Umbilical cord arterial lactate compared with pH for predicting neonatal morbidity at term

To test the hypothesis that umbilical cord arterial lactate is superior to pH for predicting short-term neonatal morbidity at term.

We conducted a prospective cohort study of all consecutive, non-anomalous, singleton, vertex, term births from 2009 to 2012 at Washington University Medical Center. Umbilical arterial lactate and pH were measured immediately after delivery, before knowledge of neonatal outcomes. The primary outcome was a composite neonatal morbidity consisting of neonatal death, intubation, mechanical ventilation, meconium aspiration syndrome, hypoxic encephalopathy, and need for hypothermic therapy. The predictive ability of lactate and pH were compared using receiver operating characteristic curves. Optimal cutoff values of lactate and pH were estimated based on the maximal Youden index.

Of 4,997 term deliveries during the study period, 4,910 met inclusion criteria. The composite neonatal morbidity occurred in 56 neonates (1.1%). The mean lactate level was nearly twofold higher in neonates with the composite morbidity (6.49 compared with 3.26 mmol/L, P<.001), whereas mean pH values were less distinct (7.19 compared with 7.29, P<.001). Lactate was significantly more predictive of neonatal morbidity than pH (receiver operating characteristic curve area: 0.84 compared with 0.78, P=.03). The optimal cutoff value for predicting neonatal morbidity was 3.90 mmol/L for lactate and 7.25 for pH. Corresponding sensitivities and specificities were also higher for lactate (83.9% and 74.1% compared with 75.0% and 70.6%, respectively).

Results of this large prospective cohort study show that umbilical cord arterial lactate is a more discriminating measure of neonatal morbidity at term than pH.

: II.