MULTICENTER STUDY ON THE CLINICAL-VALUE OF FETAL PULSE OXIMETRY .2. COMPARED PREDICTIVE VALUES OF PULSE OXIMETRY AND FETAL BLOOD ANALYSIS

OBJECTIVE: Our purpose was to compare the predictive value of intrapar tum fetal pulse oximetry with that of fetal blood analysis for an abno rmal neonatal outcome in case of an abnormal fetal heart rate. STUDY D ESIGN: A prospective multicenter observational study was conducted fro m June 1994 to November 1995. Fetal oxygen saturation was continuously recorded with a Nellcor N-400 fetal pulse oximeter in case of an abno rmal fetal heart rate during labor. Simultaneous readings of fetal oxy gen saturation and fetal blood analysis obtained before birth (i.e., e ither al full dilatation or before cesarean section when indicated) we re compared with the neonatal status. The criteria for an abnormal neo natal outcome were (1) an umbilical arterial blood pH less than or equ al to 7.15 and (2) a combined Variable including 5-minute Apgar score less than or equal to 7, umbilical arterial pH less than or equal to 7 .15, secondary respiratory distress, transfer in a neonatal care unit, or neonatal death. RESULTS: At a 7.20 threshold for fetal scalp pH an d 30% for fetal oxygen saturation (i.e., the 10th percentile in the st udy population), the predictive Value of fetal pulse oximetry was simi lar to that of fetal blood analysis for an arterial umbilical pH less than or equal to 7.15 and for an abnormal neonatal outcome (positive p redictive value 56% vs 55%, negative predictive value 81% vs 82%, sens itivity 29% vs 35%, and specificity 93% vs 91%, respectively). The rec eiver-operator characteristic curve showed similar performance of eith er technique for cutoff values less than or equal to 7.20 for fetal bl ood pH and less than or equal to 30% for fetal oxygen saturation, wher eas fetal pulse oximetry became superior at higher thresholds. CONCLUS ION: The predictive Value of intrapartum fetal pulse oximetry can be f avorably compared with that of fetal blood analysis. Randomized contro lled management trials can now be performed to assess potential clinic al benefits of this new tool.