Intrapartum ultrasound in predicting labour outcome

Introduction: I sought to determine the acceptability and feasibility of intrapartum ultrasound (USS) and the predictive value of clinical and ultrasound parameters in the first stage of labour for intrapartum Caesarean delivery (ICD) based on single and repeat assessments. External validation of a single assessment labour prediction model and evaluation of a research based “Intrapartum app” was performed. Methods: A prospective observational study was carried out at Queen Charlotte’s and Chelsea Hospital between 24-42 weeks’ gestation to assess the acceptability and feasibility of intrapartum ultrasound. Pre-assessment and post-assessment modified validated acceptability questionnaires were evaluated. The predictive value of intrapartum ultrasound in the first stage of labour for Caesarean delivery was assessed in a prospective longitudinal observational cohort in nulliparous term (37-42 weeks) labouring women. Transabdominal ultrasound scan was performed to assess fetal Doppler, amniotic fluid, fetal head position and transperineal ultrasound for head-perineum distance (HPD), caput succedaneum and moulding of the fetal head. These assessments were repeated at the next digital vaginal examination (VE). Results: 119 women were recruited to assess the acceptability of intrapartum ultrasound, 104 completed both pre and post-assessment questionnaires. The negative experience score was higher for VE compared to USS pre (10 and 5, p<0.0001) and post assessment (8 and 4, p<0.0001). The feasibility of intrapartum USS was assessed with paired vaginal and USS assessments performed in 192 women. If a cervix measured ≤6cm on VE, there was a smaller difference between USS and VE measured cervical dilatation (bias -0.173, 95% limits of agreement -2.51 to 2.16cm reduced to -1.96 to 1.89cm). There was low agreement (bias -0.232, 95% limits of agreement -5.31 to 4.84 clock hours) on fetal head position and fair agreement for the presence of caput succedaneum (76%, p<0.05). Fetal head station and head perineum distance were negatively correlated (r=-0.57, p<0.0001). In the first stage prediction study, 270 patients were recruited, 269 patients were evaluable of whom 219 (81%) had repeat assessments. Intrapartum Caesarean delivery was required in 79 (29%) patients. On external validation, the new population performed less well (AUC 0.80 (95% CI 0.74 to 0.86) compared to that on which the 2015 single assessment model (AUC 0.85 (95% CI 0.678-1.000) was based. The length of labour was shorter for those patients predicted to be at “high” likelihood of vaginal delivery compared to those with a “medium” and “low” likelihood (log rank test, p<0.01) using the “Intrapartum app”. The main predictors at the first assessment were HPD (adjusted OR 1.08/mm, 95% CI 1.04-1.13, p<0.0001) and cervical dilatation (adjusted OR 0.78/cm, 0.66-0.92, p=0.0025). When considering repeat assessments, cervical dilatation change was the most important predictive variable (adjusted OR 0.10, 0.03-0.26, p