In 2003, the cesarean delivery rate in the United States was 27.6%, the highest ever reported. Rates have been increasing steadily in recent years, exhibiting a one-third rise since 1996.1 Given that one third of cesarean deliveries are due solely to a history of a prior cesarean, experts have proposed that an effective way to reduce the rate of cesarean deliveries in the United States is to encourage women to attempt a vaginal birth after cesarean (VBAC) delivery.2 One of the objectives of Healthy People 2010 calls for a reduction in the rate of repeat cesarean births from 72% in 1998 to a target of 63% in 2010, while recognizing the need to carefully monitor outcome data among VBAC attempts to ensure that mothers and infants are not exposed to an increase in morbidity and mortality.3
Counseling women with a history of a prior cesarean about mode of delivery is complex. Patients need to be presented not only with accurate success rates but also with an appropriate estimate of the level of risk associated with each option and outcome. Prior studies that have attempted to predict VBAC failures and morbidity have been largely unsuccessful.4,5 Investigators have now turned to identifying individual risk factors that may be related to VBAC outcomes. One such category of patients that has been a concern for some clinicians and researchers is the postterm VBAC candidate.
Although VBAC data have been examined in different patient populations and clinical scenarios, little has been published about VBAC attempts in women whose pregnancy has reached or extended past the estimated date of delivery (EDD). The few studies that have examined this issue have limitations that make the information difficult to apply universally in clinical practice.6–9 For example, all studies recruited patients from a single institution, which may limit the generalizability of their findings. Due to the small number of patients, these studies also had limited power to detect a clinically relevant difference in the rare complication of uterine rupture. In addition, no study reported the proportion of patients in a cohort of women who are beyond their EDD who elected a trial of labor after a cesarean delivery.
Our objective was to examine VBAC patterns in women whose pregnancies have reached or extended beyond the estimated date of delivery. We studied how frequently these women attempt VBAC, how frequently VBAC failures occur, and whether women attempting VBAC past their EDD have a higher rate of complications, including uterine rupture.
MATERIALS AND METHODS
To assess the effect of post-EDD gestational age on the success and safety of VBAC, we performed a planned secondary analysis of a retrospective cohort study of women from 17 community and university-based hospitals in the northeastern United States. The local institutional review board at each center approved the study. Patients who were included in the cohort delivered their index pregnancy during a 5-year span between 1996 and 2000. We excluded patients with a prior vertical (classical) or unknown type of uterine incision and women with an unknown or inaccurate gestational age estimate at delivery. Details of the study are published elsewhere,10 but a brief description follows. More than 25,000 pregnant women with a history of a previous cesarean delivery were identified by International Classification of Diseases, 9th Revision codes. Trained abstractors reviewed the medical records and collected data. We recorded basic demographic information, including maternal age, race, socioeconomic status, and medical history. We also collected information about both the current pregnancy and the prior cesarean delivery. We recorded circumstances and outcomes of the current pregnancy, including an attempted trial of labor, elective repeat cesarean delivery, and successful or failed VBAC. We also examined and recorded complications of delivery. Uterine rupture was defined as a complete separation of the uterine muscle demonstrated at laparotomy in conjunction with either maternal compromise (signs or symptoms of acute bleeding or hemoperitoneum) or fetal compromise (nonreassuring fetal heart rate patterns). This variable was recorded from the medical record by the trained abstractors and reviewed by the principal investigator to ensure accurate classification of the outcome. Other complications examined included the need for maternal transfusion, bladder injury, and other major operative complications, including bowel injuries and vascular injuries. We also compared the groups for a composite maternal morbidity variable that included all of the above complications excluding blood product transfusions. Failure of VBAC is defined as a failed trial of labor after a prior cesarean delivery.
We were primarily interested in 3 comparisons. First, we examined the effect of post-EDD gestational age on the risk of VBAC failure. Second, we examined the effect of post-EDD gestational age on the risk of maternal morbidity among patients who attempted a VBAC. Third, we assessed whether, among those at or beyond 40 weeks, maternal complications were increased in patients who attempt VBAC compared with those who elect a repeat cesarean delivery. For the first 2 primary aims, we identified the exposed group using 2 different gestational age cutoffs: 40 0/7 or more weeks and 41 0/7 or more weeks of gestation at the time of delivery. We completed the analysis for each outcome in 3 stages, including descriptive statistics, an unadjusted analysis, and a multivariate analysis. To assess the independent effect of post-EDD gestational age on these outcomes, we developed multivariate logistic regression models to estimate odds ratios adjusted for all identified confounding variables. We included in each model those independent variables that were statistically significant in the univariate analysis or deemed clinically relevant. For each outcome we created an initial model and removed variables in a backward stepwise fashion until an appropriate model was created to assess the independent association between post-EDD gestational age and the outcome of interest. We performed all statistics with Stata 8 software (Stata Corporation, College Station, TX).
