Materials and Methods
We reviewed medical records of all parturients admitted for trials of labor after cesarean at Brigham and Women's Hospital from July 1984 through June 1996. We abstracted data on current trials of labor and prior cesareans. Our study design and record review have been reported.1 We also reviewed records of all nulliparas who delivered from December 1994 to August 1995.
The prior cesarean group was limited to women with one prior cesarean because additional deliveries, either vaginal or cesarean, might affect risks of cesarean and lengths of labor. Women were included who had singleton, cephalic, term pregnancies and were in spontaneous labor.
Women with single prior cesareans were categorized by the indications for the primary cesarean: breech, failure to progress, nonreassuring fetal testing, and other. The last group included those with unknown prior cesarean indications. If women had multiple indications for their cesarean, the primary indication was used. Most women had their prior deliveries at our institution, so prior cesarean indications were usually determined from the medical records. If records were not available, indications were determined from the prenatal record for the index pregnancy.
The overall rates of cesareans for nulliparas were compared with rates of women with single prior cesareans. The rates of cesareans for each prior indication subgroup were then compared with the rates in nulliparas. Cesarean rates for failure to progress and nonreassuring fetal testing in the index pregnancy are also reported for each prior indication subgroup.
Demographic and pregnancy characteristics of women with prior cesareans were compared with those of nulliparas. Statistical significance for comparisons of categoric variables was determined using χ2 test or Fisher exact test, as appropriate. Continuous variables were compared using the Student t test. Statistical significance was defined as P < .05.
To evaluate the effects of prior cesarean indications on labor management, we determined the lengths of labor before cesareans for those women who had repeat cesareans for failure to progress in the index pregnancy. Labor length was defined as the time from admission to delivery. Labor lengths by prior indications were compared using analysis of variance. Post hoc comparisons were done with the Tukey test.
Multiple logistic regression analyses were done to determine the risk of cesarean after a trial of labor overall and according to the prior indication, while controlling for confounding factors. Because the time periods for the two groups differed (1984–1996 for prior cesareans and 1994–1995 for nulliparas), we also performed an analysis that included only women who delivered in 1995.
The prior cesarean group comprised 2207 women and the nulliparous group 1617. Demographic data for the two groups are shown in Table 1. Nulliparas were younger than women with prior cesareans. The rates of cesarean were 28.7% (634 of 2207) for the prior cesarean group and 13.5% (219 of 1617) for the nulliparas (P = .001). Labor characteristics of the index pregnancies are presented in Table 2. Although women in the two groups delivered at similar gestational ages, women with prior cesareans had larger infants, were less likely to be admitted with cervical dilatation more than 4 cm, were slightly less likely to have ruptured membranes on admission, were less likely to have oxytocin augmentation, and were more likely to have epidural anesthesia during labor.
Multiple logistic regression was done to determine the risk of cesarean delivery among women with prior cesareans compared with nulliparas while controlling for maternal age, admission cervical dilatation, length of labor, public-aid insurance, artificial rupture of membranes, use of epidural, use of oxytocin, rupture of membranes at admission, and birth weight. Women who had prior cesareans were more than twice as likely to have cesareans again as were nulliparas (odds ratio [OR] 2.6; 95% confidence interval [CI] 2.2, 3.2). We also performed a multiple logistic regression analysis on women with trials of labor after cesareans in 1995 (the year in which the data for nulliparas were collected). The association for this period was similar to that for the overall population (OR 2.2; 95% CI 1.5, 3.3).
Cesarean rates by indications for the prior cesareans were compared with rates for nulliparas (Table 3). When the first cesarean was done for breech presentation, the rate of repeat cesarean was 13.9%, similar to the rate for nulliparas (13.5%). Cesarean rates in index pregnancies were highest when the first cesareans were done for failure to progress (37.3%). Rates were intermediate when prior indications were nonreassuring fetal testing (25.4%) or other (24.8%).
Multiple logistic regression was done to determine the risk of cesarean after a trial of labor compared with the rate in nulliparas according to the indications for prior cesareans. The model controlled for maternal age, gestational age, dilatation at admission, length of labor, birth weight greater than 4000 g, and use of epidural anesthesia. The cesarean rates for each subgroup of women with previous cesarean deliveries were compared with those of nulliparas. Women who had prior cesareans for failure to progress were four times more likely to have cesareans as nulliparas (OR 4.5; 95% CI 3.6, 5.5); women whose prior cesareans were for nonreassuring fetal testing were twice as likely to have cesareans again as nulliparas (OR 2.2; 95% CI 1.6, 2.9). Women with prior cesareans for breech had a risk for cesarean similar to that of nulliparas (OR 0.95; 95% CI 0.7, 1.3).
Failure to progress and nonreassuring fetal testing were the two major indications for cesarean in the index pregnancy for both groups (Table 3). Nulliparas and women with prior indications for breech had similar cesarean rates for failure to progress and nonreassuring fetal testing in index pregnancies. Cesarean rates for failure to progress and nonreassuring fetal testing were higher when the previous cesareans were for failure to progress or nonreassuring fetal testing.
Mean lengths of labor before cesareans for failure to progress in the index pregnancies were shorter for women whose prior cesareans were for failure to progress (11.5 ± 5.8 hours) than for women whose prior cesareans were for nonreassuring fetal testing (13.4 ± 5.7 hours), breech (13.9 ± 5.6 hours), and other indications (15.1 ± 7.2 hours). Analysis of variance found significant differences in the lengths of labor by prior indication (P = .003). Tukey test indicated significant differences (P < .05) in lengths of labor for women with prior indications of failure to progress compared with the other three groups (breech, nonreassuring fetal testing, and other).
