Share this article on:

Predicting Failed Trial of Labor After Primary Cesarean Delivery

Dinsmoor, Mara J. MD, MPH; Brock, Ellen L. MD, MPH

doi: 10.1097/01.AOG.0000110544.42128.7a
Original Research

OBJECTIVE: To apply published scoring systems retrospectively to patients who had undergone a trial of labor after cesarean delivery to estimate whether there was a score at which a trial of labor should be discouraged.

METHODS: Patients with 1 previous cesarean delivery who then delivered between January 1, 1998, and December 31, 1998, were studied. An investigator blinded to outcome assigned scores using 3 different scoring systems. Student t test, χ2, analysis of variance, and receiver operating curve analysis were performed. P < .05 was significant.

RESULTS: Seventy-six percent (117/153) of trial of labor patients had a vaginal birth after cesarean delivery. Successful vaginal birth after cesarean delivery patients had significantly different mean scores using all 3 scoring systems, but none of the systems accurately predicted failed trial of labor resulting in cesarean delivery. Unfavorable scores were associated with high rates of major complications.

CONCLUSIONS: An unfavorable score predicting a high rate of complications and more failed trials of labor may help in counseling patients considering trial of labor. A better system to predict the success or failure of trial of labor is needed.


Failed trial of labor is not accurately predicted by scoring systems, although the risk of complications is high in those with unfavorable scores.

From the Department of Obstetrics and Gynecology. Medical College of Virginia Physicians and Hospitals of the Virginia Commonwealth University Health System. Richmond, Virginia.

Received July 8, 2003. Received in revised form November 3, 2003. Accepted November 11, 2003.

Address reprint requests to: Mara J. Dinsmoor, MD, MPH, Department of Obstetrics and Gynecology, Evanston Northwestern Healthcare, 2650 Ridge Avenue; Room 1600WH, Evanston, IL 60201; e-mail:

Beginning in the mid-1980s, when trial of labor and vaginal delivery after a low transverse cesarean delivery were shown to be reasonably safe, pregnant women (and their obstetricians) have been encouraged in most cases to attempt a vaginal birth after cesarean delivery (VBAC). Success rates vary but average approximately 60–80%.1 A number of studies have found that circumstances surrounding the first cesarean delivery strongly influence the likelihood of successful VBAC.1–5 For example, patients whose previous cesarean delivery was for a “nonrecurring” indication, such as malpresentation or placenta previa, are much more likely to have a successful VBAC than those patients whose first cesarean delivery was for labor dystocia.3,4 Furthermore, concerns have been raised that perhaps the impetus to encourage a trial of labor has become too aggressive, resulting in some patients undergoing a trial of labor that was ill-advised. There have been several reports of increased complications, prolonged hospital stays, and increased costs after failed trial of labor, although overall, trial of labor appears to be less costly than elective repeat cesarean delivery.2,6,7 This has led some investigators to attempt to identify those patients who are at high risk for failed trial of labor.

Before beginning the present study in 1999, we identified 3 previously published scoring systems. Troyer and Parisi8 developed a scoring system (range 0–4) based on their analysis of 264 patients undergoing a trial of labor. They found 4 variables that significantly reduced the success rate of VBAC, including previous dysfunctional labor, absence of a previous vaginal delivery, nonreassuring fetal status at admission, and need for labor induction. Ninety-two percent of patients having none of these variables delivered vaginally, whereas only 46% of patients with all 4 variables delivered vaginally. Flamm and Geiger9 analyzed 5,003 patients undergoing a trial of labor, randomly selecting half of the patients to develop the scoring system (range 0–10), and then retrospectively applying the scoring system to the second half. Variables in their scoring system included maternal age, previous vaginal delivery, indication for previous cesarean delivery, and cervical dilation and effacement at admission. They found that 95% of patients with a score of 8–10 successfully delivered vaginally compared with 49% of those who scored 0–2. Finally, Alamia and colleagues (Alamia V, Meyer B, Selioutski O, Vohra N. Can a VBAC scoring system predict uterine rupture in patients attempting a trial of labor? [abstract] Obstet Gynecol 1999;93:64S) published results of using a VBAC scoring system in an attempt to predict uterine rupture during a trial of labor. The scoring system (range 0–10) was devised using 589 patients undergoing a trial of labor over a 2-year time period. Variables included in their scoring system included previous vaginal delivery, indication for previous cesarean delivery, fetal station, cervical dilation and effacement at admission, and the presence of spontaneous labor. In their report, they then evaluated the 9 cases of uterine rupture that occurred during that same period. They found that patients with a uterine rupture were 9 times more likely to have a score of 4 or less. To our knowledge, these scoring systems have not been applied to other patient populations outside the study institutions in either a prospective or retrospective fashion.

