Computerised Interpretation of Fetal Heart Rate During Labour (INFANT): A Randomised Controlled Trial

INFANT is a decision-support software that assesses fetal heart rate quality and contraction patterns alongside the Guardian electronic information capture system. INFANT color codes fetal heart rate patterns based on abnormalities (mild, moderate, severe), but does not make recommendations for clinical actions. This randomized controlled trial aimed to find whether the rate of poor neonatal outcomes would be affected by the assistance of decision-support software in the interpretation of cardiotocographs. The study took place at 24 sites in the United Kingdom and Ireland and included pregnant women 16 years or older, with a singleton or twin pregnancy, at 35 weeks' gestation or more, with no known gross fetal abnormality who were judged to require continuous electronic fetal heart ratemonitoring (EFM) on the basis of their current practice. Participants were randomly assigned to either receive decision support using INFANT or no decision support. One primary outcome was poor neonatal outcome, which included intrapartum stillbirth or early neonatal death (excluding lethal congenital anomalies), neonatal encephalopathy, admission to the neonatal unit within 24 hours for 48 hours or more with evidence of respiratory illness, feeding difficulties, or encephalopathy with evidence of compromise at birth. Another primary outcome, within a subset of surviving children, was assessment of neurodevelopmental outcome at 2 years of age. Secondary outcomes included Apgar scores, duration of hospital stay, the need for neonatal resuscitation, and metabolic acidosis of cord blood samples, among others. Of 47,062 recruited to the study between 2010 and 2013, 1020 women were excluded from analysis of the primary outcome. For the longer-term outcomes, follow-up data at 2 years were complete for 6707 children. Between the 2 groups, incidence of poor neonatal outcome did not differ significantly. In the decision-support group, 172 (0.07%) of 22,987 babies experienced poor neonatal outcome, and in the no-decision-support group, 171 (0.07%) of 23,055 babies experienced poor neonatal outcome (adjusted risk ratio, 1.01; 95% confidence interval, 0.82-1.25). No differences were noted between the groups in terms of components of the primary out-come or any of the secondary outcomes. In addition, no differences were noted in 2-year outcomes between the 2 groups. When a subset of the primary outcome cases was reviewed, it was found that a suboptimal response to the fetal heart rate abnormalities occurred 38% of the time, 14 of 35 in the study group and 13 of 36 in the control group. There was no evidence supporting the hypothesis that decision-support software would affect the incidence of suboptimal care. The authors conclude that the opportunities for improving neonatal outcomes likely lie in guidance of how to clinically respond to fetal heart rate tracings as opposed to the interpretation of the fetal heart rate tracings themselves.