Continued versus discontinued oxytocin stimulation in active labour and neonatal morbidity

Oxytocin is used for augmentation and induction of labour and is the most commonly used drug during labour.1,2 The stimulation requires a delicate balance between the wish for labour to progress and the serious risk of uterine hyperstimulation reducing placental blood perfusion and oxygen flow to the fetus.3 It is hypothesised that once the active phase of labour is established, there is sufficient production of endogenous hormones to maintain progression of labour, at which time one could discontinue the medication to reduce the risk of adverse effects.4 This hypothesis was tested in open-label, randomised trials which reported a reduced risk of caesarean section when oxytocin was discontinued in the active phase as compared with continuous use.4–6 The randomisation in these studies took place hours or even days before establishing the interventions, and a considerable number of participants had a caesarean section in the latent phase.7 When only data from women in the active phase of labour were analysed, discontinuation had little or no effect on caesarean delivery rates.7 Therefore, attention towards the neonatal consequences of the two interventions is necessary. The trial used an open-label design, and the women were randomly assigned between discontinuous oxytocin (infusion stopped at cervical dilation ≥6 cm) or continuous oxytocin until delivery. 2459 women were included, of whom 289 (11·8%) were excluded after randomisation due to ineligibility (n=92), withdrawn consent (n=3), or caesarean section before receiving the interventions (107 in the discontinued group vs 87 in the control group), leaving 2170 (88·2%) women in the modified intention-to-treat analysis. The STOPOXY8 and CONDISOX9 trials find that discontinued oxytocin in the active phase of labour does not lead to a reduction in neonatal complications or in caesarean deliveries, but data should be analysed as part of a meta-analysis or an individual patient data meta-analysis.