Vaginal birth after cesarean delivery has been advocated as a means to control the increasing rates of operative delivery in the United States. Through meticulous patient selection and careful monitoring in labor, this practice is felt to have a low risk of complications such as uterine rupture. Uterine rupture is an infrequent complication of vaginal birth after cesarean that can result in severe morbidity and mortality for both mother and fetus. It has been reported in both the scarred and unscarred uterus. The occurrence of uterine rupture has been quoted as occurring in 0.2–1.5% of all vaginal birth after cesarean (VBAC) deliveries in the United States.1
Multiple studies in the literature have documented increased risks of uterine rupture among certain subsets of patients. An increased risk of rupture has been found with the use of prostaglandins or oxytocin for induction or augmentation of labor,2,3 in patients with a short interdelivery interval,4,5 in patients in whom a single-layer closure of the previous uterine incision was performed,5,6 and in patients at older maternal ages.7 Despite the rarity of this complication, the diagnosis of uterine rupture remains a challenging one for obstetricians. Few controlled studies have attempted to address the signs and symptoms of this complication.
Fetal heart rate monitoring has been advocated as a method of differentiating uterine ruptures from successful VBAC patients. Multiple studies have attempted to define fetal heart rate characteristics that herald the occurrence of uterine rupture, but the rare nature of this event makes such definitions difficult. The most commonly cited fetal heart rate abnormality is prolonged bradycardia.7–11 Other observations noted include late decelerations,7,9 prolonged decelerations,7 variable decelerations,7,11 and a general diagnosis of “abnormalities” in the fetal heart rate.12 Most of the data on this subject come from case series, and few controlled studies on fetal heart rate changes are available. We therefore undertook this study to attempt to define fetal heart rate characteristics that may be indicative of impending uterine rupture.
MATERIALS AND METHODS
This is a case-control study of the fetal heart rate abnormalities associated with VBAC with uterine rupture versus a control group of VBAC patients without uterine rupture. The medical and billing records for all women delivering at the University of Washington Medical Center from January 1, 1984, to October 31, 2001, and at Swedish Medical Center from January 1, 1983, to October 31, 2001, were reviewed for cases of uterine rupture noted by International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes 665.0 (rupture of pregnant uterus before onset of labor) through 665.1 (rupture of pregnant uterus during labor). Both hospitals are tertiary care institutions with a level III Neonatal Intensive Care Unit. For the purposes of initial screening, uterine rupture was defined as a full-thickness uterine wall separation with extrusion of fetal or placental parts into the abdomen, associated with any clinical condition necessitating laparotomy such as acute blood loss, severe abdominal pain, or fetal distress. Inclusion criteria for the study were as follows: gestational age more than 24 weeks, a history of 1 to 3 prior low transverse uterine incisions, operative confirmation of the diagnosis of full-thickness uterine rupture, and availability of hospital records and fetal heart rate tracings. Each case of uterine rupture was then matched with 3 controls selected from a group of successful VBAC deliveries. For the purposes of the study, control patients with a successful vaginal delivery without any indication for laparotomy were assumed to have an intact uterus. All controls were matched for gestational age and delivered at the same institution as the corresponding case within 1 year. Patients undergoing VBAC were identified by using ICD-9-CM code 654.2 (previous cesarean delivery) and cross-referencing this list with a list of vaginal deliveries. Maternal demographic characteristics and information about the pregnancy history were extracted from the medical records. Institutional review board approval for this study was obtained at both institutions before the beginning of the chart review process.
Each available fetal heart rate tracing was randomly assigned a study number and blinded by removing any identifying characteristics and covering any writing on the strip with small paper tabs to prevent examiner bias. Three independent examiners (one chief resident, one second year maternal-fetal medicine fellow, and one senior attending) then analyzed each tracing over a period beginning four hours before the second stage of labor and then throughout the second stage of labor. To facilitate the correct evaluation of the fetal heart rate pattern, the current National Institute of Child Health and Human Development criteria for fetal heart rate changes were included on each data collection sheet.13 Each examiner was asked to evaluate the fetal heart rate for the presence or absence of fetal tachycardia (sustained FHR above 160 beats per minute [bpm]), mild-moderate variable decelerations (abrupt decrease in FHR for less than 2 minutes), severe variable decelerations (abrupt decrease in FHR to less than 70 bpm for less than 2 minutes), late decelerations (gradual decrease in FHR that occurs after the peak of the contraction), prolonged decelerations (decrease of FHR for a period lasting more than 2 minutes but less than 10 minutes), fetal bradycardia (baseline FHR less than 110 bpm for more than 10 minutes), and loss of uterine tone.
Before data analysis, it was determined that a fetal heart rate pattern would be considered positive if 2 of the 3 examiners rated it present on a tracing. The fetal heart tracings were then unblinded and the data from the case group analyzed and compared with the control group. Continuous variables were analyzed by using Student t test or Mann–Whitney test. Categorical variables were analyzed by using χ2 test or Fisher exact test as appropriate by using Stata 7 statistical software (Stata Corp, College Station, TX).