The entire cohort included 25,076 women who were pregnant with a history of at least 1 prior cesarean and delivered at 1 of the study hospitals. There were 13,614 women who attempted a VBAC past 24 weeks of gestation, for an overall VBAC attempt rate of 54%. We excluded 2,027 women (14.9%) because accurate gestational age data were lacking, and therefore, the exposure could not be precisely assessed. The baseline demographic characteristics for the women excluded for inaccurate gestational age data were similar to the remainder of the cohort (data not shown). The remaining 11,587 women who comprised the VBAC study population were divided into 2 groups based on whether they attempted VBAC at less than 40 weeks or 40 or more weeks of gestation. The demographic characteristics of these 2 VBAC groups are shown in Table 1. Women who attempted a VBAC past the EDD were less likely to have had a prior vaginal delivery, were more likely to require induction of labor, and less often had pregnancy complications such as diabetes, preeclampsia, or chronic hypertension. This decreased rate of pregnancy complications seen in the subgroup of patients past the EDD is likely due to the fact that most physicians will deliver women with these complications by their due date. The patients who attempted a VBAC after their EDD also were less likely to smoke cigarettes and were more frequently white. As expected, the mean birthweight of babies born to women attempting a VBAC past the EDD was significantly higher compared with women attempting a VBAC before 40 weeks of gestation.
Women who were at or past their EDD were more likely to attempt VBAC (79.1% compared with 44.7%, relative risk [RR] 1.77, 95% confidence interval [CI] 1.73–1.81). However, in the unadjusted analysis, VBAC failure was significantly increased in the women attempting VBAC at 40 or more weeks of gestation (Table 2). Rates of major operative complications (including bowel, bladder, and vascular injuries) and the composite morbidity outcome were also increased in the unadjusted analysis for women who attempted a VBAC at 40 or more weeks of gestation. The multivariate analysis of VBAC failure confirmed the unadjusted analysis, but the differences between groups in overall maternal morbidity (operative injuries and uterine rupture) did not remain statistically significant after adjusting for confounders (Table 3). We also performed a subanalysis, restricting the comparison group to only those women who attempted a VBAC at term (37 0/7 weeks to 39 6/7 weeks). The odds ratios and confidence intervals were not significantly changed, both for VBAC failure and for VBAC complications (data not shown, available on request).
Next, we examined the subgroup of women who attempted VBAC at or after 41 weeks of gestation, comparing them with the remainder of the cohort. The risk of a failed VBAC was significantly increased in the women attempting VBAC at or after 41 weeks, even after adjusting for potential confounders. However, the rates of uterine rupture and other maternal morbidities were not significantly increased (Table 4).
Finally, we examined the entire cohort of women who delivered on or after the EDD, to compare complication rates of women with post-EDD gestations who attempted a trial of labor with those women who opted for repeat cesarean delivery (Table 5). Women who opted for a trial of labor were over 13 times more likely to have a uterine rupture than those women undergoing elective cesarean delivery. However, this relative risk did not significantly differ from the cohort of women attempting VBAC before 40 weeks of gestation, who had a RR of uterine rupture of 21 (P = .7, Table 6). Major operative complications were increased in the women attempting VBAC, but these did not reach statistical significance. However, when uterine ruptures were combined with operative complications, the overall composite morbidity was increased in women attempting VBAC compared with those electing for repeat cesarean delivery (Table 5). Similar to a general population of VBAC candidate patients before the due date (Table 6), the patients in the post-EDD cohort who had a failed trial of labor had the highest risk for uterine rupture and other major delivery-related morbidities, both in the subgroups that were 40 or more weeks and those that were 41 or more weeks (Tables 5 and 7). Importantly, the relative risks for maternal morbidity in these post-EDD cohorts of patients who failed or succeeded in VBAC are not in excess of the relative risks for the entire cohort of patients who delivered before the EDD (Table 6).
