all embryos from the air jacket incubator) and was verified pregnant at last, the clinical pregnancy rate would be 50%.
Statistical analysis was performed using Student’s t-test or Fisher’s exact test with GraphPad Prism 7 software. Data were expressed as mean ± SEM. As for comparing the proportion of pregnancies, dichotomous outcomes data were showed as frequency and percentage. The differences between two groups were represented by computing the odds ratio with 95% confidence interval, and Fisher’s exact tests was used. P<0.05 was considered statistically significant.