Algert, Charles S. MPH1,2; Morris, Jonathan M. MBChB, PhD1,2; Simpson, Judy M. PhD3; Ford, Jane B. PhD2,3; Roberts, Christine L. MBBS, DrPH2
Internationally, concern about the increasing rate of cesarean delivery, both primary and repeat, has resulted in much research into the risks associated with attempted vaginal birth after cesarean (VBAC). An important consideration when counseling women contemplating a trial of labor for VBAC is the risk of uterine rupture.1 The risk of uterine rupture is increased when VBAC is attempted compared with elective repeat cesarean delivery.1,2 However, the absolute risk of uterine rupture is low, and guidelines recommend that most women with one prior cesarean delivery with a low-transverse incision should be offered a trial of labor.3,4
Because the advantages of a vaginal delivery after a primary cesarean delivery are less certain than in previous decades, it has become more important to identify women who may be at increased risk of an adverse event if VBAC is attempted. Despite the plethora of research on risk factors for uterine rupture, only a few studies have examined potentially predictive factors that pertain to the primary cesarean delivery, and none of these looked at the onset of labor for that pregnancy.5–7 This is despite the fact that both induction and augmentation of labor for a trial of labor have been associated with an increased risk of uterine rupture in most studies.1,2,5,8–11 Prior vaginal birth has been much studied as a risk factor for uterine rupture in birth after cesarean delivery,5,9,12–15 but the generalizability of these results has been unclear because most studies include some women who already had a successful VBAC.
Our aim was to estimate the effect of factors from the primary cesarean delivery that might be associated with uterine rupture in a trial of labor for the subsequent delivery.
MATERIALS AND METHODS
Data were obtained from two linked, anonymized New South Wales, Australia, population databases. New South Wales is the largest state in Australia, with approximately 85,000 births per year. The New South Wales Midwives Data Collection is a legislated surveillance system of all births reported by the attending midwife or doctor. The New South Wales Admitted Patient Data Collection has records of all hospital discharges in New South Wales, including International Classification of Diseases (ICD) diagnosis codes related to the admission. Women with a primary (first) cesarean delivery were identified from the records of all deliveries from 1994–2004. Those women who then could be linked to a subsequent delivery in the period 1998–2004 were retained as potentially eligible for the trial-of-labor study population. The study was approved by the New South Wales Population and Health Services Research Ethics Committee and the University of Sydney Human Research Ethics Committee.
Deliveries immediately after a primary cesarean delivery were eligible for inclusion as an attempted trial of labor if the birth was a singleton between 37 and 41 weeks, inclusive. Exclusions included breech presentations, grand multiparae, and women whose first cesarean delivery was a classical cesarean.3 Cesarean deliveries after spontaneous labor at 37, 38, or 39 weeks but in which failure to progress was not recorded were excluded unless there were ICD-10 diagnostic codes indicating long or complicated labor because it was uncertain whether a trial of labor was intended. We have previously demonstrated that, among women who had planned to deliver by cesarean, 11% go into labor before the planned date.16
Where “trial of labor” is used in the Materials and Methods and Results sections, it refers specifically to the birth immediately after the primary cesarean delivery. Onset of labor at the trial-of-labor birth was categorized as spontaneous, augmented spontaneous, or induced. Other pregnancy and delivery factors at the trial-of-labor birth were years between the primary cesarean delivery and the trial-of-labor birth, maternal medical conditions (preeclampsia, other hypertensive disorder, gestational or preexisting diabetes), and birth weight of the newborn. Women who had one to three vaginal births before their primary cesarean delivery were categorized as prior vaginal births. To estimate the effect of factors from the primary cesarean delivery on the subsequent trial-of-labor birth, the onset of labor for the primary cesarean delivery was categorized as spontaneous, induced, or without labor. The primary cesarean delivery also was categorized as either preterm (less than 37 weeks) or term and the delivered birth weight as less than 4,000 g or 4,000 g or more.5 When compared with medical records, the delivery factors analyzed in this study are reliably reported on the Midwives Data Collection.17 Only three deliveries in the trial-of-labor study population did not have the onset of labor recorded, and only two had mode of delivery information missing.