During the study period, a total of 45,113 deliveries occurred at the 2 institutions combined. Of these, 48 confirmed uterine ruptures were identified, 20 of which occurred at the University of Washington Medical Center and 28 at Swedish Medical Center. Thirty-six of these uterine ruptures met inclusion criteria for the study. Of the 12 cases excluded from the study, 2 were removed due to previable gestational age, 2 were removed due to the absence of prior uterine scars, 1 was removed due to 2 prior classical cesarean deliveries, and 7 were removed due to the lack of availability of the fetal heart tracings at the time of study. One hundred eight controls were then selected randomly, and 8 of these were later removed due to the lack of available fetal heart rate tracings (Figure 1). This left 36 cases and 100 controls for use during this study. For the purposes of power calculation, we assumed an incidence of 25% for each particular fetal heart rate abnormality among the controls. With this incidence, a sample of this size would provide this study with 79% power to detect an increase in incidence of each fetal heart rate abnormality to 50% in the case group (with α = .05).
There were no significant differences noted in any of the demographic characteristics (Table 1). Of particular note, there were no significant differences in the number of prior cesarean deliveries or the interdelivery interval for each group. As shown in Table 2, no differences in infant weight were noted between the 2 groups. The uterine rupture group showed significant differences toward increased use of labor induction and increased prostaglandin use. However, no differences in rates of oxytocin use were noted. In addition, the uterine rupture group showed increased rates of severe abdominal pain, significant vaginal bleeding, loss of fetal station, palpable uterine defect, and use of regional anesthesia.
Cesarean delivery rates differed between the 2 groups. Only 11% (4 of 36) of the uterine rupture group had a vaginal delivery. Of the 4 ruptured patients who delivered vaginally, 1 had an operative vaginal delivery with forceps, and 3 delivered without assistance. Among the group of ruptured patients, no maternal deaths were noted. All deliveries in both groups resulted in viable infants, with a 25% neonatal intensive care unit (NICU) admission rate noted among uterine ruptures and a 4% rate among control pregnancies. The indication for laparotomy varied widely among our study group. Of the 36 uterine ruptures, 23 laparotomies were performed for fetal distress, 6 for abdominal pain, 5 for failure to progress, and 2 for acute maternal blood loss. Two of the 36 uterine ruptures required a hysterectomy at the time of laparotomy. Both of these patients had one prior cesarean delivery and no vaginal deliveries.
Figures 2 and 3 and Tables 3 and 4 show the results of the evaluation of fetal heart rate tracings. The totals shown are the consensus results of 2 of 3 examiners. As can be seen in these tables, no significant differences were noted in the rates of fetal tachycardia, mild-moderate variable decelerations, severe variable decelerations, late decelerations, prolonged decelerations, or loss of uterine tone in either the first or second stage of labor. The only significant difference noted was the presence of fetal bradycardia in both stages. The interobserver variation was measured with kappa statistics, with the fair-to-moderate agreement among the 3 examiners noted in each category. It is of note that the evaluators had the greatest agreement about prolonged decelerations in both stages of labor (κ = 0.5 to 0.6). In the first stage, moderate agreement was seen in fetal bradycardia (κ = 0.5), late decelerations (κ = 0.4), and severe variable decelerations (κ = 0.5), with fair agreement noted in mild-moderate variables (κ = 0.3) and fetal tachycardia (κ = 0.3). In the second stage, good agreement was noted in prolonged decelerations (κ = 0.6) and mild-moderate variables (κ = 0.6), with moderate agreement in fetal tachycardia (κ = 0.4) and fair agreement in all other categories (κ = 0.3).
A nonreassuring fetal heart rate pattern is often cited in the literature as a sign of impending uterine rupture.7,9–12 One difficulty when reviewing the literature on this subject is the different definitions used by some authors. For instance, fetal bradycardia may be classified as prolonged deceleration in some articles.7 The most common pattern cited is fetal bradycardia, which is seen in 33.3% to 70.3% of all cases of symptomatic uterine rupture.7,9–11 The next most common abnormalities are late decelerations, which are reported in 19.0% to 87.5% of cases,7,10,11 and variable decelerations, which are reported in 5.1% to 19.0% of cases. Most of these studies do not elaborate on the methods used in the actual evaluation of fetal heart tracings. Reporting of these findings is biased by the retrospective nature of these studies and by the somewhat subjective nature of reading fetal heart rate tracings. In addition, none of these studies included a control group for comparison of fetal heart tracings. Many try to limit their review to cases of “symptomatic” uterine rupture, which can include abnormalities in fetal heart rate or maternal clinical conditions requiring intervention. However, by making fetal heart rate changes a part of the definition, these studies may be inflating the rates of the different abnormal heart rate patterns they are attempting to study. Our study addresses this by using only uterine rupture cases that required a laparotomy for any maternal or fetal indication.
