Abstract Study question Does the timing of intracytoplasmic sperm injection (ICSI) relative to ovulation trigger impact live birth rates and neonatal outcomes? Summary answer ICSI timing relative to ovulation trigger impacts neither live birth rates nor neonatal outcomes. What is known already When conducting ICSI, the timing of the insemination relative to ovulation trigger is difficult to standardise in clinical practice, especially in busy ART laboratories. Flexibility when scheduling ICSI procedures would therefore be beneficial from a laboratory management perspective. However, the impact of this timing on live birth rates and neonatal outcomes remains inconclusive in the literature. Study design, size, duration This retrospective study included a total of 1547 consecutive single autologous embryo transfer cycles (maternal age 35.2 ± 4.8 years) performed at Fertility North between January 2017 and August 2021. ICSI was performed immediately after denudation without timing restrictions relative to ovulation trigger. A total of 422 (27.3%) transfers resulted in a live birth, including 415 singletons, 5 twin deliveries, and 2 lost to follow-up. Only singleton babies were included for neonatal outcome analysis. Participants/materials, setting, methods ICSI timings relative to ovulation induction trigger (mean ± SD 42.5 ± 1.4 hours, range 37.0-46.5 hours) were extracted from the electronic witnessing database. Cycles were grouped according to ICSI timing quartiles (groups Q1-4: <41.7, 41.7-42.7, 42.8-43.4 and 43.5+ hours). Logistic/linear regressions were used to assess impacts of ICSI timing on the subsequent live birth and birthweight outcomes in the singleton babies. Mixed effect model was used to account for clustering effect from repeat cycles by same patients. Main results and the role of chance Q1 group had significantly lower live birth rate (22.7%) but higher maternal age (37.3 ± 5.2 years) in comparison to Q2 (28.5% and 34.5 ± 4.6, P < 0.05 respectively), Q3 (30.2% and 34.8 ± 4.5, P < 0.05 respectively) and Q4 groups (27.7% and 34.4 ± 4.4, P < 0.05 respectively). However, following multivariate logistic regression adjusting for potential confounders including maternal age, body mass index, aetiology, sperm origin (ejaculated or surgical), sperm type (donor/partner), embryo stage at transfer (cleavage or blastocyst), and fresh/frozen transfer; observed differences in live birth rates became no longer significant (Q2 vs Q1, aOR = 0.982, 0.692-1.394, P = 0.919; Q3 vs Q1, aOR = 1.058, 0.884-1.266, P = 0.538; Q4 vs Q1, aOR = 0.969, 0.859-1.094, P = 0.615; Q3 vs Q2, aOR = 1.153, 0.830-1.600, P = 0.396; Q4 vs Q2, aOR = 0.963, 0.818-1.134, P = 0.649; Q4 vs Q3, aOR = 0.836, 0.601-1.164, P = 0.289). Gestational age and birthweight of singletons arising from 4 groups were also comparable (Q1 38.1 ± 2.2 weeks and 3209.1 ± 581.4 grams, Q2 38.2 ± 1.6 weeks and 3159.9 ± 508.2 grams, Q3 38.7 ± 1.8 weeks and 3304.7 ± 546.9 grams, and Q4 38.4 ± 2.0 weeks and 3297.5 ± 526.9 grams; P > 0.05 respectively). Multivariate linear regression further confirmed the absence of significant impacts (standardised coefficient = 0.001, P = 0.980) by ICSI timing on the birthweight of resulting newborns, taking into account gestational age and baby sex plus all above mentioned 7 potential confounders. Limitations, reasons for caution Our study is limited by its retrospective nature, where it is impossible to control all known and unknown confounding factors. The varied denudation duration as a result of different egg numbers was unable to be analysed, although ICSI was aimed to be performed immediately after completion of denudation. Wider implications of the findings Our findings indicate a wide time window in human oocytes to initiate fertilization via ICSI. This data offers reassurance in applying flexible ICSI timing without impacting live birth rates and neonatal outcomes. This is clinically valuable by providing evidence-based guidance to daily practice and patient counselling. Trial registration number not applicable
Read full abstract