Women were categorized as suffering a uterine rupture if their hospital discharge data contained the ICD-10 diagnostic codes O71.0 (rupture of uterus before onset of labor) or O71.1 (rupture of uterus during labor). A New South Wales validation study of maternal morbidity18 included ascertainment of whether uterine ruptures reported in medical records also were included in the hospital discharge database. Four uterine ruptures were found in the sample of 1,200 deliveries, all of which were identified in the hospital discharge database; two cases of asymptomatic dehiscence were noted at cesarean delivery and neither was coded as a rupture.
Crude relative risks (RRs) and 95% confidence intervals (CIs) for uterine rupture were calculated for exposures in both the trial-of-labor birth and the primary cesarean delivery. Adjusted odds ratios were determined using multivariable logistic regression models and used as estimates of the adjusted RRs. Labor for the trial-of-labor birth was simplified as either spontaneous or induced/augmented; for the primary cesarean delivery it was simplified as either spontaneous or induced/elective cesarean delivery based on similarity of risk for these factors. An indicator representing patient payment status (private or public) also was included in the models. The final model of uterine rupture included only those factors from the primary cesarean delivery and trial-of-labor deliveries that added a statistically significant contribution. Finally, we performed a sensitivity analysis in which spontaneous labor for the primary cesarean delivery was replaced by an indicator for prior vaginal birth and/or spontaneous labor for the primary cesarean delivery, compared with women who had neither.
From 1994 to 2004, there were 940,939 deliveries in New South Wales, of which 198,897 (21.1%) were cesarean. Of these, 113,820 could be definitely identified as primary cesarean deliveries and 28,552 could be linked to a subsequent birth in the period 1998–2004 and satisfied the study criteria. For these linked births after cesarean delivery, 17,076 (59.8%) were delivered by repeat cesarean without labor and 1,316 (4.6%) were a second cesarean delivery at 37, 38 or 39 weeks after spontaneous labor and could not definitely be categorized as an intended trial of labor. This left 10,160 (35.6%) women who attempted a trial of labor after their primary cesarean delivery (Fig. 1). Characteristics of both the trial-of-labor birth and the preceding primary cesarean delivery are shown in Table 1. Of the women who attempted a trial of labor, 8,456 (83.2%) had spontaneous labor, 1,701 (16.8%) were induced, and three were “missing” for onset of labor. Women who were induced had higher rates of any hypertensive disorders (12.4%) and of any (gestational or preexisting) diabetes (6.4%). These rates were similar to the rates of any hypertension (6.4%) and any diabetes (6.9%) among the 17,076 women who delivered by planned repeat cesarean. Women in whom labor was induced were also less likely to have had spontaneous labor before their primary cesarean delivery than women who did not require induction for their trial of labor (28.4% compared with 47.0%). Of the women who had a prior vaginal birth, 529 (54.9%) had a planned cesarean delivery for their primary cesarean delivery and 280 (52.9%) of these were nonvertex presentations. Of women with no prior vaginal birth, 2,915 (31.7%) had a planned cesarean delivery for their primary cesarean delivery and 1,763 (60.8%) of these were nonvertex presentations.
Overall, 65.3% of the trials of labor resulted in a vaginal birth. Rates of VBAC were similar for the 7,330 women with unaugmented spontaneous labor (66.5%) and the 1,091 women with augmented labor (67.1%). For the 1,334 women whose labor was induced with oxytocin without prostaglandins, the VBAC rate was 61.8%. Induction including prostaglandins resulted in a VBAC rate of 50.2% for 299 women. Use of prostaglandins decreased rapidly during the study period: 5.4% of the trial-of-labor births recorded use of prostaglandins in 1998, but only 1.3% recorded prostaglandins in 2004. For the 59 women recorded as being induced by “other” techniques (eg, Foley catheter), the VBAC rate was 42.4%. Taking into account obstetric history, the 4,124 women who had spontaneous labor before their primary cesarean delivery but no prior vaginal birth had a VBAC rate of 64.6%. The 3,364 women who did not have this history but who did go into spontaneous labor for their trial of labor had a VBAC rate of 64.5%. The 965 women who had a prior vaginal birth had a VBAC rate of 84.8%. The 1,704 women who had no history of spontaneous labor or vaginal birth, and who required induction or augmentation to proceed with a trial of labor, had a VBAC rate of 60.0%.
