Despite ongoing improvements in perinatal care and neonatal survival, severe intraventricular hemorrhage and periventricular leukomalacia (PVL) continue to be major complications of extreme prematurity, associated with severe morbidity in the postnatal period. The pathophysiology of severe intraventricular hemorrhage and PVL is complex and multifactorial,1 and although the incidence of these outcomes has gradually decreased over the last few decades,1,2 the trends in rate reduction seems to have reached a plateau in recent years.3
Decreased gestational age is a known neonatal risk factor for severe intraventricular hemorrhage, PVL, and mortality. In addition, other factors like chorioamnionitis4 and preeclampsia5 have been associated with significant neonatal morbidity and mortality. The use of antenatal steroids is the only perinatal intervention that has been consistently shown to improve neonatal outcome and decrease the risk for severe intraventricular hemorrhage in premature infants.6
Mode of delivery, specifically cesarean delivery, has been postulated to have a theoretical advantage over vaginal delivery in premature infants. This benefit may be the result of the avoidance of prolonged labor, allowing a less traumatic birth.7 However, how to deliver a very preterm baby has been a controversial topic in the obstetric and neonatal community for decades, and delivery mode depends on the medical judgment of the obstetrician. Effective care cannot be based on meta-analysis of well designed randomized controlled trials because none of the attempts have come to conclusion.8,9 The question of what delivery mode is less harmful for the fetus is most relevant in those infants with borderline viability. Hysterotomy at early gestational age is technically more difficult, and is more likely to be complicated with injury to the bladder and uterine artery10 and an overall increased risk for the mother and the unborn child in their medical future.11–13
In breech presentation, the cesarean mode of delivery has been supported by the available evidence in very preterm gestation.14–16 However, the optimum mode of delivery of the early preterm fetus even in breech presentation is also controversial. Several reports do not advocate a policy of cesarean delivery for early preterm breech delivery and do not report associations with increased survival without disability or handicap.17,18
The literature about the impact of mode of delivery in extremely premature fetuses in a vertex presentation is equivocal,19,20 and a prospective randomize approach to answer this question has been unsuccessful.8 Because of the conflicting evidence in the literature and major changes in perinatal practice, we decided to investigate whether vaginal mode of delivery is associated with greater risk for poor short-term outcome in premature infants weighing less than 1,251 g. We hypothesized that vaginal delivery is a predictor of poor short-term outcome in infants less than 1,251 g at birth, and that the negative impact is significant in infants less than 751 g at birth.
MATERIALS AND METHODS
After receiving institutional review board approval, we performed a historical cohort study of infants less than 1,251 g born at 2 Emory University perinatal centers, Grady Memorial Hospital and Emory Crawford Long Hospital in Atlanta, during the last 4 years (January 1, 2000 to December 31, 2003). This study period was chosen because there was no temporal trend in cesarean delivery rates and major perinatal practices. There were no significant hospital-specific trends observed. The data were obtained from our existing prospectively recorded perinatal database. The quality of the data was assured by review of the infants’ charts when available performed in detail by the first author. A total of 429 records were retrieved and reviewed; these included the 397 records of infants who met the inclusion criteria. We did not review the information from maternal records. We excluded those infants with major congenital anomalies, outborn infants admitted to our neonatal intensive care units, and those who received only comfort-care measures until their demise.
Outcome variables measured included death, severe intraventricular hemorrhage, PVL, and combined poor short-term outcome. Severe intraventricular hemorrhage was defined as grade III or IV based on Papile's classification,21 on cranial ultrasonograms read by a pediatric radiologist who was unaware of the mode of delivery. Periventricular leukomalacia was also diagnosed in the same way at 1 month of age. Combined poor short-term outcome was defined as the presence of death, severe intraventricular hemorrhage, or PVL at any point during the admission to the neonatal intensive care unit (NICU). We analyzed these outcomes in the whole cohort (< 1,251 g), and also on infants who survived to discharge. In both groups, data were also analyzed by birth weight categories at 250-g intervals.
We recorded neonatal demographics such as birth weight, gestational age, gender, small for gestational age, race, and Apgar scores at 1 and 5 minutes of age. We also collected perinatal variables including the use of antenatal steroids, multiple gestation, breech presentation, prolonged rupture of membranes, presence of chorioamnionitis, use of magnesium sulfate before delivery, and delivery room cardiopulmonary resuscitation.
