The rate of cesarean deliveries increased to 26% in 2002, the highest rate ever reported in the United States.1,2 The vaginal birth after cesarean (VBAC) rate declined 23% between 2001 and 2002, from 16.4 per 100 women to 12.6.1,2 These changes are attributable partially to concerns for maternal and neonatal morbidity in those who attempt a trial of labor.3,4 Data suggest that the risk of uterine rupture is higher in those who fail the VBAC attempt.5 Studies in the past have tried to elucidate prognostic factors for the success of VBAC and uterine rupture, including prior and current obstetric history,6,7 neonatal birthweight,8 maternal weight,9 and nonclinical factors, such as hospital or physician characteristics.10 However, data are not as clear regarding the effect of other factors on maternal morbidity, such as demographics or gestational age.11 The impact of prematurity on the success of VBAC and complications is unclear. It is possible that, given the association between preterm labor and intrauterine infection,12 prematurity might be a factor that influences the success and morbidity of a trial of labor. Both subclinical and clinical chorioamnionitis may result in complications, such as infectious morbidity, particularly in those who attempt to deliver vaginally after a prior cesarean delivery. Chorioamnionitis might also lead to uterine contraction abnormalities and, thus, a failed VBAC attempt.13 Conversely, given the lower weight of preterm infants, these pregnancies may be more likely to deliver vaginally and thus experience a higher VBAC success rate when compared with term gestations.
Our main objective was to compare the success of VBAC and uterine rupture rates between preterm and term gestations in women with a history of one cesarean delivery who attempted a VBAC. Our hypothesis was that a preterm VBAC attempt was more likely to be successful and have a lower rate of complications than a term VBAC attempt. We also compared morbidity rates between women with preterm gestations with one prior cesarean delivery that either underwent an elective repeat cesarean delivery or attempted a trial of labor. Information on success and complication rates would be useful for clinicians and patients facing the decision whether to pursue a trial of labor at a preterm gestation.
MATERIALS AND METHODS
To establish whether preterm gestation is independently associated with VBAC success and complication rates, we performed a secondary analysis of a retrospective cohort study of VBAC delivery. We reviewed medical records of all women with a history of a cesarean delivery who either attempted a VBAC or underwent a repeat cesarean delivery between 1995 and 2000 in 6 community and 11 university and/or residency-affiliated hospitals. Institutional review board approval was obtained from each institution's ethics committee. Subjects were identified by an International Classification of Disease code-based search as having had a previous cesarean delivery. Trained abstractors used standard fixed instructions to collect detailed information from the medical record, including demographics, medical history, current and prior obstetric history, neonatal birthweight, major and minor complications, treatment, and outcome of the index pregnancy. Records were reviewed for 25,079 women with a prior cesarean delivery. Within this cohort, 13,613 women attempted a VBAC. Because multiple gestations or a history of more than 1 cesarean delivery might alter the selection of patients for a trial of labor or the management of labor, we limited the analysis to women with singleton gestations and a history of only one prior cesarean delivery. Breech deliveries were excluded from the analysis. The analysis comparing term and preterm VBAC patients was limited to 12,463 women.
We categorized patients who attempted a VBAC by gestational age. Gestational age was determined by the treating obstetrician. Preterm gestation was defined as a gestational age between 240/7 and 366/7 weeks. Although the etiology of preterm delivery (whether spontaneous preterm labor, preterm premature rupture of membranes, or iatrogenic preterm delivery) was not available, we did have information on whether labor was induced, which most likely occurs in indicated preterm deliveries. We also identified other exposures that could influence the rate of VBAC success and complications, such as medical illnesses (chronic hypertension, pregestational and gestational diabetes), tobacco use, labor method (spontaneous, induced, or augmented), and obstetric history (whether women had a previous vaginal delivery). Chronic hypertension was defined as elevated blood pressure diagnosed before pregnancy and treated with antihypertensive medications. Preeclampsia was defined as pregnancy-induced hypertension with proteinuria requiring magnesium sulfate for seizure prophylaxis. Pregnancy-induced hypertension was defined as elevated blood pressures in the third trimester without proteinuria. Information was collected on the hospital setting (university or community hospital) and residency program affiliation.