Our aim was to apply the different scoring systems to patients in our institution who had undergone a trial of labor after primary cesarean delivery. We sought primarily to validate the scoring system in our population, and to estimate by receiver operating curve analysis if there was a score, for example at which 75% of patients needed a repeat cesarean delivery, at which a trial of labor could be discouraged.

Back to Top | Article Outline


For the study period of January 1, 1998, to December 31, 1998, two different databases were queried in an effort to include all potential study candidates. The database maintained by the hospital medical record department was searched for all obstetrical deliveries coded as a previous cesarean delivery (International Classification of Diseases-9th Revision, Clinical Modification 654.21) In addition, the perinatal database, a computerized record of all deliveries occurring in our hospital since 1992, was searched for all patients delivering in 1998 recorded as “VBAC candidates.”

Maternal hospital charts were reviewed for all potential candidates, at which time data were collected on maternal demographics and characteristics, the indication for the previous cesarean delivery, and the events of labor and delivery. Major complications, as noted in the chart, were recorded, and included the development of intra-amniotic infection or endometritis, uterine dehiscence or rupture, or hemorrhage requiring transfusion.

After chart review and data abstraction, an investigator (ELB) blinded to the ultimate outcome of the pregnancy reviewed each patient's data. Each patient was then scored using the point systems described by Troyer and Parisi,8 Flamm and Geiger,9 and Alamia and colleagues (Alamia V, et al. Obstet Gynecol 1999;93:64S). This study was reviewed and approved by the Institutional Review Board of the Medical College of Virginia of Virginia Commonwealth University. Subject confidentiality was maintained at all times.

Statistical analysis was performed using SPSS 10 (Chicago, IL) for the personal computer. Data were analyzed using χ2 analysis for discrete variables and Student t test and analysis of variance for continuous variables. Receiver operating curve analysis was performed using SAS (Cary, NC). A P value of < .05 was considered significant.

Back to Top | Article Outline


Using the combined searches described above, a total of 290 patients who delivered between January 1, 1998, and December 31, 1998, were identified as having had a previous cesarean delivery. After chart review, 36 patients were not considered VBAC candidates, 16 as a result of a previous classical cesarean delivery, 17 as a result of malpresentation in the current pregnancy, and 1 each as the result of an anomalous fetus, placenta previa, and human immunodeficiency virus infection. Twenty-nine patients were not included because of a history of 2 or more previous cesarean deliveries, and 31 were eliminated because they had chosen to have an elective repeat cesarean delivery. The hospital charts were missing or incomplete for 41 (12%) patients. The remaining 153 patients, all carrying singleton fetuses, were included in the study. In 80 cases (53%), the indication for the previous cesarean delivery could be confirmed by reviewing the operative note and/or records for the previous delivery. In 91% (n = 73) of the patients, the reported reason for the previous cesarean delivery was confirmed. The most common error in recall was citing fetal distress as an indication for cesarean delivery when arrest of labor was the actual indication, a situation found in 8% of the operative notes reviewed.

In general, the demographics of the study group were similar to those of the general obstetric population at the Medical College of Virginia, with a large proportion being indigent, a majority (68%) being African American and most (65%) receiving prenatal care in hospital and health department clinics. The most frequent indication for their primary cesarean delivery was active phase arrest (39%), with fetal distress (24%) and malpresentation (20%) also being common indications.

Successful VBAC occurred in 76% of the patients in our study. There were no significant differences between the 2 groups (successful VBAC and failed VBAC) in maternal age, race, site of prenatal care, or insurance status (Table 1). Compared with those women who ultimately had a repeat cesarean delivery (failed VBAC), we found that those who delivered vaginally weighed less and were more likely to have had a previous vaginal delivery, although neonatal birth weights were not significantly different. Women who delivered vaginally were more likely to have entered into spontaneous labor and had more advanced cervical dilation and effacement and a lower station at the time of admission to labor and delivery. Compared with women who failed their trial of labor, successful VBAC patients also had significantly different mean scores using all 3 scoring systems.

Table 1

Table 1

Comparing patients with successful VBAC to those with a failed VBAC attempt, patients who delivered vaginally were less likely to develop intra-amniotic infection or endometritis and were less likely to have a uterine dehiscence or rupture (Table 2). Major complications were twice as common in those patients who had a failed VBAC.

Table 2

Table 2

Using receiver operating curve analysis, we found that none of the 3 scoring systems was adequately sensitive in predicting a failed trial of labor for clinical use (Figure 1). For example, although there appears to be a “break point” in the Flamm scoring system at a score of 5, 67% of patients with a score of 5 or less still delivered vaginally. Examined individually, all 3 scoring systems were accurate in predicting those women who would ultimately deliver vaginally. Conversely, they were not accurate in predicting those patients who ultimately required repeat cesarean delivery (Table 3).

