Prediction of fetal distress

Abstract

Introduction: Identification of the increased risk of intrapartum fetal compromise prior to labour is an obstetric challenge. Globally, intrapartum hypoxia remains a major contributor to stillbirth, hypoxic ischemic encephalopathy and cerebral palsy. For parents and families, the psychosocial and financial impact ($1.5 billion in Australia) of these complications are profound and long-lasting. The majority of these catastrophic events occur despite a lack of obvious risk factors.Objectives: To explore the relationship between feto-placental and feto-maternal Doppler ultrasound variables and maternal serum placental growth factor (PlGF) concentration with adverse perinatal outcomes (caesarean for intrapartum fetal compromise, CS-IFC; operative delivery of any kind for intrapartum fetal compromise, Op-IFC; or composite neonatal outcome, CNO) in low risk pregnancies from 36 weeks’ gestation. Additionally, to determine the screening performance of these variables.Methods: A prospective, observational, panel study of low risk women was conducted at Mater Mothers’ Hospital, Brisbane. Women were eligible to participate if they were normotensive with an appropriately grown singleton fetus, uncomplicated pregnancy and anticipating a vaginal delivery beyond 36 weeks’ gestation. Women underwent fortnightly assessment of various feto-placental and feto-maternal Doppler parameters, estimated fetal weight and maternal PlGF quantification. Maternal characteristics, ultrasound and PlGF data, intrapartum and neonatal outcomes were recorded. For each assessment and for the last assessment preceding delivery, ultrasound and PlGF variables were assessed for their distribution and association with outcomes. Additionally, the rate of change in the individual variables was tested for association with outcomes for the different gestation intervals at assessment. Finally, the screening performance of ultrasound and PlGF parameters were tested in isolation and combination for raw and gestation-adjusted (regression and centiles) variables.Results: From May 2014 to September 2016, 483 women participated in the study, with exclusions predominantly for missing data and planned cesarean delivery. For the Doppler parameters, the following associations were demonstrated: the umbilical artery pulsatility index (PI) was higher at 40 weeks’ gestation for babies in the adverse Op-IFC outcome group; the middle cerebral artery PI was lower in the adverse outcome groups for CS-IFC (40 weeks’ gestation), Op-IFC (36, 38, 40 weeks’ gestation and last assessment) and CNO (last assessment) groups; the cerebroplacental ratio was lower in the adverse outcome groups for CS-IFC (40 weeks’ gestation and last assessment), Op-IFC (38 and 40 weeks’ gestation and last assessment) and CNO (40 weeks’ gestation and last assessment); weight-adjusted umbilical venous flow rate was lower in the adverse outcomes groups for CS-IFC (38 weeks’ gestation), Op-IFC (38 weeks’ gestation and last assessment) and CNO (last assessment); the uterine artery PI was higher in the adverse outcomes groups for CS- IFC (40 weeks’ gestation and last assessment), Op-IFC (38 weeks’ gestation and last assessment) and CNO (38 weeks’ gestation and last assessment); estimated fetal weight centile was lower in the adverse outcomes group for Op-IFC (38 weeks’ gestation and last assessment). The rate of change in the cerebroplacental ratio, weight-adjusted umbilical venous flow rate and uterine artery PI was different between those with and without the CS-IFC, Op-IFC and CNO outcomes.Maternal PlGF was lower at last assessment preceding delivery amongst those with Op-IFC outcome and all assessments for the CNO outcome. However, sub-analysis of women who had PlGF assay performed on the Alere Triage platform demonstrated that maternal PlGF levels were lower at 38 and 40 weeks’ gestation and at last assessment preceding delivery. The rate of change in maternal PlGF, irrespective of assay platform, was no different between the outcome groups.Screening performance, as determined by the area under the receiver-operator characteristic (AUROC) curves, was highest for CS-IFC when gestation-adjusted centiles for the cerebroplacental ratio (or middle cerebral artery PI), uterine artery PI and PlGF levels (Alere platform) were used in combination (AUROC curve = 0.94 and 0.95, respectively). In screening for Op-IFC, the highest AUROC curve (0.76) was produced by a combination of the middle cerebral artery PI, weightadjusted umbilical venous flow rate, uterine artery PI and nulliparity that were not adjusted for gestation at assessment. In screening for CNO, the equal highest AUROC curve (0.67) was produced by either a combination of the cerebroplacental ratio, uterine artery PI and gestation at delivery, unadjusted for gestation at assessment, or a combination of gestation-adjusted (centile) for PlGF (DELFIA platform), uterine artery PI and gestation at delivery.Conclusion: Various ultrasound parameters and lower maternal PlGF levels are associated with CSIFC, Op-IFC and CNO. The rate of change in these parameters did not differ between the outcome groups. The combination of predictors (ultrasound and maternal PlGF) for CS-IFC and Op-IFC provided significant improvement in screening performance compared to their use in isolation and may provide benefit as part of a screening algorithm. Significantly, these results are an improvement on current intrapartum fetal heart rate monitoring.