ABSTRACT
OBJECTIVES: To identify risk factors for Cesarean delivery and non-reassuring fetal heart tracing (NRFHT) in pregnancies with a small-for-gestational-age (SGA) fetus undergoing induction of labor and to design and validate a prediction model, combining antenatal and intrapartum variables known at the time of labor induction, to identify pregnancies at increased risk of Cesarean delivery. METHODS: This was a retrospective cohort study of non-anomalous, singleton gestations with a SGA fetus that underwent induction of labor, delivered in a single tertiary referral center between January 2011 and December 2016. SGA was defined as estimated fetal weight (EFW) < 10th percentile. The primary outcome was to identify risk factors associated with Cesarean delivery. The secondary outcome was to identify risk factors associated with NRFHT. Univariate and multivariate analyses were used to determine which clinical characteristics, available at the time of admission, had the strongest association with Cesarean delivery and NRFHT during labor induction. The predictive value of the final models was assessed by the area under the receiver-operating-characteristics curve (AUC). Sensitivity and specificity of the models were also assessed. Internal validation of the models was performed using 10 000 bootstrap replicates of the original cohort. The adequacy of the models was evaluated using the Hosmer-Lemeshow goodness-of-fit test. RESULTS: A total of 594 pregnancies were included. Cesarean delivery was performed in 243 (40.9%) pregnancies. Significant risk factors associated with Cesarean delivery, and included in the final model, were maternal age, gestational age at delivery and initial method of labor induction. The bootstrap estimate of the AUC of the final prediction model for Cesarean delivery was 0.82 (95% CI, 0.78-0.86). The model had sensitivity of 64.2%, specificity of 86.9%, positive likelihood ratio (LR) of 4.9 and negative LR of 0.41. The model had good fit (P = 0.617). NRFHT complicated 117 (19.7%) pregnancies. Significant risk factors for NRFHT included EFW < 5th percentile, abnormal umbilical artery Doppler studies (pulsatility index > 95th percentile or absent/reversed end-diastolic flow) and gestational age at delivery. The final prediction model for NRFHT had an AUC of 0.69 (95% CI, 0.63-0.75) and specificity of 97.0%. CONCLUSION: We identified several significant risk factors for Cesarean delivery and NRFHT among SGA pregnancies undergoing induction of labor. Clinicians may use these risk factors to guide patient counseling and to help anticipate the potential need for operative delivery. Copyright © 2019 ISUOG. Published by John Wiley & Sons Ltd.
