The counseling for patients considering vaginal birth after cesarean delivery (VBAC) has become increasingly more complex. As our knowledge of VBAC has expanded,1,2 so has the scope of counseling. Factors in the decision regarding delivery mode may include the likelihood of a successful vaginal delivery, risk of VBAC-associated morbidities, and family size. It is clear that if patients attempt VBAC and do not experience the rare event of uterine rupture, their greatest risk for morbidity is associated with a failed VBAC attempt.3 To that end, several researchers have tried to develop predictive models for VBAC success or failure with mixed results.4–7 Factors such as prior vaginal delivery,8 spontaneous labor,9 and nonrecurring indication for prior cesarean delivery9 are associated with relative increase in VBAC success. One factor that is less clearly characterized is the effect of maternal race on the likelihood of VBAC success and the risk for associated morbidities. We sought to estimate the role of race on VBAC success and VBAC-associated morbidities, including uterine rupture and major surgical adverse events.
MATERIALS AND METHODS
A large, multi-center, retrospective cohort study of patients with at least one prior cesarean delivery was conducted to estimate the rate of and risk factors for maternal morbidities associated with VBAC, including uterine rupture. Within the large cohort of over 25,000 patients, this secondary analysis examined the role of maternal race in 13,706 patients with a history of at least one prior cesarean delivery who elected to attempt VBAC. Patients of black race were compared with patients of other racial groups with respect to rate of VBAC attempt, rate of VBAC failure, risk for uterine rupture, and other VBAC-related major morbid events. The methods of the large, retrospective cohort study have been described previously in detail,1 but a brief description follows.
The study was conducted in the northeastern United States from 1995 to 2000, with institutional review board approval at all 17 participating medical centers (both community and tertiary). Study subjects were identified by International Classification of Disease, 9th Revision (ICD-9) codes for “previous cesarean delivery, delivered,” and data were extracted from medical charts by trained research nurses using standardized, closed-end data collection forms. For quality assurance, we performed re-extraction in 3% of the charts. Data collected included maternal demographics, medical and obstetric history, and social and family history. Details of the index pregnancy, antepartum course, labor and delivery events, complications, and maternal outcomes were also recorded. Maternal race was determined by patient self-report. For this analysis, we excluded women without a reported race or those who did not fall into one of four groups that we examined: black, Asian, white, or Hispanic.
Primary outcomes measured in this study included rates of VBAC failure, uterine rupture, and composite maternal morbidity. Uterine rupture was defined a priori to strictly identify and distinguish cases from asymptomatic uterine dehiscence and uterine “windows” found incidentally at surgery. Thus, cases of uterine rupture sustained a full-thickness scar separation and had clinical evidence of rupture, including at least one of the following: acute maternal hemorrhage, maternal hypotension (systolic blood pressure less than 70 mm Hg or diastolic blood pressure less than 40 mm Hg), maternal heart rate greater than 120 beats per minute, blood in the peritoneal cavity at the time of laparotomy, or nonreassuring fetal heart rate tracing immediately preceding surgery. A binary composite maternal morbidity outcome variable was created, defined as at least one of the following: uterine rupture, bladder or bowel injury, or uterine artery laceration.
Descriptive statistics were used to estimate the frequencies of races within the cohort and to assess the race-specific rates of VBAC attempt and outcomes. Based on the results of the descriptive and unadjusted analyses, the racial study groups were collapsed into a binary variable including black race and a category of all other races studied (hereafter referred to as the “other race” category). The characteristics and outcomes of the two groups were compared using Student t test or χ2, as appropriate. Stratified analysis was conducted to identify potential confounding variables and interaction. Finally, logistic regression models were fit to estimate the association between maternal race and the risk for each of the three outcomes, while adjusting for relevant and significant confounding variables. Stratified sub-analysis was also performed based on whether patients experienced a VBAC failure. Regression models were created in a backward, stepwise fashion, removing potentially confounding variables that were not significant. Because hospital site was a potential confounder beyond the level of individual patient data and could have resulted in nonindependence within hospital groups, it was included as a cluster variable in the final models. All statistical analyses were completed using STATA 8, Special Edition (College Station, TX).
Of the 25,005 patients with at least one prior cesarean delivery enrolled in the cohort study, 23,213 (95.7%) specified one of the four race categories studied. Of these, 14,911 (61.5%) patients were white, 6,561 (27.0%) were black, 1,256 (5.2%) were Hispanic, and 485 (2.0%) were Asian, with varying rates of VBAC attempt (Fig. 1). Because Asian and Hispanic races were the least prevalent, and were similar to white patients with respect to risk factors and outcomes (data not shown), we defined race as a binary variable, with black and other race categories. When the two study groups, defined as black and other race women who attempted VBAC, were compared for baseline sociodemographics and history (Table 1), we found some statistically significant differences, but not all differences are clinically significant. Black women, on average, were younger, more likely to have a diabetic or hypertensive comorbidity, more likely to give birth earlier and to have a smaller infant, and more likely to smoke tobacco than the women of other races. The other race group of women were more likely to deliver at a university hospital and more likely to have had prior abdominal surgery, but they were of lower average gravidity and were less likely to have had a prior vaginal delivery or to have had more than one prior cesarean delivery.