Table 3

Table 3

Figure 1

Figure 1

Although not predictive of ultimate route of delivery, unfavorable scores were associated with a significantly higher rate of complications, even among those who delivered vaginally (Table 4). Patients with an unfavorable Troyer/Parisi score of 3 (n = 18) had complicated deliveries in 50% of cases, including 40% of successful VBAC and 63% of failed trial of labor resulting in cesarean delivery. Thirty-seven percent of patients with an unfavorable Flamm score of 5 or less (n = 97), including 32% of patients who delivered vaginally and 47% of those who delivered by repeat cesarean delivery, had major complications. Those patients who had an unfavorable Alamia score of 4 or less (n = 89) had complications in 35% overall, including 31% of patients who delivered vaginally and 43% of those who delivered by repeat cesarean delivery. In 2 of the 3 scoring systems, the rate of major complications was significantly higher in patients with “unfavorable” scores than in those with “favorable” scores.

Table 4

Table 4

Back to Top | Article Outline


Our study confirms the findings of previous authors, with similar VBAC success rates in our population at the different scoring levels. Of note, however, is the fact that even those patients with the most unfavorable scores still had an approximately 50% success rate for vaginal delivery after a trial of labor. It has been proposed that a clinical test to determine eligibility for a trial of labor should have both a sensitivity and specificity of at least 85–90% to be clinically useful.10 Although not of much use in identifying those patients who are at high risk for a failed trial of labor, some patients with unfavorable scores may elect to undergo an elective repeat cesarean delivery if advised of the lower success rate (50% or less) and the high rate of major complications. Alternatively, the high success rate for vaginal delivery and the relatively low rate of complications may reassure patients with favorable scores.

Our data also confirm the findings of a meta-analysis performed by Rosen and colleagues,2 who reported an increase in febrile morbidity, uterine dehiscence, and low 5-minute Apgar score in patients with a failed trial of labor compared with those with an elective repeat cesarean delivery. Although we did not record neonatal outcome in this study, we also found an increase in febrile morbidity and uterine rupture in the failed trial of labor group. Despite the increase in complications and hospital stay after a failed trial, Traynor and Peaceman6 have reported that hospital charges are higher after elective repeat cesarean delivery compared with trial of labor. Similarly, Grobman and colleagues11 have reported that a trial of labor appears to be more cost-effective than elective cesarean delivery. To optimize patient safety and optimize maternal and fetal outcomes, better systems to predict the success or failure of a vaginal trial of labor are still needed.

Limitations of this study include the fact that the scoring systems are being applied in a retrospective fashion, using data abstracted from charts. One of the scoring systems used was in fact designed to predict uterine rupture rather than success or failure of a trial of labor (Alamia V, et al. Obstet Gynecol 1999;93:64S). In some cases, the indication for the previous cesarean delivery could not be verified, and in a small number of cases, review of the contemporary labor record and/or operative note revealed a different indication than what was cited in the subsequent pregnancy. There is some inherent subjectivity in assigning scores that might be reduced if a single examiner in a prospective fashion did all exams and evaluations.

An ideal scoring system would allow patients to be evaluated and counseled at 37–39 weeks, before the onset of labor. Those women who are very likely to need a repeat cesarean delivery would be scheduled for such, whereas the practitioner and patient would await spontaneous labor in those patients who were more likely to deliver vaginally.

Back to Top | Article Outline


1.American College of Obstetricians and Gynecologists. Evaluation of cesarean delivery. Washington (DC): American College of Obstetricians and Gynecologists; 2000.
2.Rosen MG, Dickinson JC, Westhoff CL. Vaginal birth after cesarean: a meta-analysis of morbidity and mortality. Obstet Gynecol 1991;77:465–70.
3.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.
4.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.
5.Demianczuk NN, Hunter DJ, Taylor DW. Trial of labor after previous cesarean section: Prognostic indicators of outcome. Am J Obstet Gynecol 1982;142:640–2.
6.Traynor JD, Peaceman AM. Maternal hospital charges associated with trial of labor versus elective repeat cesarean section. Birth 1998;25:81–4.
7.Stone J, Lockwood CJ, Berkowitz GS, Lynch L, Alvarez M, Lapinski RH, et al. Morbidity of failed labor in patients with prior cesarean section. Am J Obstet Gynecol 1992;167:1513–7.
8.Troyer LR, Parisi VM. Obstetric parameters affecting success in a trial of labor: Designation of a scoring system. Am J Obstet Gynecol 1992;167:1099–104.
9.Flamm B, Geiger AM. Vaginal birth after cesarean delivery: An admission scoring system. Obstet Gynecol 1997;90:907–10.
10.Macones GA. The utility of clinical tests of eligibility for a trial of labour following a caesarean section: a decision analysis. Br J Obstet Gynaecol 1999;106:642–6.
11.Grobman WA, Peaceman AM, Socol ML. Cost-effectiveness of elective cesarean delivery after one prior low transverse cesarean. Obstet Gynecol 2000;95:745–51.
© 2004 The American College of Obstetricians and Gynecologists