Sample size and statistical analysis
From a previous cohort of women with severe early PE (31) we estimated
the interval from admission to delivery to be 12 days in non-complicated
and 5 days in complicated pregnancies. Assuming this difference between
women in the highest quartile of angiogenic delta values and those in
the 1st-3rd quartiles, for an
alpha-risk of 5% and a power of 80%, 56 women would be needed
(log-rank test)(32).
Variables were checked for normal distribution by Kolmogorov-Smirnov
test. Paired comparisons of the angiogenic values between admission and
delivery were performed by non-parametric Wilcoxon sign test.
Comparisons between uncomplicated and complicated pregnancies were
performed by Student-T (assuming unequal variances), Mann-Whitney U,
Pearson Chi-squared and Fisher-F, as appropriate.
The likelihood of maternal complications was modelled by logistic
regression (with robust estimation of the standard errors). Levels of
angiogenic factors were log-transformed. The predictive performance was
determined by receiver–operating characteristic (ROC) curve analysis.
The interval from admission to delivery was graphed by Kaplan-Meier
procedure and tested for differences in the survival time by the
angiogenic delta values quartile using Log Rank (Mantel-Cox) test. The
interval to delivery was modelled by multivariate Cox regression
checking for the assumptions of proportionality and log-linear
relationship.
Statistical analyses and graph constructions were performed using STATA
13.0 (StataCorp LT, Texas, USA) and R 3.1.2 (The R Foundation for
Statistical Computing) [package “pROC”].