J. CHRISTOPH RAGETH, MD, CLAUDIA JUZI, MD and HEIDI GROSSENBACHER, MD FOR THE SWISS WORKING GROUP OF OBSTETRIC AND GYNECOLOGIC INSTITUTIONS
From the Spital Limmattal, Schlieren, and Kantonsspital Münsterlingen, Münsterlingen, Switzerland.
Objective: To examine the risks of vaginal delivery after previous cesarean and to find criteria to help decide whether a trial of labor or an elective repeat cesarean should be preferred.
Methods: We evaluated 29,046 deliveries after previous cesarean registered in a pooled database of 457,825 deliveries used to assess quality control in gynecology and obstetrics departments in Switzerland.
Results: Among the 17,613 trial-of-labor cases logged (attempt rate 60.64%), the success rate was 73.73% (65.56% after inducing labor and 75.06% after the spontaneous onset of labor). The following complications were significantly more frequent in the previous-cesarean group: maternal febrile episodes (relative risk [RR] 2.77; 95% confidence interval [CI] 2.52, 3.05), thromboembolic events (RR 2.81; CI 2.23, 3.55), bleeding due to placenta previa during pregnancy (RR 2.06; CI 1.70, 2.49), uterine rupture (92 cases; RR 42.18; CI 31.09, 57.24), and perinatal mortality (118 cases, including six associated with uterine rupture; RR 1.33; CI 1.10, 1.62). The postcesarean group also showed a 0.28% rate of peripartum hysterectomy (81 cases; RR 6.07; CI 4.71, 7.83). There was one maternal death in the group, compared with 14 maternal deaths in the group without previous cesarean (no statistical significance). The risk of uterine rupture for patients with previous cesareans was elevated in the trial-of-labor group compared with the group without trial of labor (RR 2.07; CI 1.29, 3.30), but all other maternal risks, including peripartum hysterectomy (RR 0.36; CI 0.23, 0.56), were lower. When comparing the women having a trial of labor, the 70 with uterine rupture more often had induced labor (24.29% compared with 13.92% in the nonrupture group; P = .013), had epidural anesthesia (24.29% compared with 8.44%; P rate tracing (32.86% compared with 8.53%; P had failure to progress (21.43% compared with 7.98%; P = .001).
Conclusion: A history of cesarean delivery significantly elevates the risks for mother and child in future deliveries. Nonetheless, a trial of labor after previous cesarean is safe. Induction of labor, epidural anesthesia, failure to progress, and abnormal fetal heart rate pattern are all associated with failure of a trial of labor and uterine rupture.
Source: Obstetrics & Gynecology 1999;93:332-337
In the absence of contraindications, a trial of labor after previous cesarean is generally endorsed,1 and several studies support this approach.2–8 In previous studies, trial-of-labor success rates ranged between 60% and 90%; the rate of repeat cesareans ranged from 10% to 40%, increasing to up to 60% in women who had had two or more cesareans.2,4–17 Even after two or more previous cesarean deliveries, the success rate remained relatively high.2,7,10,14 Nevertheless, the percentage of women opting for a trial of labor after previous cesarean varied considerably (16.3–90%).9 The problem remains the rate of nonelective repeat cesareans, particularly when the procedure has to be performed as an emergency measure because of complications during labor. Recent studies have indicated that a trial of labor may lower the incidence of minor complications (eg, wound infections) but not the number of major complications (eg, uterine rupture, septicemia, and hysterectomy), many of which occur during unsuccessful trials of labor.5 An unsuccessful trial of labor must be considered a failure because it merely adds to the expenditure and frustration for both patients and medical staff.
Numerous authors have offered criteria to help decide whether a trial of labor might be successful12,15–20; however, there are very few reliable criteria that indicate when a trial of labor will fail.19 In this study, we examined the risks of delivery after previous cesarean and tried to identify additional criteria to help decide whether a trial of labor or an elective repeat cesarean should be proposed to an individual patient.