The primary outcomes of interest included VBAC success and maternal morbidity. Major complications were defined as uterine rupture; operative injury such as bladder, ureter, and bowel injury; and uterine artery laceration. Minor complications were defined as excessive bleeding requiring transfusion of packed red blood cells and postpartum fever. Fever was defined as a temperature of 38°C or higher 24 hours after delivery. A variable for composite major morbidity was created for patients with 1 or more major complications. Uterine rupture was defined as symptomatic rupture leading to a nonreassuring fetal heart rate pattern, bleeding, pain, and an operative intervention. Every case of uterine rupture was confirmed using laparotomy. Asymptomatic uterine dehiscence or uterine “windows” were not included in the definition of uterine rupture. These determinations were made by an independent review of the records by a study investigator.
For statistical analysis, we used Stata 8.0 SE software (Stata, College Station, TX). We performed 2 separate analyses. The first analysis was performed to assess whether gestational age (categorized as term compared with preterm) influenced VBAC success and failure. The second analysis was performed to compare rates of maternal morbidity between preterm deliveries that occur in a VBAC attempt compared with a repeat cesarean delivery. Comparisons were made with the Student t test for continuous variables and χ2 analysis or the Fisher exact test for categorical variables. VBAC success and morbidity rates by gestational age category and mode of delivery were calculated using χ2 analysis. Multivariable logistic regression models, controlling for confounding, were developed to evaluate both the effect of prematurity on VBAC success and morbidity and the effect of delivery mode (VBAC attempt compared with elective cesarean delivery) on morbidity among preterm gestations. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were derived from the models.
Within the entire cohort of 25,079 women, 20,156 had a history of only 1 prior cesarean delivery. Among these women, 12,463 (61.8%) attempted a VBAC; of these, 11,492 (92.2%) delivered at term, whereas 971 (7.8%) delivered preterm. Mean gestational ages for the term and preterm groups were 39.2 weeks and 33.9 weeks of gestation, respectively. Significant differences between the term and preterm gestations included maternal age, gravidity, race, insurance status, tobacco use, medical conditions (including chronic hypertension and diabetes), and obstetric history (Table 1). Women in the preterm group were more likely to have spontaneous labor than induced or augmented labor. Preterm gestations were more commonly treated in hospitals associated with a university or with a residency program (Table 1).
The VBAC success rate in patients at term with 1 prior cesarean delivery was 74%, and the success rate in preterm patients was 82%. The rate of uterine rupture was 0.9% among women attempting a VBAC at term, whereas the rate was 0.3% among women attempting a VBAC at a preterm gestation (unadjusted relative risk 0.34, 95% CI 0.11–1.06).
Logistic regression models were developed to evaluate the effect of prematurity on the success and morbidity of a VBAC attempt controlling for potential confounding variables. The findings of the unadjusted analysis were unchanged in our multivariable analysis; preterm VBAC patients were more likely to succeed than term VBAC patients (Table 2). There also was a suggestion of lower uterine rupture rates in the preterm VBAC group, although this finding did not reach statistical significance (adjusted OR 0.28, 95% CI 0.07–1.17; P = .08). The adjusted odds ratio for composite morbidity was similar between term and preterm gestations (Table 2).
As a separate analysis, we compared the rate of morbidity among patients with preterm gestations who attempted a trial of labor compared with those who underwent an elective repeat cesarean delivery. Among the cohort of 20,156 women with a history of 1 prior cesarean delivery, 1894 (9.4%) delivered preterm and 971 (51% of total preterm gestations) attempted a trial of labor. As we described previously, the rate of uterine rupture among women with preterm gestations attempting a VBAC was 0.3%. The rate of uterine rupture among women with preterm gestations undergoing an elective repeat cesarean delivery was 0.1%. Rates of major complications were similar between these 2 groups (Table 3). However, the rates of minor complications, such as transfusion and postpartum fever, were higher in those women who underwent an elective cesarean delivery at a preterm gestation (Table 3). Logistic regression models were developed to evaluate the effect of a VBAC attempt among women with preterm gestations on maternal morbidity. The adjusted OR for composite morbidity among preterm gestations was 0.44 (95% C.I. 0.14–1.36), which was not significantly different between the groups. However, the adjusted odds ratio for minor complications revealed a 70–80% reduction in risk among women with preterm gestations that attempted a trial of labor (Table 3).
