No differences in CS rates were reported
in seven out of eight RCTs. In one study there was a
significantly lower CS rate in the induction group w39x. One
possible explanation for this finding was published four
years later w40x by the same authors in a secondary analysis
of data from the same RCT. In the expectant management
group 554 (32.4%) women were actually induced with a CS
rate of 33.6%, whereas in the expectant non-induced group
the CS rate was 20.1%. The CS rate was higher in nulliparous
women irrespective of allocation and was highest (42%)
among nulliparous women randomized to expectant management,
but required induction of labor. In contrast to the Canadian
study, the majority of RCTs reported a significant
reduction in CS for either abnormal intrapartum fetal heart
rate patterns or poor labor progress in both treatment and
control groups, when spontaneous labor ensued. The Canadian
multicenter PT pregnancy trial has been criticized for
the methodology and the conclusions w64x.