There were 39 uterine ruptures recorded among the 10,160 women who attempted a trial of labor, for an overall rate of 0.38%. Compared with spontaneous labor, trial of labor that was induced or augmented had significantly increased crude risks of uterine rupture: augmented labor RR 4.09 (95% CI 1.79–9.32), induced with oxytocin RR 4.42 (95% CI 2.07–9.41), induced with prostaglandins RR 4.84 (95% CI 1.41–16.61). For any induced/augmented compared with spontaneous trial of labor, the uterine rupture RR was 4.24 (95% CI 2.23–8.07). The onset of labor for the primary cesarean delivery also had a statistically significant effect on the crude risk of uterine rupture in the subsequent trial-of-labor birth. Using spontaneous labor before the primary cesarean delivery as the comparison group, the risks of uterine rupture in the subsequent trial of labor were: induced labor RR 2.42 (1.00, 5.83), planned primary cesarean delivery RR 2.73 (1.24, 6.03). The combined risk if the primary cesarean delivery was preceded by either induction or no labor, compared with spontaneous labor, was RR 2.61 (95% CI 1.24–5.49). Crude and adjusted risks of uterine rupture, by maternal and pregnancy characteristics of both the trial-of-labor birth and the primary cesarean delivery, are shown in Table 2. Maternal hypertensive disorders and diabetes in the trial-of-labor pregnancy, birth weight for the trial of labor, and prior vaginal birth were not statistically significant predictors in any of the logistic regression models.
Of the factors related to the primary cesarean delivery (onset of labor, preterm delivery, birth weight, interval from primary cesarean delivery to trial of labor), only onset of labor was retained as statistically significant. The final model of longitudinally linked deliveries showed that, even after adjusting for maternal age, patient payment status, and onset of labor for the trial-of-labor birth, there was an independent effect of the onset of labor before the primary cesarean delivery on the risk of uterine rupture. The adjusted RR of uterine rupture in the trial-of-labor pregnancy was 2.25 (95% CI 1.06–4.79) if there was no labor before the primary cesarean delivery or labor was induced. Conversely, if the onset of labor before the primary cesarean delivery was spontaneous, the adjusted RR of uterine rupture was 0.45 (95% CI 0.21–0.95) compared with induction or planned cesarean delivery. The sensitivity analysis of the larger group of women who had either spontaneous labor before their primary cesarean delivery and/or a prior vaginal birth, compared with women who had neither, resulted in an adjusted RR of uterine rupture of 0.44 (95% CI 0.22–0.90). In absolute terms, among women with a history of either spontaneous labor or vaginal birth, there was one uterine rupture for every 460 deliveries (0.2%). Among those without this history who also required induction or augmentation to proceed with their trial of labor, there was one uterine rupture for every 95 deliveries (1.1%).
Our results indicate that a history of spontaneous labor before a primary nonclassical cesarean delivery may more than halve the risk of uterine rupture in a subsequent trial of labor. This is in addition to the already recognized reduction in risk if the trial-of-labor delivery commences with spontaneous labor.1,2,5,8–11 Our results will help inform women about their individual risk of uterine rupture when they deliberate over mode of delivery in a pregnancy after a primary cesarean delivery. In particular, these data provide encouraging support to those women who have a history of spontaneous labor and who are considering a first trial of labor.