Gestational age was assigned by the best obstetric estimate (last menstrual period and early fetal ultrasonogram, when available). If this information was not available or if there was significant discrepancy among the parameters, the modified Ballard examination22 was used for gestational age as recorded in the admission physical examination. Small for gestational age was defined as birth weight less than the 10th percentile for the assigned gestational age. We defined prolonged rupture of membranes as membranes ruptured for more than 18 hours before the time of birth. Clinical chorioamnionitis was diagnosed by the obstetrical team and transcribed into our neonatal database. Delivery room cardiopulmonary resuscitation was defined as the need for chest compression or epinephrine treatment in the delivery room.
Demographics were summarized as proportions and means with standard deviations. Univariate analyses were performed with the χ2 for categorical variables, with Fisher exact test when expected cell counts were less than five, and t test for continuous variables. Multivariable logistic regression models were used to obtain adjusted odds ratios (ORs) with 95% confidence intervals (CIs) comparing outcomes between infants born by vaginal delivery and those born by cesarean delivery; only adjusted ORs are reported. This analysis was done using the entire cohort (< 1,251 g), the surviving cohort, and 250-g birth weight subgroups for the entire and surviving cohorts. Variables that were statistically different between vaginal delivery and cesarean delivery groups in the entire cohort or in any 250-g subgroup were included as covariates in the models. Statistical significance was set at P < .05. Statistical analyses were performed using SAS 8.02 (SAS Institute Inc, Cary, NC).
A total of 529 infants with a birth weight less than 1,251 g were admitted to the two perinatal centers during the 4-year study period. Of them, 95 infants who were outborn, 32 infants who received comfort-care measures only, and 5 infants with major congenital anomalies were excluded. Therefore, 397 infants met inclusion criteria; 175 (44%) of whom were born vaginally and 222 (56%) of whom were born by cesarean delivery. There was similar distribution of birth weights and gestational ages between the two groups, around 900 g and 27 weeks, respectively. The number of males was equally distributed, and the use of antenatal steroids, magnesium sulfate, and delivery room cardiopulmonary resuscitation was similar between groups. There were significant differences in the presence of multiple gestation, breech presentation, and prolonged rupture of membranes (Table 1). The distribution of birth weight by 250-g categories was similar between the two groups (Fig. 1).
In this cohort, the outcomes studied were related to previously described risk factors. In the logistic regression models, gender was a major risk factor for poor neonatal outcome in infants less than 1,251 g, regardless of mode of delivery. The combined poor short-term outcome occurred in 35% of males compared with 19% of females (P = .006). For all infants less than 1,251 g in breech presentation, 40% had combined poor short-term outcome if born vaginally compared with 29% if born by cesarean delivery (P = .24). In multiparous deliveries, 23% developed the combined poor short-term outcome if born vaginally compared with 18% if born by cesarean delivery (P = .17). Breech presentation and multiple gestation births were associated with increased risk for poor outcomes by stratification analyses. These factors were adjusted for by multivariable logistic regression when analyzing the association of mode of delivery to poor outcomes in the different cohorts reported. All ORs reported below are adjusted after logistic regression analysis.
The analysis of all infants less than 1,251 g who were studied showed no significant differences between groups in any of the outcome variables; severe intraventricular hemorrhage occurred in 16% of those born vaginally and in 15% of those born by cesarean delivery, and PVL occurred in 5% and 2%, respectively. The proportion for mortality (20% versus 18%) and combined poor short-term outcome (29% versus 25%) were not different between groups. Demographics and characteristics of infants less than 1,251 g are shown in Table 1.
In the subcategory of infants between 1,001 and 1,250 g, 3% of those delivered vaginally developed severe intraventricular hemorrhage compared with 8% of those born by cesarean delivery. The proportion for PVL, mortality, and combined poor short-term outcome was 4% versus 0%, 3% versus 7% and 7% versus 14%, respectively. None of these trends favoring vaginal delivery in this birth weight category were significant after logistic regression analysis.
Among infants between 751 and 1,000 g, the proportion of severe intraventricular hemorrhage was the same (15% in both groups). The proportions for PVL (6% versus 3%), mortality (13% versus 11%), and combined poor short-term outcome (26% versus 20%) were not significantly different after regression analysis.
