Predicting the duration of induction of labour in nulliparous women

This study examined if the modified Bishops score (MBS) at the start of an induction of labour (IOL) (VE-1) or if the MBS after pharmacological/mechanical ripening (VE-2) was the better predictor of the duration of induced labour and whether there was one component of the score that was most predictive of time to delivery. The measures at VE-2 were correlated more strongly with the time to birth, than VE-1 measures. At both VE-1 and VE-2, component measures (especially position of the cervix) showed weak correlation compared to composite measures. Omitting position from the composite score resulted in a simplified modified Bishops score (sMBS) that had the highest correlation coefficients. A model comprising sMBS and 5 clinical variables explained similar to 73% of the variance. The vaginal examination findings prior to an IOL do not impact how long the labour may take. A more favourable cervix similar to 12 h later (measured using a 4-item composite of dilatation, length, consistency and station) predicts a quicker induced labour.Impact statement What is already known? Induction of labour (IOL) is a common obstetric intervention in Australia. The IOL process can be a protracted and sometimes frustrating experience for women, and it may not result in a vaginal birth. A 'failed induction' or 'failure to progress' are relatively common indications for caesarean section (CS) in this setting where, despite many hours of an oxytocin infusion, the woman does not establish or progress in the active phase of labour. What does this study add? The measures at VE-2 were correlated more strongly with the time to birth, than VE-1 measures. At both VE-1 and VE-2, component measures (especially position of the cervix) showed weak correlation compared to composite measures. Omitting position from the composite score resulted in a simplified modified Bishops score (sMBS) that had the highest correlation coefficients. A model comprising sMBS and 5 clinical variables explained similar to 73% of the variance.