The study population was a carefully selected cohort of women who could be categorized as having attempted a trial of labor at term for the birth immediately after a primary cesarean delivery. The study population did not include women who had already had one or more successful VBAC deliveries, which would have affected their risk of uterine rupture.14 The overall rate of uterine rupture of 0.38% in this trial-of-labor population is consistent with values reported by other studies in which uterine rupture rates vary greatly, ranging from 1% or higher among women who required an induction for a trial of labor1,9,10 down to 0.1% in a study where the protocol was to wait for spontaneous labor or else recommend cesarean delivery if the cervix was unripe.19
The mechanism by which labor protects against uterine rupture could be that suggested by a study of prior preterm birth: that labor might result in a better developed lower uterine segment, such that the uterine incision during cesarean delivery is less likely to involve contractile muscle.7 That study reported a tendency toward increased risk of uterine rupture for prior preterm delivery among women with a trial of labor. We did not find an increased risk associated with preterm birth at the primary cesarean delivery, but the CI around the effect estimate was wide and does not preclude such an increased risk. Of interest was that none of the primary cesarean deliveries in which the newborn birth weight was 4,000 g or more resulted in a trial of labor with uterine rupture. This factor was not statistically significant, but the direction of the effect was consistent with another study that reported a decreased risk if the prior cesarean delivery birth weight was more than 4,000 g.5
We did not find a statistically significant effect for prior vaginal birth. Previous studies that suggest that prior vaginal birth is protective against uterine rupture included women who had one or more successful vaginal deliveries subsequent to their primary cesarean delivery, which would decrease the risk of uterine rupture.14 Our study specifically reports on whether a vaginal birth before the primary cesarean delivery is protective against uterine rupture in a subsequent trial of labor. In the sensitivity analysis where women were more broadly categorized as having spontaneous labor before the primary cesarean delivery and/or a prior vaginal birth, the reduction in risk was significant and marginally greater than that for spontaneous labor alone.
Our finding that maternal age 35 years or older at trial of labor is associated with an increased risk of uterine rupture is consistent with findings from most previous studies,5,20,21 although those studies also included women who had successful VBAC deliveries as well as women attempting a first trial of labor. A study that analyzed the risk of uterine rupture specifically for the subgroup of women with no prior vaginal delivery found no evidence of increased risk with increasing maternal age.22
One of the strengths of this study is that it is a population-based cohort in which women were tracked prospectively from primary cesarean delivery to next delivery using validated data collections. The accuracy and reliability of the demographic and delivery data in this study are high.16–18 Further, our finding that induction at trial of labor is a risk factor for uterine rupture is consistent with previous results. A potential weakness is that the outcome of uterine rupture is identified from the hospital ICD-10 codes and not from case notes. A Massachusetts study of the use of ICD-9 codes for rupture of the uterus for the period from 1990–1997 reports a positive predictive value of 50.7% for any birth with a 665.0 or 665.1 code but does not state what the positive predictive value is among the subgroup of births with a prior cesarean delivery.23 Some uterine ruptures may have been missed, and some of the reported ruptures may actually have been scar dehiscence. Because scar dehiscences in cesarean deliveries would be observed, there is the potential for more of this misclassification error to have occurred in failed VBAC, which could result in some overestimation of the risk for induced and augmented labor. Another limitation is that we did not have specific information about cervical dilation or duration of labor.
In conclusion, a primary cesarean delivery preceded by spontaneous onset of labor is associated with a decreased risk of uterine rupture for those women who attempt VBAC at their next delivery. This is in addition to a larger decrease in risk of uterine rupture if labor for the trial-of-labor birth commences spontaneously and is not augmented. The women in the study were specifically those who attempted a first trial of labor after a primary cesarean delivery, and the results should be generalizable to other women contemplating a trial of labor at term and can inform protocols for treating women who intend a first trial of labor.
1. 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.
2. Lydon-Rochelle M, Holt VL, Easterling TR, Martin DP. Risk of uterine rupture during labor among women with a prior cesarean delivery. N Engl J Med 2001;345:3–8.