In the subcohort of infants less than 751 g, there were significant differences in the proportion of severe intraventricular hemorrhage and of combined poor short-term outcome after multivariable logistic regression analyses. Among infants born vaginally, 41% developed severe intraventricular hemorrhage compared with 22% in the cesarean delivery group (OR 2.79, 95% CI 1.08–7.72). The combined poor short-term outcome occurred in 67% of those born by vaginal delivery, compared with 41% among those born by cesarean delivery (OR 2.95, 95% CI 1.25–6.95). There was a statistically nonsignificant trend toward an increased risk for PVL (5% versus 3%) and mortality (55% versus 35%) in infants born vaginally, compared with abdominal delivery. Demographics and characteristics of this subcohort are shown in Table 2.
In the analysis of infants who survived to discharge from the NICU the results were different. Among all survivors less than 1,251 g, PVL occurred more frequently in those delivered vaginally (5% versus 1%; OR 11.53, 95% CI 1.66–125). In the tiniest of survivors (birth weight < 751 g), 24% of vaginally delivered infants had severe intraventricular hemorrhage compared with 9% in the cesarean delivery group (OR 8.18, 95% CI 1.58–42.20). Demographics and characteristics of survivors weighing less than 751 g are shown in Table 3.
The question concerning mode of delivery as a predictor of poor outcome in premature infants has been difficult to answer. The limited available data on mode of delivery in extremely small infants provide conflicting evidence, which has fueled a long-lasting controversy. There have been attempts to answer this question with prospective randomized studies. However, none has been completed due to problems with recruiting of patients, crossover and ethical considerations.8 To date there has been no specific prospective randomized trial exclusively for extreme premature infants to assess the impact of the mode of delivery.
Despite the lack of evidence to support one delivery mode over the other, hysterotomy rates have increased consistently in the last decade in premature deliveries. Cesarean delivery rate at 23 weeks rose from 15.9% in 1995 to 28.2% in 2002 without major changes in demographic or obstetric factors.23–24 This trend toward cesarean delivery in premature deliveries may have an impact in neonatal outcomes, because it is known that obstetric management influences the outcome of extremely low birth weight infants. Willingness to intervene apparently results in greater likelihood of intact survival.25,26 We believe that questioning viability before delivery might influence the obstetric and neonatal approach toward these vulnerable infants, dooming them to a self-fulfilled prophecy of poor neonatal outcome with less neonatal intervention and exclusion of cesarean delivery.
This large cohort study was designed a priori to stratify the impact of mode of delivery at different birth weight categories and assess mode of delivery as a predictor of poor outcome in infants less than 1,251 g. We conducted our analysis based on birth weight because of the objective nature and reliability of this variable in the context of a retrospective design. However, the distribution of small-for-gestational-age infants was similar among groups, and the results were similar when the logistic model included gestational age instead of birth weight. The results show that when the whole cohort is examined, vaginal delivery is not associated with worse poor short-term outcome. However, after weight stratification there is a consistent trend of poorer outcomes in those infants with birth weight equal or less than 1,000 g when delivered vaginally. These differences become significant for the tiniest babies at the borderline of viability. When born vaginally, the extremely premature infants (< 751 g) have over 2 times the risk to present severe intraventricular hemorrhage and combined poor short-term outcome. Furthermore, the survivors in this smallest birth weight category had a risk of severe intraventricular hemorrhage about 8 times higher if born vaginally. Periventricular leukomalacia was the short-term outcome that was most consistently associated with vaginal delivery; the differences were statistically significant in infants less than 1,251 g who survived to discharge from the NICU, with 11 times higher risk for developing PVL.
