Intrapartum nonreassuring fetal heart rate tracing and prediction of adverse outcomes: interobserver variability.

Abstract

We determined interobserver variability in the classification of fetal heart rate (FHR) tracing with periodic deceleration as being reassuring or nonreassuring and in the ability to predict emergency cesarean delivery (ECD) or umbilical arterial pH < 7.00. Five clinicians reviewed 100 FHR tracings 1 hour before abnormalities and, if applicable, the hour before delivery. We calculated weighted Kappa coefficients (WKC) to assess interobserver variability and likelihood ratio of FHR tracing to identify ECD and low pH. Among 100 parturients, 46% of the women had ECD, and 2% of the women had low pH. The WKC for the classification of the FHR tracing as reassuring or nonreassuring in early labor was -0.12 and before delivery was 0.15. The WKC for ECD was 0.26 and for low pH was 0.21. The likelihood ratio for these 2 outcomes was < 2.0. There was poor agreement among the clinicians who classified FHR as reassuring vs nonreassuring; they could not identify which parturient would have ECD or a newborn infant with low pH.