Table 3.
Effect of TNORCs on IVF/ICSI outcomes adjusted for the covariates women’s age, women’s tobacco smoking, and male infertility due to oligo-, astheno-, and/or teratozoospermia, cryptozoospermia or azoospermia, or idiopathic male infertility
| Explanatory variable | IVF/ICSI outcomes | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| n | No. of cycle cancellations before oocyte retrievalb | n | No. of cycle cancellations before embryo transferb | n | No. of clinical pregnanciesb | n | No. of clinical pregnancy lossesb | n | LBc | |
| TNORC | 1433 | P ≤ 0.0005a | 1433 | P ≤ 0.0005 | 1433 | P ≤ 0.0005 | 893 | P ≤ 0.0005 | 1433 | P ≤ 0.0005 |
| 1 | 691 | 0.00 (0.00–0.01)d | 691 | 0.16 (0.12–0.21) | 691 | 0.73 (0.67–0.78) | 493 | 0.03 (0.03–0.05) | 691 | 0.69 (0.63–0.74) |
| 1e | 1 | 1 | 1 | 1 | ||||||
| 2 | 398 | 0.12 (0.07–0.22) | 398 | 0.35 (0.27–0.46) | 398 | 0.74 (0.67–0.82) | 261 | 0.16 (0.14–0.19) | 398 | 0.62 (0.56–0.67) |
| 90.13 (14.69–552.89)f | 2.22 (1.95–2.54) | 1.03 (0.98–1.07) | 4.86 (3.59–6.58) | 0.74 (0.66–0.82) | ||||||
| 3 | 344 | 0.19 (0.10–0.37) | 344 | 0.49 (0.43–0.55) | 344 | 0.49 (0.43–0.56) | 139 | 0.44 (0.38–0.50) | 344 | 0.29 (0.25–0.34) |
| 137.84 (23.37–813.12) | 3.07 (2.28–4.13) | 0.67 (0.61–0.73) | 13.07 (9.50–17.99) | 0.19 (0.15–0.24) | ||||||
The covariates appearing in the GEE models are fixed at the following values: women’s age = 35.0349 years; number of cigarettes smoked by women per day for the 3 months before starting the first intended oocyte retrieval cycle = 3.59; oligo-, astheno-, and/or teratozoospermia = 0.44; cryptozoospermia or azoospermia = 0.12; idiopathic male infertility = 0.42
aThe adjusted/corrected Benjamini & Hochberg’s significance is P ≤ 0.05, based on five statistical tests applied
bCount variable that considers cumulative number of events per woman
cBinary variable that takes into account the occurrence = 1 or non-occurrence = 0 of a LB per woman
dValues are marginal means and 95% confidence intervals in parenthesis
eReference value
fValues are exponentiated regression coefficients and 95% confidence intervals in parenthesis. Note that exponentiated regression coefficients should be interpreted as rate ratios or odds ratios depending on whether a Poissonb or a logisticc regression is applied, respectively [20]