We acknowledge several limitations to our study. One of them is the retrospective nature of the design. However, potential confounding variables were identified (birth weight, gender, antenatal steroid use, multiple gestation, breech presentation, and prolonged rupture of membranes) and statistical adjustment was done in the logistic regression model. Nevertheless, there might be other important confounders that we were unable to control for. In addition, we were not able to separate the patients according to the duration of labor, but exact determination of labor duration at this gestational age is elusive. Chorioamnionitis is a known risk factor for poor neonatal outcomes. In our study, we relied on clinical diagnosis transcribed to our neonatal database. Clinical criteria for the diagnosis of chorioamnionitis are somewhat insensitive and not supported by histologic evidence for infection.27 In a recent study, maternal chorioamnionitis by histology was detected in 45% of 101 placentas, but only 33% of these had a diagnosis of clinical chorioamnionitis.28 In our study, the proportion of mothers with this diagnosis was somewhat lower, at 22%, with no significant differences observed between the two groups (Tables 1–3).
The results of this study are different from many previous reports because others have included larger infants and have lacked the power to analyze differences among birth weight categories.20 Different from previous published studies comparing the impact of mode of delivery and to remove source of bias, we carefully identified and excluded from the analysis those infants labeled as not viable and who received comfort care. Furthermore, in our study population, the rate of intubations and cardiopulmonary resuscitation in the delivery room was similar between the two groups of infants less than 751 g (Table 2). Thus, the variability seen among the infants at the borderline of viability likely reflects the impact of mode of delivery and not a less aggressive prenatal and or postnatal management.
The most feared outcome for the practitioners and families alike is an infant that survives with very poor quality of life. White matter injury is an ominous sign in the premature infant that correlates well with cerebral palsy and poor neurodevelopmental outcome. Fortunately, PVL incidence in the extremely low birth weight infant is relatively low, about 6%.29 In our study population, the increased rate of PVL in survivors less than 1,251 g delivered vaginally was consistent with results reported in low birth weight twins delivered vaginally.30
To elucidate the true impact of mode of delivery in preterm infants, a prospective randomized trial should be conducted; however, this trial is likely to present challenges with recruitment, crossover, and ethical dilemmas. Until more evidence is available, it is not clear if cesarean delivery is the preferable mode of delivery for the extreme premature infant. However, in future clinical studies, when other interventions are evaluated, it may be useful to control for mode of delivery, particularly in the tiniest of infants. Long-term consequences of the differences found in the current study are being analyzed in the cohort of survivors who are being seen in our developmental follow-up program.
1. Thorp JA, Jones PG, Clark RH, Knox E, Peabody JL. Perinatal factors associated with severe intracranial hemorrhage. Am J Obstet Gynecol 2001;185:859–62.
2. Phillip AGS, Allan WC, Tito AM, Wheeler LR. Intraventricular hemorrhage in preterm infants: declining incidence in the 1980s. Pediatrics 1989;84:797–801.
3. Horbar JD, Badger GJ, Carpenter JH, Fanaroff AA, Kilpatrick S, LaCorte M, et al. Trends in mortality and morbidity for very low birth weight infants, 1991–1999. Pediatrics 2002;110:143–51.
4. Alexander JM, Gilstrap LC, Cox SM, McIntire DM, Leveno KJ. Clinical chorioamnionitis in the prognosis for very low birth weight infants. Obstet Gynecol 1998;91:725–29.
5. Chibber RM. Severe preeclampsia and the very low birth weight infant: the controversy over delivery mode continues. J Reprod Med 2002;47:925–30.
6. Leviton A, Kuban KC, Pagano M, Allred EN, Van Marter L.Antenatal corticosteroids appears to reduce the risk of postnatal germinal matrix hemorrhage in intubated low birth weight newborns. Pediatrics 1993;91:1083–8.
7. Ment LR, Oh W, Phillip AG, Eherenkranz RA, Duncan CC, Allan W, et al. Risk factors for early intraventricular hemorrhage for low birth weight infants. J Pediatr 1992;121:776–83.
8. Grant A, Penn ZJ, Steer PJ. Elective or selective cesarean delivery of the small baby? A systematic review of the controlled trials. Br J Obstet Gynaecol 1996;103:1197–200.
9. Grant A, Glazener CM. Elective versus selective caesarean section for delivery of the small baby (Cochrane Review). In: The Cochrane Library, Issue 1, 2001. Oxford: Update Software.
10. Nielsen TF, Hokegard KH. Cesarean section and intraoperative surgical complications. Acta Obstet Gynecol Scand 1984;63:103–8.
11. Rosen MG, Dickinson JC, Westhoff CL. Vaginal birth after cesarean: a meta-analysis of morbidity and mortality. Obstet Gynecol 1991;77:465–72.