3. Vaginal birth after previous cesarean delivery. ACOG Practice Bulletin No. 54. American College of Obstetricians and Gynecologists. Obstet Gynecol 2004;104:203–12.
4. Minkoff H, Powderly KR, Chervenak F, McCullough LB. Ethical dimensions of elective primary cesarean delivery. Obstet Gynecol 2004;103:387–92.
5. 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.
6. Rochelson B, Pagano M, Conetta L, Goldman B, Vohra N, Frey M, et al. Previous preterm cesarean delivery: identification of a new risk factor for uterine rupture in VBAC candidates. J Matern Fetal Neonatal Med 2005;18:339–42.
7. Sciscione AC, Landon MB, Leveno KJ, Spong CY, Macpherson C, Varner MW, et al. Previous preterm cesarean delivery and risk of subsequent uterine rupture. Obstet Gynecol 2008;111:648–53.
8. Delaney T, Young DC. Spontaneous versus induced labor after a previous cesarean delivery. Obstet Gynecol 2003;102:39–44.
9. Grobman WA, Gilbert S, Landon MB, Spong CY, Leveno KJ, Rouse DJ, et al. Outcomes of induction of labor after one prior cesarean. Obstet Gynecol 2007;109:262–9.
10. Ravasia DJ, Wood SL, Pollard JK. Uterine rupture during induced trial of labor among women with previous cesarean delivery. Am J Obstet Gynecol 2000;183:1176–9.
11. Zelop CM, Shipp TD, Repke JT, Cohen A, Caughey AB, Lieberman E. Uterine rupture during induced or augmented labor in gravid women with one prior cesarean delivery. Am J Obstet Gynecol 1999;181:882–6.
12. Gyamfi C, Juhasz G, Gyamfi P, Stone JL. Increased success of trial of labor after previous vaginal birth after cesarean. Obstet Gynecol 2004;104:715–9.
13. Hendler I, Bujold E. Effect of prior vaginal delivery or prior vaginal birth after cesarean delivery on obstetric outcomes in women undergoing trial of labor. Obstet Gynecol 2004;104:273–7.
14. Mercer BM, Gilbert S, Landon MB, Spong CY, Leveno KJ, Rouse DJ, et al. Labor outcomes with increasing number of prior vaginal births after cesarean delivery. Obstet Gynecol 2008;111:285–91.
15. Zelop CM, Shipp TD, Repke JT, Cohen A, Lieberman E. Effect of previous vaginal delivery on the risk of uterine rupture during a subsequent trial of labor. Am J Obstet Gynecol 2000;183:1184–6.
16. Roberts CL, Bell JC, Ford JB, Morris JM. Monitoring the quality of maternity care—how well are labour and delivery events reported in population health data? J Paediatr Child Health 2008;44:A106.
18. Roberts CL, Cameron CA, Bell JC, Algert CS, Morris JM. Measuring maternal morbidity in routinely collected health data: development and validation of a maternal morbidity outcome indicator. Med Care 2008;46:786–94.
19. Gonen R, Nisenblat V, Barak S, Tamir A, Ohel G. Results of a well-defined protocol for a trial of labor after prior cesarean delivery. Obstet Gynecol 2006;107:240–5.
20. Gregory KD, Korst LM, Cane P, Platt LD, Kahn K. Vaginal birth after cesarean and uterine rupture rates in California. Obstet Gynecol 1999;94:985–9.
21. Shipp TD, Zelop C, Repke JT, Cohen A, Caughey AB, Lieberman E. The association of maternal age and symptomatic uterine rupture during a trial of labor after prior cesarean delivery. Obstet Gynecol 2002;99:585–8.
22. Bujold E, Hammoud AO, Hendler I, Berman S, Blackwell SC, Duperron L, et al. Trial of labor in patients with a previous cesarean section: does maternal age influence the outcome? Am J Obstet Gynecol 2004;190:1113–8.
23. Use of hospital discharge data to monitor uterine rupture-Massachusetts, 1990-1997. MMWR Morb Mortal Wkly Rep