We compared the morbidity among preterm patients who had either a successful or unsuccessful VBAC attempt. Composite morbidity was higher in the women who failed the VBAC (adjusted OR 11.7, 95% CI 2.6–52.1), which is consistent with previously published data in term gestations. Postpartum fever and other major operative morbidities were also more common in the preterm women who failed the VBAC attempt (data not shown).
In our retrospective cohort study, preterm patients had a higher VBAC success rate than term patients. We also found that there was a trend toward a lower uterine rupture rate in preterm patients who attempted a VBAC. Major operative morbidity was not higher in women with preterm gestations that attempted a VBAC compared with those who elected to have a repeat cesarean delivery. Although we hypothesized that preterm women attempting a VBAC might have a higher success rate than term patients attempting a VBAC, one also might hypothesize that the success among preterm patients undergoing VBAC could be lower if infection were playing a role in the etiology of labor. Previously published studies from our cohort have not evaluated preterm delivery as an independent risk factor for VBAC morbidity, and this study addresses the unique question of the effect of prematurity in the success and morbidity of VBAC.
The reason for the increased VBAC success among women with preterm gestations may be related to the lower birthweight compared with term gestations. Although prematurity has not been identified as a prognostic factor for VBAC success in previous studies, data suggest that greater birthweight may negatively impact the likelihood of a successful VBAC.8 Therefore, it is plausible that lower birthweight may explain our findings. We did not examine the effect of fetal weight on the success of a VBAC for several reasons. Ultrasonography, although a useful aid in estimating fetal weight, has a known margin of error in the third trimester, making counseling for VBAC on the basis of estimated fetal weight difficult. We also feel that obstetricians are better able to determine gestational age than fetal weight and thus feel that counseling for VBAC on the basis of gestational age is more appropriate.
Among preterm gestations, we found a trend toward lower uterine rupture rates. However, this finding should be interpreted with caution because our larger cohort study was not powered to address this question. We also found that major complications were similar among those who either attempted a VBAC or underwent an elective repeat cesarean delivery. Our findings are in contrast to data from McMahon et al,5 in which major complications were higher for those who attempted a trial of labor. Our results may be the result of a type II error because our study was not powered to find differences in the rates of these complications among women with preterm gestations. We also found that minor complications, both the need for transfusion and postpartum fever, were higher for those who underwent an elective repeat cesarean delivery than for those who attempted a trial of labor. This finding is similar to data from McMahon et al,5 in which the risk of postpartum fever was 25% higher in the elective cesarean delivery group.5 However, the rate of blood transfusion was not different between those women attempting a VBAC and an elective cesarean delivery.5 Our data specifically addressing prematurity may reflect the pathophysiology in these pregnancies. Intrauterine infection plays a role in the preterm labor process,12 potentially contributing to both postpartum hemorrhage and febrile morbidity in the women who undergo surgery. Composite morbidity was higher for those preterm patients who failed their VBAC attempt. This finding is consistent with previously published data on the morbidity of a failed VBAC.5
Previous studies addressing VBAC success and uterine rupture rates have been limited by differences in outcome definitions.11 Strengths in our study include the large number of patients in the cohort and the fact that we used clear definitions for our primary and secondary outcomes of interest, including uterine rupture, established before the study. However, one study limitation is that complete information was not available on the specific indication for preterm delivery. Patients in preterm labor with a ripe cervical examination have a higher likelihood of a successful trial of labor than preterm patients whose delivery is indicated by preeclampsia, chronic maternal disease, or placental abruption. Because the need for induction or augmentation of labor may influence the success and morbidity of the trial of labor, we controlled in the multivariable analysis for method of labor and for any known maternal comorbidity that might necessitate preterm delivery. These steps should minimize bias that might result from not having data on the indication for preterm delivery. Our results showed that the success of a VBAC attempt was still higher for women with preterm gestations than term gestations. When we compared women with preterm gestations that either underwent a VBAC or had an elective cesarean delivery, major complications were similar between the groups. Information on the indication for preterm delivery might have been helpful in clarifying who is at most risk for morbidity with a VBAC attempt.