In this large retrospective cohort study of over 25,000 VBAC candidates, we determined that women who attempt VBAC beyond their EDD are 36% more likely to fail a trial of labor, but importantly, they are no more likely to experience delivery-related major morbidity than patients who attempt VBAC before the EDD. The findings of increased VBAC failure without an increase in overall morbidity persisted when we analyzed a subcohort of patients who attempted VBAC at 41 or more weeks. However, although the failure rate is increased in women attempting a VBAC after the due date, the success rate of 69% is within the range of overall VBAC success that has been reported in the literature.2 In the group of patients who delivered after their EDD, patients who attempted VBAC had a 13-fold increase in uterine rupture and a 2-fold increase in overall morbidity compared with patients who elected a repeat cesarean. It is of great significance that these relative risks of the post-EDD cohort are not in excess of or disproportionate to those found in a generalized population of VBAC candidates (Table 6). In addition, one must remember that although the RR of morbidity is increased in VBAC patients, the absolute risk is small–-in women past 40 weeks of gestation, the risk of uterine rupture is 1.1% in women attempting VBAC compared with 0.08% in those electing repeat cesarean delivery.
The data available to date to assist practitioners in counseling women about the possible risks of a VBAC attempt at or after the EDD consist primarily of 4 studies. An early case study reported a successful VBAC rate of 73% with no significant increase in maternal or neonatal morbidity.6 This study, however, reported women who had reached 42 weeks of gestation and is not as relevant for current practice, because most women are delivered before 42 weeks due to the potential for increased perinatal morbidity. In our cohort, the limited number of women who reached 42 weeks did not allow us to report accurately the relative risks of VBAC attempts in that subgroup. Callahan et al,7 in a case–control study, compared women attempting VBAC past 40 weeks of gestation with women of the same gestational age with no history of prior cesarean delivery. Among the cases, the VBAC success rate was similar to that reported in the literature (66%), but the choice of a comparison group without a history of prior cesarean limits the application of the study results to patient counseling. Zelop et al8 compared women attempting VBAC at or before 40 weeks of gestation with those attempting VBAC beyond 40 weeks. This study included 2,775 women, of whom 1,271 delivered after 40 weeks. They found that women beyond 40 weeks were more likely to have a failed VBAC (35.4% compared with 26.7%, P < .001). In the adjusted analysis, gestational age more than 40 weeks remained a risk factor for a failed VBAC both for spontaneous (OR 1.5, CI 1.2–1.8) and induced labor (OR 1.5, CI 1.1–2.2). However, the authors report that the risk of uterine rupture was not increased in the cohort of women attempting VBAC past 40 weeks of gestation, although this finding is limited because of the relative small sample size. Assuming a uterine rupture rate of 1% in the group that delivered less than 40 weeks, the study can detect no less than a 2.5-fold increase in rupture. Most recently, Hammoud et al9 reported their analysis of a cohort of patients attempting VBAC examined by gestational age category. They reviewed records of 329 patients 41 or more weeks of gestation attempting VBAC, comparing those with 2 groups of earlier gestational ages (24–36 6/7 weeks and 37–40 6/7 weeks of gestation). They noted, as in our study, a lower rate of VBAC success with advanced gestational age. However, they reported a significantly elevated risk of uterine rupture in the most advanced gestational age category, with an odds ratio of 2.85 in women attempting VBAC at 41 or more weeks. This is in contrast to our findings in the present study, which showed no increased risk of uterine rupture in those attempting VBAC at 40 or more weeks of gestation or 41 or more weeks of gestation. The Hammoud data came from a single institution and involved a smaller number of patients, thus making the findings less applicable to the general population.
Our study has several advantages. Our cohort comprised almost 6,000 women who remained undelivered by the estimated date of delivery. This gives us greater statistical power to detect smaller but clinically relevant differences in some of the potentially catastrophic complications of VBAC. For example, in a post hoc power analysis, we determined that this study has 80% power to detect an odds ratio of 1.6 for uterine rupture and 2.0 for other rare major morbidities in the post-EDD VBAC group compared with the pre-EDD VBAC group. Furthermore, this study has 80% power to detect among the post-EDD group of patients a 2-fold increase in major morbidity with VBAC compared with elective cesarean delivery. Many of the articles written on the safety of VBAC have been conducted in tertiary care centers,11 and may not be applicable to community hospitals. Our data were collected both from university centers and community hospitals, in both teaching facilities and those without resident training programs. This makes our findings more broadly applicable to the general population. In addition, the cohort study design provides an advantage over some of the previous investigations. The cohort study allows us to examine multiple outcomes and adjust for many confounding variables. The study design also enables us to investigate 2 related, but distinct issues. First, we are able to compare VBAC outcomes in advanced gestation patients to those who deliver before the EDD. The information on VBAC safety and success afforded by this comparison is most useful to the health care system in developing policy. Second, we are able to assess safety across delivery schemes among patients of advanced gestation, namely VBAC compared with repeat cesarean. Although this information is also important to the clinician and health system, it is most relevant to the patient faced with the decision of whether to attempt a VBAC. Given the large number of patients, we were able to examine carefully the rarer complications of VBAC failures such as bladder and bowel injuries. Finally, in this study, the data were obtained by trained abstractors who were following strict, detailed, a priori definitions of the outcome and potential confounding variables, minimizing the potential for bias.
