[Fetal pulse oximetry: clinical practice]

To assess in current practice the application of our protocol of using fetal pulse oximetry during labor, to evaluate whether fetal scalp blood sampling can be reduced and to determinate reliability of fetal pulse oximetry on the prediction of poor neonatal outcomes.

Prospective observational unicenter cohort including 449 patients during two years. All pregnancies were singleton, greater than or equal to 37 weeks' gestation, cephalic presentation, and had non reassuring fetal heart rate. The poor neonatal outcome was defined by one of the followings: arterial umbilical cord pH

The use of fetal pulse oximetry was concordant with our protocol in more than 80% of cases. The frequency of fetal scalp blood sampling was significantly reduced from 9.9 to 8.6% after the introduction of our protocol. With a 30% threshold, diagnostical values of fetal oximetry for a poor neonatal outcome were 9.1% for sensitivity, 93.1% for specificity, 79.4% for negative predictive value and 25.9% for positive predictive value. With a 40% threshold, the diagnostic values were 74, 51.6, 88.2 and 28.9% respectively.

The strict application of our protocol allow a less aggressive management of labor with a significant decrease in fetal scalp blood sampling. This study shows that with a 40% threshold, fetal pulse oximetry could be considered as a reliable tool for the management of labor with no increase of poor neonatal outcomes. On the other hand, the threshold which could determine whether an abnormal fetal heart rate needs immediate intervention still remains unclear.