In Table 2 we show the rates, unadjusted relative risks, and adjusted odds ratios for failed VBAC attempt, uterine rupture, and composite maternal morbidity in black women as compared with the other race group of women who attempted VBAC. Although more likely to attempt VBAC (62.2% compared with 52.2%, adjusted odds ratio [OR] 1.26, 95% confidence interval [CI] 1.04–1.52, P=.020), black women were significantly more likely to have a failed VBAC trial (unadjusted relative risk [RR] 1.11, 95% CI 1.04–1.18) than other women, and this increased risk for an unsuccessful trial persisted after adjusting for significant confounding variables (adjusted OR 1.50, 95% CI 1.29–1.74).
There was a significant difference in risk of uterine rupture between the two groups (Table 2). Black women were 40% less likely to experience a uterine rupture than other women (unadjusted RR 0.54, 95% CI 0.35–0.85). After adjusting for spontaneous or induced labor, prior vaginal delivery, oxytocin exposure, and hospital site, black women had a significantly lower risk of uterine rupture than nonblack women (adjusted OR 0.59, 95% CI 0.39–0.89). The indication for prior cesarean delivery was considered in the models, but not included in the final regression models because the indication (such as dystocia) did not alter the association between black race and any of the primary outcomes.
Finally, because it is well established that VBAC failure is associated with the greatest morbidity and because black patients appear to have a higher rate of unsuccessful VBAC attempt, we examined the subgroup of women who had failed VBAC attempt (Table 3). Among women who experienced a failed VBAC attempt, black women were significantly less likely to experience a uterine rupture than other women after adjusting for the confounding effects of labor induction or augmentation (as compared with spontaneous labor) and more than one prior hysterotomy.
Race is significantly associated with VBAC failure and risk of uterine rupture. Although previous studies have demonstrated that black women are more likely to attempt VBAC, limited results have been published on the outcomes of those VBAC trials. We found that, despite a small but significant decrease in rate of VBAC success in black women as compared with women of other racial groups, black women are 40% less likely to experience a uterine rupture. This adds to a growing body of evidence that race likely plays a significant role in many aspects of labor and VBAC trials, including length of labor,10 rate of VBAC attempt,11 VBAC success,7 VBAC failure,12 and VBAC-associated morbidities. The possible explanations for these disparities are many, ranging from inherent genetic and anatomic differences to differences in fetal parameters such as weight, disparities in counseling, management, clinical decision making, access to health care, and acceptance of care plans.
Several investigators have reported a difference in length of labor between racial and ethnic groups. Greenberg et al10 performed a retrospective cohort study of over 27,000 singleton births and reported that black women had shorter second stages of labor (median: 50 compared with 92 minutes, P<.001, for nulliparas; 15 compared with 20 minutes, P<.001, for multiparas) and were significantly less likely to have a prolonged second stage (adjusted OR 0.54, 95% CI 0.45–0.68 for nulliparas; adjusted OR 0.61, 95% CI 0.49–0.65 for multiparas) when compared with other ethnic groups. When investigators described the possible etiologies of these disparities, they highlighted the spectrum of factors that could explain or contribute, including racial differences in pelvic anatomy.
If the anatomic racial differences in pelvimetry impart a faster second stage of labor for black women, it is not intuitive why black women would be more likely to experience a failed VBAC attempt. Yet Hollard et al12 retrospectively studied 54,146 women, 8,030 of whom had had a prior cesarean delivery, and reported ethnic differences in VBAC success. They found that black women were threefold less likely to experience a successful VBAC trial when compared with white women (adjusted OR 0.37, 95% CI 0.27–0.50). The investigators postulated that perhaps placental factors that predisposed the fetus to labor intolerance explained the difference in success rates between the groups. However, our study found a higher VBAC failure rate in black women, even when adjusting for some of the comorbidities and patient characteristics that have been associated with placental insufficiency, making the theory for this discrepancy postulated by Hollard and colleagues less tenable.
It has been repeatedly shown that the majority of morbidity associated with VBAC is associated with a failed trial, emphasizing the importance of VBAC candidate selection. Regardless of the reason for the small but significantly increased rate of VBAC failure in black women in our cohort, it should follow that black women would experience a greater rate of VBAC-associated morbidity, including uterine rupture. However, this was not the case in our study. Black women were significantly less likely to experience a uterine rupture than other women attempting VBAC. Furthermore, although many of the already highlighted possible differences may also affect the discrepant rate in uterine rupture, there is significant data to support ethnic differences in pelvic connective tissue as demonstrated by the differences in rates of pelvic organ prolapse,13 collagen composition,14 and, perhaps, even wound healing after a hysterotomy.
The large sample size of this cohort study enabled us to look at the effect of race on relatively rare outcomes in patients attempting VBAC, namely uterine rupture, while adjusting for relevant confounding effects. The a priori definition of the primary outcome allowed us to distinguish symptomatic uterine rupture from less clinically severe events. Additionally, we considered the possibility of bias at both patient-level data, and institutional-level data in our statistical approach. Using a multi-level analytic model allowed us to statistically assess and account for potentially significant confounding effects that could arise from clustering effects of hospital site in addition to confounding effects of individual patient characteristics.
