In total, 27 472 deliveries meeting our inclusion criteria were recorded between April 1st 2012 and March 31st, 2019. Of these, 15 752 women were managed expectantly, 9 712 women had a plan for induction of labour and 2008 had a no-labour caesarean section (Table 1). A total of 14 487 women had a spontaneous vaginal delivery, 3 439 women had an operative vaginal delivery, and 9 546 women had a no-labour caesarean section (Table 1 and Fig. 1). Clinical and demographic data are described in Table 1.
The mean age of women included in our cohort was 29 years. In our cohort, 17 032 (62.0%) women had class I obesity, 6 529 (23.8%) had class II obesity, 2 363 (8.6%) had class III obesity, 855 (3.1%) had class IV obesity and 693 (2.5%) had class V obesity. The overall rate of caesarean section in our study was 34.7%. Of women who were managed expectantly, 3 631 (23.1%) had a caesarean section. Of women who were induced, 3 907 (40.2%) had a caesarean section delivery.
Table 2 depicts the association between planned mode of delivery (expectant management, induction of labour and no-labour caesarean section) and adverse birth outcomes. Overall, no-labour caesarean section reduced the risk of adverse events by 41% (aRR 0.59, 95%CI [0.50, 0.70]). Similarly, the WAOS showed a trend towards improved outcomes with no-labour caesarean section compared to expectant management (beta − 0.96, 95%CI [-1.87, -0.06]). There was no statistically significant increase in risk of adverse outcomes when comparing induction of labour to expectant management (aRR 1.03, 95% CI [0.96, 1.10]).
In our study population, there were a total of 29 intrapartum or in-hospital newborn deaths, 12 of which occurred in the expectant delivery group and 17 in the induction of labour group. None were recorded in the no-labour caesarean section group. There was a 30% reduction in risk of adverse neonatal outcomes with no-labour caesarean section compared to expectant management (aRR 0.70, 95% CI [0.57, 0.87]) and this risk reduction was observed for all BMI classes (Fig. 2).
We observed a trend towards increased neonatal adverse events with induction of labour compared to expectant management, but this was not significant (aRR 1.10, 95% CI [0.99, 1.21]). The WAOS showed a statistically significant increase in neonatal adverse events with induction of labour compared to expectant management (beta 0.61, 95% CI [0.16, 1.06]). The increase in neonatal risk with induction of labour was most observed in women of obesity classes III, IV and V (Fig. 2) suggesting a small linear trend between increasing BMI and increasing neonatal adverse outcomes with induction of labour.
The observed decrease in neonatal adverse events was dependent on gestational age (Fig. 2). At 37 weeks of gestation, no-labour caesarean section and induction of labour were both strongly associated with an increase in adverse neonatal outcomes across all BMI categories. At 39 weeks, the relationship was inversed and both induction of labour and no-labour caesarean section resulted in decreased neonatal adverse events for women of BMI category I, II, IV and V when compared to expectant management.
There were no maternal deaths recorded within our cohort. Comparing induction of labour to expectant management, there was no statistically significant difference in the relative risk of uterine rupture, blood transfusion, unanticipated operative procedure, or maternal ICU admission. Induction of labour appeared to provide a protective effect against 3rd and 4th degree lacerations compared to expectant management (aRR 0.85, 95% CI [0.75, 0.97]). Overall, there was no statistically significant difference in adverse maternal outcomes when comparing induction of labour to expectant management (aRR 0.95, 95% CI [0.85, 1.05]). There was a statistically significant increase in the risk of unanticipated operative procedure when comparing no-labour caesarean section to expectant management (aRR 1.92, 95% CI [1.32, 2.77]), but overall, no-labour caesarean section appeared to protect against adverse maternal outcomes (aRR 0.46, 95% CI [0.35, 0.60]). Conversely, the WAOS score showed a trend towards increased maternal adverse events with no-labour caesarean section compared to expectant management, but this was not statistically significant (beta 0.24, 95%CI [-0.06, 0.54]).
Results and observed associations held true when the complete case analysis was compared to the analysis with multiple imputation (Table S1).