The Consortium on Safe Labor was a study conducted by the Eunice Kennedy Shriver
National Institute of Child Health and Human Development, National Institutes of Health and has been described in detail elsewhere.11
In brief, this was a retrospective cohort study involving 228,668 deliveries from 12 clinical centers and 19 hospitals representing 9 American Congress of Obstetrics and Gynecology (ACOG) districts. Women gave birth between 2002 and 2008. Institutional Review Board approval was obtained by all participating institutions. Women could have more than one pregnancy in the cohort, so to avoid intra-person correlation we only included the first pregnancy for a total of 206, 969 women. One site that did not include indications for induction was excluded. There were 16,910 deliveries between 34 0/7 and 36 6/7 weeks gestation. Of the late preterm births, 15,136 were singletons and 1,774 were multiple gestations (1,712 twin, 61 triplet, 1 quadruplet). We compared the 15,136 singleton gestations born late preterm to 170,593 singleton deliveries between 37 0/7 and 41 6/7 weeks gestation.
Demographic data, medical history, labor and delivery information as well as obstetrical, post partum and neonatal outcomes were extracted from patient electronic medical records from each institution. Data from the neonatal intensive care unit was collected and linked to the newborn record. Maternal and newborn discharge ICD-9 codes were also collected for each delivery. Data was transferred in electronic format from each site and was mapped to common categories for each pre-defined variable at the data coordinating center. Investigators at each site completed surveys on hospital and physician characteristics. Data inquiries, cleaning and logic checking were performed. Validation studies were performed for four key outcome diagnoses and the electronic medical records were found to be a highly accurate representation of the medical charts (> 95% for the majority of variables).11
We defined the following categories of precursors for late preterm delivery: “spontaneous labor”, “preterm PROM)”, “indicated” delivery and “unknown.” The admission reason and indication for induction or cesarean delivery were used to identify the precursors for delivery. We identified any other maternal, fetal, obstetrical or demographic variables associated with the outcome of the pregnancy. We classified the precursors for delivery using the following hierarchy. Women who presented in spontaneous labor were included only in the “spontaneous preterm birth” category even if they had other pregnancy complications (e.g. maternal diabetes). Women with premature rupture of the membranes and not in labor were included as “preterm PROM”. Thus, if a women presented with both preterm PROM and in spontaneous labor she was counted only once in the “spontaneous labor” category. If a woman did not present in spontaneous labor or with preterm PROM, we then identified all potential maternal, fetal or obstetrical complications of pregnancy, and included these in the “indicated” category. A woman could have more than one pregnancy condition in the “indicated” category. If there was no other reason, then we identified those women who were admitted to labor and delivery for an unspecified “fetal” or “maternal” reason and not in spontaneous labor or preterm PROM. These two categories were the only ones in the “indicated” category that were exclusive, and are specified as “Admission for fetal reason, not otherwise specified”, or “Admission for maternal reason, not otherwise specified”. The “unknown” category included elective inductions or cesarean sections as identified by the site with no other obstetrical, fetal or maternal conditions, as well as if there was no reason for induction or cesarean section provided and there were no other obstetrical, fetal or maternal conditions of the pregnancy.
We calculated the incidence of neonatal outcomes according to category of precursor of late preterm delivery stratified by the gestational age at delivery. Finally, we identified precursors for preterm delivery that could have been managed to deliver until 37 weeks of gestation and then potentially up to 40 weeks of gestation according to available evidence and expert opinion, and re-categorized these as “soft precursors”. Soft precursors included suspected macrosomia without maternal diabetes12
; uncomplicated gestational13
and chronic hypertension14
; and history of fetal, maternal or obstetrical complication in a prior pregnancy (e.g., history of preeclampsia in a previous pregnancy). We also identified “elective” induction or cesarean delivery as indicated by the site, with no other fetal, maternal or obstetrical complication. For late preterm birth, the soft precursors category came from both the original “indicated” and “unknown” categories. We did not include spontaneous labor or preterm PROM because those were likely the indications for the preterm delivery (i.e., soft indications without spontaneous or preterm PROM).
In order to examine how delaying delivery for women with a soft precursor would affect neonatal outcomes, we considered that a woman with a soft indication might either be delivered late preterm or expectantly managed until term. We compared neonatal morbidity and mortality between late preterm births for a soft indication and neonates born at 37 – 40 weeks of gestation with the same array of (soft) precursors. For pregnancies ≥ 37 weeks of gestation, a woman with a soft precursor who was being expectantly managed could have gone into spontaneous labor or had rupture of membranes. Thus, we included all women at term with a soft precursor at that gestational week of delivery for comparison (i.e., soft precursors with or without spontaneous or PROM). Stillbirths at term were also included in the comparison group.
We performed logistic regression to calculate the odds ratios for neonatal morbidities (oxygen use, transient tachypnea of the newborn, mechanical ventilation, respiratory distress syndrome, apnea, pneumonia or sepsis, and admission to the NICU) and perinatal mortality with delivery at 37, 38, 39, or 40 weeks gestation compared to delivery at 34–36 weeks of gestation (the referent group) in women with a soft or elective precursor for delivery. We combined 34–36 weeks gestation together as the referent group because our goal was not to show the difference between neonatal outcomes at 34, 35 and 36 weeks gestation, but rather to investigate whether neonatal outcomes improve with each advancing week of delivery beyond preterm. A backwards elimination was performed and variables considered included maternal race, parity, body mass index, type of insurance, and smoking. All variables were significant with a P-value < .05 and therefore were adjusted for in the final model. Chi-square, Cochran-Armitage test for trend, analysis of variance and Kruskal-Wallis tests were also performed where appropriate. Statistical analysis was performed using SAS (SAS Institute Inc., Cary, NC) version 9.1.