Although defining race by self-report has been previously validated, we only considered five categories of race and by doing so may have introduced information bias or bias by ethnicity. Race may be a surrogate marker for complex genetic differences that most likely do not fit into five clear exclusive categories, and thus group assignment may have biased our results. However, this bias would most likely have caused patients with similar genetic make-ups to be considered in different groups in our analysis, most likely biasing the results toward the null. Two other additional potential weaknesses are worth noting. Because the primary cohort study was designed to examine maternal morbidities associated with VBAC attempt, we are unable to assess neonatal outcomes. Additionally, despite statistical considerations of hospital setting as a surrogate marker for practice patterns and controlling for multiple confounding factors, the potential for confounding by indication cannot be completely removed. That is, counseling at the level of the individual patient and practitioner may have biased election to attempt VBAC. This type of selection bias cannot be entirely accounted for statistically. Importantly, although this type of bias may over- or underestimate the true relative risk, it still allows a generalizable or relevant risk estimate of a factor as it is used among other previously known factors in standard clinical practice.
Although the exact biologic or nonbiologic explanation for the decreased rate of uterine rupture in black women undertaking VBAC trials compared with women of other races remains unknown, the clinical implications are very relevant, even in the absence of known sources. Knowing that the rates of VBAC-associated morbidities, such as uterine rupture, vary greatly by race, it may be appropriate to consider more patient-specific risk-benefit counseling. Specifically, a patient may find that the knowledge of being in a higher or lower risk group for VBAC-associated morbidities may influence her preference for delivery mode. Grobman and collegues7 successfully developed a predictive nomogram approach to refining VBAC failure risk estimation for use in counseling individual patients by incorporating multiple risk factors, including race. Future research to develop a bedside predictive tool for uterine rupture or severe morbidity that incorporates race and other risk factors is needed. In short, risks of VBAC trials differ dramatically by race, and a one-size-fits-all approach to planning delivery modality in patients with a history of a prior cesarean should be re-evaluated.
1. Macones GA, Peipert J, Nelson DB, Odibo A, Stevens EJ, Stamilio DM, et al. Maternal complications with vaginal birth after cesarean delivery: a multicenter study. Am J Obstet Gynecol 2005;193:1656–62.
2. Landon MB, Hauth JC, Leveno KJ, Spong CY, Leindecker S, Varner MW, et al. Maternal and perinatal outcomes associated with a trial of labor after prior cesarean delivery. N Engl J Med 2004;351:2581–9.
3. McMahon MJ, Luther ER, Bowes Jr, WA Olshan AF. Comparison of a trial of labor with an elective second cesarean section. N Engl J Med 1996;335:689–95.
4. Srinivas SK, Stamilio DM, Stevens EJ, Odibo AO, Peipert JF, Macones GA. Predicting failure of a vaginal birth attempt after cesarean delivery. Obstet Gynecol 2007;109:800–5.
5. Gonen R, Tamir A, Degani S, Ohel G. Variables associated with successful vaginal birth after one cesarean section: a proposed vaginal birth after cesarean section score. Am J Perinatol 2004;21:447–53.
6. Vinueza CA, Chauhan SP, Barker L, Hendrix NW, Scardo JA. Predicting the success of a trial of labor with a simple scoring system. J Reprod Med 2000;45:332–6.
7. Grobman WA, Lai Y, Landon MB, Spong CY, Leveno KJ, Rouse DJ, et al. Development of a nomogram for prediction of vaginal birth after cesarean delivery. Obstet Gynecol 2007;109:806–12.
8. Cahill AG, Stamilio DM, Odibo AO, Peipert JF, Ratcliffe SJ, Stevens EJ, et al. Is vaginal birth after cesarean (VBAC) or elective repeat cesarean safer in women with a prior vaginal delivery? Am J Obstet Gynecol 2006;195:1143–7.
9. Landon MB, Leindecker S, Spong CY, Hauth JC, Bloom S, Varner MW, et al. The MFMU Cesarean Registry: factors affecting the success of trial of labor after previous cesarean delivery. Am J Obstet Gynecol 2005;193:1016–23.
10. Greenberg MB, Cheng YW, Hopkins LM, Stotland NE, Bryant AS, Caughey AB. Are there ethnic differences in the length of labor? Am J Obstet Gynecol 2006;195:743–8.
11. 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.
12. Hollard AL, Wing DA, Chung JH, Rumney PJ, Saul L, Nageotte MP, et al. Ethnic disparity in the success of vaginal birth after cesarean delivery. J Matern Fetal Neonatal Med 2006;19:483–7.
13. Howard D, Delancey JO, Tunn R, Ashton-Miller JA. Racial differences in the structure and function of the stress urinary continence mechanism. Obstet Gynecol 2000;95:713–7.
© 2008 The American College of Obstetricians and Gynecologists
14. Leppert PC, Catherino WH, Segars JH. A new hypothesis about the origin of uterine fibroids based on gene expression profiling with microarrays. Am J Obstet Gynecol 2006; 195:415–20.