Statistical analysis
Comparisons among all the four groups or between any two groups were
performed using the Ļ2 test (Pearson Chi-square for cases where none of
the cells has expected counts less than 5; Likelihood Ratio for cases
with one or more cells had expected counts less than 5).
P<0.05 was considered statistically significant for
comparisons among all four groups and P<0.008 was considered
statistically significant for comparisons between any two groups, where
Bonferroni-corrected
p-value was used for multiple pairwise comparisons (the alpha-level was
divided by the number of pairwise
comparisons, the number of pairwise comparisons for all four groups was
6, i.e., equal to
0.05/6)4.
In addition, comparisons between the double cleavage-stage embryo
transfer group (C-2) and the single blastocyst-stage embryo transfer
group (B-1) were performed. First, unadjusted logistic regressions were
performed to calculate unadjusted odds ratios (ORs) and 95%
confidenceĀ intervals (CIs) of B-1 VS C-2 for each maternal or neonatal
outcome. Then, multivariable logistic regressions were performed to
calculate adjusted odds ratios (aORs) and 95% CIs, which included
maternal age, BMI, infertility type, cause of infertility, duration of
infertility, type of fertilization in the current cycle, and previous
ART cycles. Furthermore, multivariable logistic regression models were
used to predict absolute risks (probabilities) for each maternal or
neonatal outcome for each maternal age group. This was graphically
presented to illustrate age-outcome relationships.
All analyses were performed using IBM SPSS Statistics 25.0 (Armonk, NY:
IBM Corp.).