Women who had trials of labor after one prior cesarean and no other deliveries were at increased risk for repeat cesareans compared with women having their first labors. The rate of repeat cesarean was strongly associated with the indication for the prior cesarean. The risk of repeat cesareans for women with prior cesareans for breech was essentially identical to that of nulliparas. Women with prior cesareans for nonreassuring fetal testing had two times the cesarean risk of nulliparas. Women with prior cesareans for failure to progress had the highest risk of cesarean.
Previous reports have noted lower rates of repeat cesareans after trials of labor for women with prior cesareans for breech.2 Our similar findings support the notion that breech presentation is a nonrecurring indication. Breech was the only prior cesarean indication in this study for which the rate was similar to nulliparas. Because women with breech fetuses do not typically labor at our institution, this group (in contrast to women with prior cesareans for other indications) likely consisted of women who had never (or minimally) labored. The similar cesarean rates between this group and nulliparas suggest that a scarred uterus alone does not alter the risk for cesarean.
The influence of prior cesareans for nonreassuring fetal testing on repeat cesarean risks is controversial.2–5 Our data suggest that women with prior cesareans for nonreassuring fetal testing have higher risks of cesareans in subsequent trials of labor, so nonreassuring fetal testing should not be considered a completely nonrecurring indication for cesarean. These women were at an increased risk of cesarean for both failure to progress and nonreassuring fetal testing. The high rate of repeat cesareans among those with prior cesareans for failure to progress (37.3%) is consistent with previous reports and primarily was caused by an increased risk of repeat cesarean for failure to progress.5–8
The limitations of our study deserve mention. We included only women with singleton pregnancies in term spontaneous labor, so our findings might not be applicable to women with induced labor, multiple gestations, or preterm delivery. Our results also might not apply to women with prior vaginal deliveries in addition to prior cesareans. We chose to look at gravidas with no prior vaginal deliveries because those deliveries affect the risk of repeat cesarean and alter the lengths of labor.9 We collected our data over two different periods (1984–1996 for those with prior cesareans and 1994–1995 for nulliparas), so changes in practice could have influenced our results. To address this issue, we performed an analysis that compared nulliparas from 1995 with those who had trials of labor after cesareans during the same year. The results of the multiple logistic regression analysis did not differ from those for the overall population.
We found that among those who had repeat cesareans for failure to progress in the index pregnancies, the durations of labor were approximately 2.5 hours shorter if their prior cesareans were also for failure to progress than if they were for other indications. It is possible that health care providers intervened with repeat cesarean earlier in women with prior cesareans for failure to progress than in those with prior cesareans for breech. Further study is required to determine whether permitting these women to labor longer would be associated with lower repeat cesarean rates and whether it would be safe.
We found that the rate of cesarean after a trial of labor overall was higher than in nulliparas, but this varied according to the indications for prior cesareans. Women with prior cesareans for failure to progress had the highest rates, whereas women with prior cesareans for breech presentation were no more likely than nulliparas to have repeat cesareans, which suggests that this is a nonrecurring indication for cesarean. The success rates of trials of labor after cesarean varied widely according to the indications for the prior cesareans, so this history provides an even better estimate of the likelihood of successful vaginal births after cesareans. The length of trials of labor for gravidas who had cesareans for failure to progress was related to the indications for prior cesareans and was shortest for women whose prior cesareans were for failure to progress, suggesting that physicians might intervene earlier in these cases.
1. Caughey AB, Shipp TD, Repke JT, Zelop C, Cohen A, Lieberman E. Trial of labor after cesarean delivery: The effect of previous vaginal delivery. Am J Obstet Gynecol 1998;179:938–41.
2. Clark SL, Eglinton GS, Beall M, Phelan JP. Effect of indication for previous cesarean section on subsequent delivery outcome in patients undergoing a trial of labor. J Reprod Med 1984;29:22–5.
3. Ollendorff DA, Goldberg JM, Minogue JP, Socol ML. Vaginal birth after cesarean section for arrest of labor: Is success determined by maximum cervical dilatation during the prior labor? Am J Obstet Gynecol 1988;159:636–9.
4. Flamm BL, Goings JR, Liu Y, Wolde-Tsadik G. Elective repeat cesarean delivery versus trial of labor: A prospective multicenter study. Obstet Gynecol 1994;83:927–32.
5. Benedetti TJ, Platt L, Druzin M. Vaginal delivery after previous cesarean section for a nonrecurrent cause. Am J Obstet Gynecol 1982;142:358–9.
6. Hoskins IA, Gomez JL. Correlation between maximum cervical dilatation at cesarean delivery and subsequent vaginal birth after cesarean delivery. Obstet Gynecol 1997;89:591–3.
7. Flamm BL. Vaginal birth after cesarean section: Controversies old and new. Clin Obstet Gynecol 1985;28:735–44.
8. Seitchik J, Rao VRR. Cesarean delivery in nulliparous women for failed oxytocin-augmented labor: Route of delivery in subsequent pregnancy. Am J Obstet Gynecol 1982;143:393–7.
9. Flamm BL, Geiger AM. Vaginal birth after cesarean delivery: An admission scoring system. Obstet Gynecol 1997;90:907–10.