Since 1983, the Swiss Working Group of Obstetric and Gynecologic Institutions has used questionnaires to collect joint prospective data for quality-control purposes. That database consists of 457,825 admissions with deliveries, representing approximately 40% of the deliveries in Switzerland from 1983 through 1996. After excluding twin pregnancies, we found 255,453 patients with a parity of greater than 1 (the reference number in the calculations that follow). Of the 255,453 patients, 29,046 had a previous cesarean delivery. The questionnaires were completed by the treating doctors when the patient was released from the hospital. In addition to information about the delivery, we collected various data regarding the patient’s history (risk factors such as smoking, drugs, allergies, cardiovascular diseases, and socioeconomic status). A customized program featuring a filtering function was used to evaluate the records. The method of storing and retrieving the data unfortunately did not allow multivariate analysis. For significance comparisons, we used the 2 test and calculated the confidence intervals (CIs) for the relative risk (RR).
As a first step, we compared the following groups: 1) patients with previous cesarean delivery and those without previous cesarean delivery, 2) patients who chose a trial of labor and those who did not choose a trial of labor (elective repeat cesarean), 3) patients who chose a trial of labor who needed to have induction of labor and those having a trial of labor who started spontaneous contractions, and 4) patients with a successful trial of labor and those with an unsuccessful trial of labor leading to emergency repeat cesarean. We then examined the uterine-rupture group and the perinatal-death group in more detail. Finally, we tried to assess the relevant differences between the groups and to set up new guidelines to improve decision making as to when an elective repeat cesarean should be the preferred option.
During the observation period, 457,825 deliveries were recorded. Among the 255,453 deliveries in women with a parity of greater than 1, 29,046 (11.37%) had a previous cesarean. The rate of choosing a trial of labor was 60.64% (Table 1). Within that trial-of-labor group, 15,154 women (86.04%) had spontaneous labor and 2459 (13.96%) required induction of labor. The success rate was 75.06% (11,374 of 15,154) for the spontaneous-labor group compared with 65.56% (1612 of 2459) for the induced-labor group. The average success rate was thus 73.73% (12,986 of 17,613).
Comparison between the trial-of-labor group and the elective repeat-cesarean group (Table 3) showed that maternal risks were consistently lower in the former, with the exception of uterine rupture. Perinatal death was, however, significantly higher for the trial-of-labor group (RR 1.74; 86 cases or 0.49%). A more detailed study of these infants revealed that 36 of them (41.86%) were less than 28 weeks of gestational age (compared with 18.75% in the group with primary repeat cesarean) and that several had malformations. Looking solely at the infants born after 28 weeks of gestational age without malformations in the previous-cesarean group, we found 43 cases of perinatal death: 10 in the primary repeat-cesarean group and 33 in the trial-of-labor group (RR 2.14; 95% CI 1.07, 4.27). This difference remained marginally statistically significant. Of these 43 cases of perinatal death, six were associated with uterine rupture (one in the elective repeat-cesarean group and five in the trial-of-labor group), nine were due to placental abruption (five occurred during labor and four before admission to the hospital, so that emergency repeat cesareans were performed), one was associated with placenta previa, three were linked to preeclampsia, 12 were caused by amniotic infection or premature rupture of membranes with or without preterm deliveries and respiratory distress syndrome, and 12 had differing reasons.
Table 4 shows that the following risk factors were statistically significant for an unsuccessful trial of labor: fetal malpresentation (RR 3.83; CI 3.53, 4.14), induced labor (RR 1.47; CI 1.37, 1.59), and fetal weight of 4000 g or greater (RR 1.45; CI 1.33, 1.58). The overall success rate in the group with fetal malpresentation was 42.31%; the subgroup with occipitoposterior presentation had a success rate of 57.96% and the subgroup with breech presentation had a success rate of 33.51%. The success rates after induced labor (65.56%) or with fetal weight of 4000 g and greater (65.93%) were still high. Combining the need to induce labor and fetal macrosomia of over 4000 g, however, led to a rate of only 57.02% for successful vaginal deliveries. In contrast, augmenting labor was not a risk factor for a failed trial of labor.