12. Hemminki E, Merilainen J. Long-term effects of cesarean sections: ectopic pregnancies and placental problems. Am J Obstet Gynecol 1996;174:1569–74.
13. Clark SL, Koonings PP, Phelan JP. Placenta previa/accreta and prior cesarean section. Obstet Gynecol 1985;66:89–92.
14. Jonas HA, Lumley JM. The effect of mode of delivery on neonatal mortality in very low birth weight infants born in Victoria, Australia: Caesarean section is associated with increased survival in breech-presenting, but not vertex-presenting, infants. Paediatr Perinat Epidemiol 1997;11:181–99.
15. Gorbe E, Chasen S, Harmath A, Patkos P, Papp Z. Very low birth weight breech infants: short-term outcome by method of delivery. J Matern Fetal Med 1997;6:155–8.
16. Gilady Y, Battino S, Reich D, Gilad G, Shalev E. Delivery of the very low birth weight breech: what is the best way for the baby? Isr J Med Sci 1996;32:116–20.
17. Wolf H, Schaap AH, Bruinse HW, Smolders-de Haas H, van Ertbruggen I, Treffers PE. Vaginal delivery compared with caesarean section in early preterm breech delivery: a comparison of long term outcome. Br J Obstet Gynaecol1999;106:486–91.
18. Penn ZJ, Steer PJ, Grant A. A multicenter randomized controlled trial comparing elective and selective caesarean section for the delivery of the preterm breech infant. Br J Obstet Gynaecol 1996;103:684–9.
19. Vakrilova L, Emilova Z, Sluncheva B, Dimitrov A, Nikolov A, Iarukova N. Problems and outcome in extremely low birth weight newborns, depending on the mode of delivery [in Bulgarian]. Akush Ginekol (Sofia) 2002;41:24–7.
20. Bauer J, Hentschel R, Zahradnik H, Karck U, Linderkamp O. Vaginal delivery and neonatal outcome in extremely low birth weight infants below 26 weeks of gestational age. Am J Perinatol 2003;20:181–8.
21. Papile LA, Burstein J, Burstein R, Koffler H. Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weight less than 1,500 gm. J Pediatr 1978;92:529–34.
22. Ballard JL, Khoury JC, Wedig K, Wang L, Eilers-Walsman BL, Lipp R. New Ballard score, expanded to include extremely premature infants. J Pediatr 1991;119:417–23.
23. Redman ME, Gonik B. Cesarean delivery rates at the threshold of viability. Am J Obstet Gynecol 2002;187:873–6.
24. McElrath TF. Cesarean delivery at the limits of neonatal viability. Clin Obstet Gynecol 2004;47:342–51.
25. Bottoms SF, Paul RH, Iams JD, Mercer BM, Thom EA, Roberts JM, et al. Obstetric determinants of neonatal survival: influence of willingness to perform cesarean delivery on survival of extremely low birth weight infants. Am J Obstet Gynecol 1997;176:960–6.
26. Lucey JF, Rowan CA, Shiono P, Wilkinson AR, Kilpatrick S, Payne NR, et al. Fetal infants: the fate of 4172 infants with birth weights of 401 to 500 grams: the Vermont Oxford Network experience (1996–2000). Pediatrics 2004;113:1559–66.
27. Smulian JC, Shen-Schwarz S, Vintzileos AM, Lake MF, Ananth CV. Clinical chorioamnionitis and histologic placental inflammation. Obstet Gynecol 1999;94:1000–5.
28. Kazzi SN, Jacques SM, Qureshi F, Quasney MW, Kim UO, Buhimschi IA. Tumor necrosis factor-alpha allele lymphotoxin-alpha+250 is associated with the presence and severity of placental inflammation among preterm births. Pediatr Res 2004;56:94–8.
29. Fanaroff Avroy A, Hack M. Periventricular leukomalacia: prospects for prevention. N Engl J Med 1999;341:1229–31.
© 2005 The American College of Obstetricians and Gynecologists
30. Salomon LJ, Duyme M, Rousseau A, Audibert F, Paupe A, Zupan V, et al. Periventricular leukomalacia and mode of delivery in twins under 1500 g. J Matern Fetal Neonatal Med 2003;13:224–9.