We must, however, acknowledge the limitations of our data. Some degree of bias can always be present in an observational study, although as previously discussed, we made every effort to minimize information bias by defining the variables and inclusion criteria before data abstraction. We had to exclude almost 15% of the patients in the cohort due to the lack of accurate gestational age data, and so we cannot overlook the possibility of bias. However, we are encouraged that selection bias was limited by excluding this portion of patients, because they were similar to the remainder of the cohort with respect to demographic and multiple historical variables. Lastly, previous work in this cohort has shown that there are many factors, both clinical and nonclinical, that influence decision making in regard to VBAC attempts.12 The observational nature of this study does not allow us to measure many of the factors that may be considered both by patients and physicians when deciding on a course of action in a pregnancy with a history of a prior cesarean delivery. These immeasurable factors have the potential for introducing selection bias, because patients or physicians may differentially select only the most promising post-EDD candidates for VBAC. For example, patients in the VBAC group who delivered after the EDD less commonly had more than 1 prior cesarean compared with the VBAC group who delivered before the EDD. Perhaps clinicians counseled post-EDD patients away from VBAC if they had more than 1 prior cesarean for fear of a suspected increase in risk of uterine rupture. We were able to assess many, but certainly not all, of these factors in the multivariate analyses.
In conclusion, practitioners can counsel women that the chances of VBAC success are decreased in gestations that continue beyond the EDD compared with those before 40 weeks. However, patients can be reassured that the success rate of almost 70% is still quite good. In addition, the risks of a VBAC attempt are not increased simply because the due date has passed.
1. Hamilton BE, Martin JA, Sutton PD. Births: preliminary data for 2003. Natl Vital Stat Rep 2004;53:1–17.
2. Vaginal birth after previous cesarean delivery. ACOG Practice Bulletin no. 54. American College of Obstetricians and Gynecologists. Obstet Gynecol 2004;104:203–11.
3. U.S. Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health. 2nd ed. Washington, D.C.: U.S. Government Printing Office; 2000.
4. Pickhardt MG, Martin JN Jr, Meydrech EF, Blake PG, Martin RW, Perry KG Jr, et al. Vaginal birth after cesarean delivery: are there useful and valid predictors of success or failure? Am J Obstet Gynecol 1992;166:1811–5.
5. Macones GA, Hausman N, Edelstein R, Stamilio DM, Marder SJ. 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.
6. Yeh S, Huang X, Phelan JP. Postterm pregnancy after previous cesarean section. J Reprod Med 1984;29:41–4.
7. Callahan C, Chescheir N, Steiner BD. Safety and efficacy of attempted vaginal birth after cesarean beyond the estimated date of delivery. J Reprod Med 1999;44:606–10.
8. Zelop CM, Shipp TD, Cohen A, Repke J, Lieberman E. Trial of labor after 40 weeks’ gestation in women with prior cesarean. Obstet Gynecol 2001;97:391–3.
9. Hammoud A, Hendler I, Gauthier RJ, Berman S, Sansregret A, Bujold E. The effect of gesational age on trial of labor after cesarean section. J Matern Fetal Med 2004;15:202–6.
10. Macones GA, Cahill A, Par E, Stamilio DM, Ratcliffe S, Stevens E, et al. Obstetrical outcomes in women with two prior cesareans: is vaginal birth after cesarean delivery a viable option? Am J Obstet Gynecol. 2005;192:1223–8.
11. Yap OW, Kim ES, Laros RK. Maternal and neonatal outcomes after uterine rupture in labor. Am J Obstet Gynecol 2001;184:1576–81.
12. Dunsmoor-Su R, Sammel M, Stevens E, Peipert JL, Macones G. Impact of sociodemographic and hospital factors on attempts at vaginal birth after cesarean delivery. Obstet Gynecol 2003;102:1358–65.