Uterine rupture was found in 0.32% (92 of 29,046 patients) of the previous-cesarean group, in 0.40% of the trial-of-labor group, and in 0.65% of the group after induction of labor. Analysis of the uterine-rupture group did not show a higher incidence of cephalopelvic disproportion (seven cases or 7.6% compared with 1696 or 5.9%; RR 1.30; CI 0.63, 2.67; P = .47) or of high birth weight (4000 g and above) (12 cases or 13.04% compared with 2888 or 9.97%; RR 1.31; CI 0.76, 2.24; P = .327). In contrast (Table 5), epidural anesthesia during labor was given more often in the group with uterine rupture (17 [24.29%] compared with 1480 [8.44%]; P rates of induced labor (RR 1.74; 17 [24.29%] compared with 13.92% in the nonrupture group; P = .013), but augmenting labor was not found to be a significant risk factor for uterine rupture (29 [41.43%] compared with 35.80% in the group without rupture; P = .327). Abnormal fetal heart rate tracing (23 [32.86%]; 3.85 times higher than in the group without rupture; P and failure to progress (15 [21.43%]; 2.69 times higher than in the group without rupture; P with uterine rupture.
A history of cesarean delivery implies a considerably elevated risk for a variety of peripartal complications for both mother and child, such as an increased frequency of extrauterine pregnancy,21 the necessity for hysterectomy, and febrile and thromboembolic complications.
Reports have described increased placental implantation disturbances (placenta previa, placental abruption) related to cesarean delivery21–24 (also Green R, Gardeil F, Turner MJ. Long-term implications of cesarean section [Letter]. Am J Obstet Gynecol 1997;176:254–5). Our data confirm these findings. We found a 2.06-fold elevated risk of vaginal bleeding with placenta previa during pregnancy in women who had a previous cesarean. We also confirmed an elevated risk of placental abruption in these women (during pregnancy, RR 1.87; during labor, RR 1.49).
Maternal death can occur from uterine rupture or placenta percreta after previous cesarean.24–26 According to the literature, uterine rupture occurred in approximately 0.5–0.8% of trials of labor (up to 1.5% when including bloodless dehiscence).4,11,27 In our trial-of-labor group, the risk of rupture was 0.40% (70 of 17,613), and in the primary repeat-cesarean group, it was 0.19% (22 of 11,433). This is a 42-fold higher risk for the previous-cesarean group compared with the group without previous cesarean (17 of 226,407, or 0.0075%). Uterine rupture during labor frequently manifests itself in fetal bradycardia or failure to progress.2,11 In our study, induction of labor raised the risk for uterine rupture from 0.40% to 0.65%. This ratio is still low, but we believe labor should be induced only if a clear indication is given. Epidural anesthesia was also associated with uterine rupture, but it remains unclear whether this is an independent risk factor because we could not perform a multivariate analysis.
Hillan28 demonstrated a marked increase in febrile morbidity after emergency cesarean. Our study revealed not only significantly elevated maternal risks (hysterectomy, thromboembolic complications, febrile morbidity, and maternal transfer to another hospital or another department) but also significantly elevated perinatal risks (5-minute Apgar score below 5, arterial pH below 7.00, neonatal transfer after birth, and perinatal death) after previous cesarean.
The frequency of trial of labor varies considerably among institutions (16.3–90%).9 Hueston and Rudy18 found that women undergoing a trial of labor were more likely to be younger, nonwhite, unmarried, living in households where all members were unemployed, and lacking private insurance. Our results confirm these findings concerning age, marital status, and insurance.
The overall success rate for vaginal birth was 73.73% (65.56% in the trial-of-labor group with induced labor and 75.06% in the trial-of-labor group without induced labor). The trial-of-labor group had fewer incidences of febrile morbidity and thromboembolic complications. However, perinatal mortality in infants older than 28 weeks’ gestation and without malformations was elevated: 0.19% compared with 0.09% in the elective repeat-cesarean group (RR 2.14; CI 1.07, 4.27; P = .031). Although this is marginally statistically significant, the absolute number is small and is in accordance with the generally slightly elevated risk for the infant during vaginal delivery compared with elective cesarean delivery. On the other hand, neonatal transfer was required less often. Uterine rupture occurred more often, but the necessity for peripartal hysterectomy did not, in contrast to the findings of McMahon et al.5 Thus, we conclude that a trial of labor is a safe procedure.
