Mode of Delivery and Neonatal Outcomes in Preterm, Small-for-Gestational-Age Newborns

Werner, Erika F.; Savitz, David A.; Janevic, Teresa M.; Ehsanipoor, Robert M.; Thung, Stephen F.; Funai, Edmund F.; Lipkind, Heather S.

Obstetrical & Gynecological Survey:
doi: 10.1097/01.ogx.0000427628.88134.e3

More than 50% of neonates born before 34 weeks of gestation in the United States are delivered by cesarean section (CS). When intrauterine growth restriction is present or the neonate is small for gestational age (SGA), vaginal delivery has been reported to be associated with increased odds of neonatal mortality. Given the increased risk of morbidity and mortality for SGA newborns and the relative impossibility of a conclusive randomized controlled trial, this study was performed to examine the association between route of delivery and neonatal outcomes, particularly neonatal death, intraventricular hemorrhage (IVH), and respiratory distress syndrome (RDS), in a large, diverse cohort of preterm SGA newborns.

Data on 1,025,903 singleton live births from 1995 to 2003 were obtained from a data set linking birth certificate and hospital discharge data in New York City. The study population included only women delivering vertex-presenting, singleton neonates between 25 and 34 6/7 weeks of gestation (n = 31,135). Of 23,144 neonates meeting the inclusion criteria, 2885 met the definition for SGA at delivery. Small for gestational age was used as a surrogate for intrauterine growth restriction. Because of the wide range of gestational ages, a subanalysis was done for those with a gestational age less than 30 weeks (n = 359). Another study population included all operative deliveries (n = 2927). Main outcomes included neonatal death before discharge, neonatal RDS, sepsis, IVH, seizure, subdural hemorrhage, or 5-minute Apgar score less than 7. Maternal characteristics included age, parity, race and ethnicity, level of education, insurance status, and prepregnancy weight. Covariates included diabetes, hypertension, and gestational age of delivery.

Of the 2885 neonates, 1214 (42.1%) were delivered vaginally, and 1671 (57.9%) were delivered by CS. All maternal and pregnancy characteristics differed significantly between the 2 groups (P < 0.05). Women who underwent CS were older, heavier, better educated, more likely to be white, and more often had diabetes or hypertension. The rate of CS rose in SGA newborns born at 34 weeks of gestation or less during the course of the study (1995–2003; 50% and 61% in 1995 and 2003). Neonates delivered by CS had a higher incidence of RDS and 5-minute Apgar scores of less than 7 compared with those delivered vaginally (odds ratio [OR], 1.49; 95% confidence interval [CI], 1.25–1.76; and OR 1.42; 95% CI, 1.07–1.87, respectively). After adjustment for maternal age, race, parity, education, insurance status, prepregnancy weight, diabetes, hypertension, and gestational age at delivery, CS was still associated with increased odds of RDS compared with vaginal delivery (adjusted OR, 1.32; 95% CI, 1.07–1.63) and tended to be associated with slightly elevated risks for all outcomes except IVH and neonatal death. When forceps and vacuum deliveries were included in the analysis, results were not altered significantly.

The fact that CS is associated with more RDS indicates that CS is not necessarily beneficial to SGA preterm neonates. The known risks and costs of CS might be justified if CS improves neonatal outcomes. However, these results show that CS compared with vaginal delivery was not associated with any statistically significant benefit for preterm SGA neonates. Further research is necessary to evaluate the effects of the rising CS rate in preterm SGA neonates. As with full-term newborns, vaginal delivery offers neonates an early respiratory advantage compared with CS.

Author Information

Departments of Gynecology and Obstetrics, Johns Hopkins University School of Medicine, Baltimore, MD; Ohio State University, Columbus, OH; Departments of Epidemiology and Obstetrics and Gynecology, Brown University, Providence, RI; Department of Epidemiology, University of Medicine and Dentistry of New Jersey School of Public Health, Piscataway, NJ; and Department of Obstetrics, Gynecology & Reproductive Sciences, Yale University School of Medicine, New Haven, CT

Article Outline
Back to Top | Article Outline


(There is no question that the dramatic rise in the cesarean delivery rate is a major concern in current health care. One third of infants are now born by cesarean delivery in the United States. The number of preterm infants delivered abdominally is even higher; more than half were delivered by cesarean as of 2005. In some cases, preterm infants are delivered by cesarean because of indications such as malpresentation. To some parents and providers, it seems logical that cesarean delivery would be less stressful for a very small and fragile fetus, although it is not at all clear in fact that this is the case. In fetuses that are preterm and/or small due to intrauterine growth restriction, there are conflicting reports as to whether cesarean delivery improves outcomes. Some data have indicated that the odds of IVH and neonatal death might be increased in the very-low-birth-weight infant born vaginally (Pediatrics 2006;118:e1836–e1844; Am J Obstet Gynecol. 1995;172(3):795–800), whereas others have not found this to be the case (Obstet Gynecol. 2008;112(1):21–28).

This issue is difficult to study, as the number of preterm deliveries of fetuses with intrauterine growth restriction is small. Doing a randomized trial to resolve the question would be nearly impossible even in a multicenter study. In this abstracted article, the authors used birth data for a nearly 10-year period from New York City. Obviously, this would allow inclusion of data on a very diverse population, socially, racially, and economically. Admittedly, a retrospective, nonrandomized study design presents some limitations, as the authors were not able to determine the indications for delivery, nor which infants were born because of indicated versus spontaneous preterm births, and these are all factors that may have major impact on outcomes. Nevertheless, the authors did a careful analysis controlling for potential confounders such as maternal age, race and ethnicity, education, insurance, and prepregnancy weight. Infants delivered by vacuum or forceps were initially excluded, but later included in the final analysis as these did not significantly alter any of the results.

There were nearly 3000 SGA infants delivered between 25 and 34 weeks’ gestation in the final analysis, of which fewer than half (42%) were delivered vaginally. Interestingly and importantly, there were no significant differences in IVH, subdural hemorrhage, seizures, or sepsis between the cesarean and vaginal delivery groups. And, as is also seen with full-term newborns, vaginal delivery had respiratory benefits, as demonstrated by a decreased risk of RDS in infants born vaginally (the adjusted OR for RDS in infants born by cesarean was 1.3).

The authors conclude that these findings point out the uncertainty of the premise that cesarean delivery provides a benefit for SGA infants. It is interesting that they are hesitant to state more strongly that these results indicate the fallacy of that premise and rather indicate a benefit of vaginal delivery. We should not hesitate to point out when well-designed studies refute findings of older literature that has become entrenched as common wisdom, which simply “seems logical.” Especially when associated with known harms, as cesarean deliveries certainly are for the mother, evidence of no benefit or additional harms to the infant are really important in making decisions regarding optimal care for our patients.—MEN)

© 2013 Lippincott Williams & Wilkins, Inc.