A risk model approach to the prediction of fetal acidemia.

Known risk factors for adverse perinatal outcome were related to severe fetal acidemia during labor and at birth. Severe fetal acidemia was defined as a fetal scalp or umbilical artery pH > 3 SD below the mean for normal pregnancy. pH values were available for 1,524 singleton births without congenital anomalies. Of these births, 83 (5.4%) had at least one pH value > 3 SD below the reference mean. Known risk factors in the group with severe acidemia were compared with those in a random sample of the remaining 94.6% births. The strongest predictor of severe fetal acidemia was an abnormal cardiotocogram with a Hammacher score > 8. Risk factors shown to lower pH values significantly were combined in a prediction model based on multiple regression analysis. Back-validation of the model showed that half of all cases of severe fetal acidemia could not be adequately predicted. It is concluded that such prediction models do not contribute to a more effective detection of severe fetal acidemia.