Another limitation in our study is not having information on other types of morbidity, such as anal sphincter tears, pulmonary complications, and hospital stay. These types of morbidity are potentially relevant in an attempted VBAC. However, information on these variables was not collected as part of the cohort study. Complete neonatal information also was not collected. Neonatal data are difficult to obtain in a retrospective fashion, particularly in such a large group of patients.
Other limitations in our retrospective cohort design include the possibility of selection bias and the potential for residual confounding. However, we feel that these are minimized because we used strict inclusion criteria, a standardized data collection process, and strict outcome definitions. Finally, these results may not be generalizable to preterm patients with more than a single previous cesarean delivery or those with a multiple gestation.
In conclusion, prematurity is associated with a higher likelihood of a successful trial of labor and a trend toward a lower uterine rupture rate. The high rate of success among preterm gestations attempting a VBAC should be discussed with patients that may be considering a trial of labor in the preterm period. In 2002, the rate of preterm birth among singletons was 10.4%.2 If one assumes a primary cesarean delivery rate of 18% among preterm patients, 72,000 women annually may face the question whether to pursue a vaginal birth in the future. In our opinion, this is a significant number of women. A VBAC attempt should be offered to preterm patients because of its favorable rate of success and lower maternal morbidity. Moreover, a successful trial of labor is associated with a lower risk of downstream maternal complications, such as placenta accreta and hysterectomy (Paré E, Quiñones J, Macones G. Am J Obstet Gynecol 2002;187:S104 [abstract]). This information is important for counseling patients with a singleton preterm gestation who may be considering a trial of labor after a single prior cesarean delivery.
1. Menacker FC, Curtin SC. Trends in cesarean birth and vaginal birth after previous cesarean, 1991–99. Natl Vital Stat Rep 2001;49:1–16.
2. Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Menacker F, Munson ML. Births: Final Data for 2002. Natl Vital Stat Rep 2003;52:1–113..
3. Flamm B. Once a cesarean, always a controversy [review]. Obstet Gynecol 1997;90:312–5.
4. Lydon-Rochelle M, Holt V, Easterling T, Martin D. Risk of uterine rupture during labor among women with a prior cesarean delivery. N Engl J Med 2001;345:3–8.
5. McMahon M, Luther E, Bowes W, Olshan A. Comparison of a trial of labor with an elective second cesarean section. N Engl J Med 1996;335:689–95.
6. Marchiano D, Elkousy M, Stevens E, Peipert J, Macones G. Diet-controlled gestational diabetes mellitus does not influence the success rates for vaginal birth after cesarean delivery. Am J Obstet Gynecol 2004;190:790–6.
7. Macones G, Hausman N, Edelstein R, Stamilio D, Marder S. Predicting outcomes of trials of labor in women attempting vaginal birth after cesarean delivery: a comparison of multivariate methods with neural networks. Am J Obstet Gynecol 2001;184:409–13.
8. Elkousy M, Sammel M, Stevens E, Peipert J, Macones G. The effect of birth weight on vaginal birth after cesarean delivery success rates. Am J Obstet Gynecol 2003;188:824–30.
9. Carroll C, Magann E, Chauhan S, Klauser C, Morrison J. Vaginal birth after cesarean section versus elective repeat cesarean delivery: Weight-based outcomes. Am J Obstet Gynecol 2003;188:1516–22; discussion 1520–2.
10. Hashima J, Eden K, Osterweil P, Nygren P, Guise J. Predicting vaginal birth after cesarean delivery: a review of prognostic factors and screening tools [review]. Am J Obstet Gynecol 2004;190:547–55.
11. Guise JM, Berlin M, McDonagh M, Osterweil P, Chan B, Helfland M. Safety of vaginal birth after cesarean: a systematic review [review]. Obstet Gynecol 2004;103:420–9.
12. Goldenberg R, Hauth J, Andrews W. Intrauterine infection and preterm delivery [review]. N Engl J Med 2000;342:1500–7.
13. Mark S, Croughan-Minihane M, Kilpatrick S. Chorioamnionitis and uterine function. Obstet Gynecol 2000;95:909–12.