Weinstein et al17 found that a few factors can lead to a successful trial of labor, including the Bishop score, history of previous vaginal deliveries, and any of the following as the reason for previous cesarean: breech presentation, preeclampsia, multiple pregnancy, and placenta previa. McMahon et al5 found an increased rate of cesarean after a failed trial of labor when the maternal age was 35 years or older, the delivery took place at a community or regional hospital, the infant’s birth weight was greater than 4000 g, and the woman had no previous vaginal delivery. Learman et al12 studied 175 trials of labor and found that the risk factors for failure were induced labor and a high fetal station, but even these women had high rates of vaginal deliveries (67% and 75%, respectively). Only one subgroup, with both induced labor and large fetuses, had a 75% risk of cesarean delivery. We found similar results in our group of women who had induced labor, with a 65.56% success rate. In the group with the combination of fetal macrosomia and induced labor, we found a trial-of-labor success rate of 57.02%, which is not very encouraging. Thurnau et al20 studied a scoring system in which the fetal head and abdominal circumferences (by ultrasonographic measurement) were compared with the maternal pelvic inlet and midpelvic circumferences (by X-ray pelvimetry) and found that cesareans were likely if these measurements were unfavorable. Flamm and Geiger29 developed an admission scoring system. A trial of labor was more often successful when the patient was younger than 40 years (odds ratio [OR] 2.58), there was a history of vaginal birth (after first cesarean, OR 3.39; before first cesarean, OR 1.53; before and after first cesarean, OR 9.11), the reason for the first cesarean was other than failure to progress (OR 1.93), cervical effacement was present (more than 75%, OR 2.72; 25–75%, OR 1.79), and cervical dilatation was 4 cm or more upon admission to the hospital (OR 2.16).
A trial of labor should not be attempted when the patient has an unknown uterine scar type, a history of uterine rupture or scars in the upper segment of the uterus, absolute cephalopelvic disproportion, placenta previa, severe myopia complicated by retinal detachment, or fetal malpresentation incompatible with a safe vaginal delivery.17 The American College of Obstetricians and Gynecologists has published guidelines for these contraindications.1 In patients with a history of multiple cesareans, the risk for uterine rupture is even higher and has been associated with fetal death and serious neonatal disorders.4–6,11,24–27,30
One of the main problems seems to be the quick decision to perform the first cesarean. Consideration of cesarean should include not only the direct risks, but also the potential for late sequelae. Because our study was not randomized and our questionnaires did not ask why previous cesareans were performed, we cannot discuss the problem any further. We would like to point out that the overall frequency of cesareans in our working group was only 14% during the past 5 years (19,833 cesareans among 141,212 deliveries).
Our data show that a trial of labor after previous cesarean is safe and can be recommended in the majority of cases. Because the success rate for a trial of labor is only 57.02% when fetal macrosomia (greater than 4000 g) is combined with the need for inducing labor, we recommend an elective repeat cesarean in these situations. Although epidural anesthesia and induction of labor are associated with uterine rupture, we believe that a previous cesarean is not a strict contraindication for epidural anesthesia or induction of labor.
The Swiss Working Group of Obstetric and Gynecologic Institutions is an association of obstetrics and gynecology departments from public hospitals in Switzerland. In 1996, there were 84 members.
Received April 13, 1998. Received in revised form September 2, 1998. Accepted September 10, 1998.
1. American College of Obstetricians and Gynecologists. Guidelines for vaginal delivery after a previous cesarean birth. ACOG committee opinion no. 64. Washington DC: American College of Obstetricians and Gynecologists, 1988.
19. Pickhardt MG, Martin JN, Meydrech EF, Blake PG, Martin RW, Perry KG, 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.
20. Thurnau GR, Scates DH, Morgan MA. The fetal-pelvic index: A method of identifying fetal-pelvic disproportion in women attempting vaginal birth after previous cesarean delivery. Am J Obstet Gynecol 1991;165:353–8.
Table 3. Comparison of the Group of Patients With Elective Repeat Cesarean and the Group With a Trial of Labor
Table 5. Analysis of the Trial-of-Labor Patients Having Uterine Rupture