The entry criteria for the cases and controls for this study were selected to provide a comparison between VBAC patients with a uterine rupture and VBAC patients without uterine rupture. By eliminating uterine ruptures that occurred in the absence of uterine scar or in patients who were previable, we limited our case group to patients who would be considered VBAC candidates. While we would have liked to include a group of VBAC candidates in whom a cesarean delivery was performed for fetal distress, it was felt that these would be difficult to identify by ICD-9-CM coding and possibly would not provide enough fetal heart rate tracings for adequate analysis. Of note, fetal heart rate tracings were not available for analysis in 14.6% of the cases identified versus 7.4% of the control group. Although we anticipated a few tracings to be unavailable due to the problems of record keeping and tracking over a long period of time in both groups, we also found that some of the case patients had fetal distress identified immediately upon presentation to the hospital, which resulted in immediate delivery without enough fetal heart rate tracing to evaluate.
This case-control study was undertaken to determine whether there are any significant findings on fetal monitoring that mark the onset of uterine rupture. Our results show that the only significant fetal heart rate pattern that separates uterine rupture from a successful VBAC is the presence of fetal bradycardia. There were no significant differences noted in variable, late, or prolonged decelerations, which contradicts some findings by other researchers.7,10,11 There was a significantly higher rate of cesarean delivery among the uterine rupture group. This is likely due to the pitfalls inherent in analyzing patients by retrospective review; often patients were not classified (or coded) as VBAC unless they had a successful vaginal delivery, whereas uterine ruptures were coded as such whether the delivery was vaginal or by cesarean. Despite this, our case and control groups were well matched in terms of maternal demographic characteristics and clinical characteristics. Of particular note, the cases and controls did not show a significant difference in interdelivery interval, a finding that has been reported by multiple authors previously.4,5 This finding bears further examination in a large population-based study. Another observation is the significant association of increased rates of induction of labor and prostaglandin use in uterine rupture patients, an association that has been discussed previously by Lydon-Rochelle et al.2 It is worth noting that the data sets for these studies overlap, because the previous article used Washington State birth certificate data from 1987 to 1996 for analysis, which would include 13 of our uterine rupture cases. Our analysis confirms anecdotal findings noted previously in uterine rupture patients, specifically the association with abdominal pain, vaginal bleeding, loss of fetal station, and a palpable uterine defect. Also noteworthy is the association of uterine rupture with increased rates of regional anesthesia, a finding that may be confounded by the increased rates of abdominal pain among the rupture group. Maternal and neonatal complications were relatively low among our uterine rupture group. No maternal deaths were reported, and only two hysterectomies had to be performed at the time of laparotomy.
Our primary outcome of interest was the evaluation of fetal heart rate patterns. This helps to remove some observer bias that has been seen in the literature previously. In addition, the authors recognize the inherently subjective nature of electronic fetal monitoring and attempt to control for this by using three examiners for each tracing. Each examiner also had the current National Institute of Child Health and Human Development criteria for interpretation of electronic fetal monitoring available12 on each evaluation sheet in an attempt to ensure that they conformed to the criteria as closely as possible. Despite these precautions, the kappa statistics show a fair to moderate degree of interobserver agreement among the 3 examiners. In addition, the kappa statistics varied between the first and second stages. The highest degree of agreement appears to have been present when evaluating prolonged decelerations. By taking a consensus view of each tracing and using that for analysis, the potential for bias in reading fetal heart rate patterns was decreased significantly over studies that have used just 1 examiner.
It is possible that fetal heart rate patterns can be influenced by a variety of other factors that were not controlled in this study, such as the presence or absence of nuchal cords or the location of placenta over the previous uterine scar. However, it is clear from these data that no distinct pattern emerges as a clear indicator of uterine rupture in VBAC patients except for fetal bradycardia. This reemphasizes the need for fetal heart rate monitoring at all times in those patients attempting VBAC. In addition, facilities attempting VBAC must have the capability of performing an emergency cesarean delivery without delay, as there may be little or no warning that rupture is imminent based on fetal heart rate monitoring alone.
1. American College of Obstetricians and Gynecologists. Vaginal birth after previous cesarean delivery. ACOG Practice Bulletin 5. Washington, DC: ACOG; 1999.
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
3. 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
4. Shipp TD, Zelop CM, Repke JT, Cohen A, Lieberman E. Interdelivery interval and risk of symptomatic uterine rupture. Obstet Gynecol
5. Bujold E, Mehta SH, Bujold C, Gauthier RJ. Interdelivery interval and uterine rupture. Am J Obstet Gynecol
6. Bujold E, Bujold C, Hamilton EF, Harel F, Gauthier RJ. The impact of a single-layer or double-layer closure on uterine rupture. Am J Obstet Gynecol
7. Leung AS, Leung EK, Paul RH. Uterine rupture after previous cesarean delivery: maternal and fetal consequences. Am J Obstet Gynecol
8. Leung AS, Farmer RM, Leung EK, Medearis AL, Paul RH. Risk factors associated with uterine rupture during trial of labor after cesarean delivery: a case-control study. Am J Obstet Gynecol
9. Farmer RM, Kirschbaum T, Potter D, Strong TH, Medearis AL. Uterine rupture during trial of labor after previous cesarean section. Am J Obstet Gynecol
1991;165(4 Pt 1):996–1001.
10. Ayres AW, Johnson TR, Hayashi R. Characteristics of fetal heart rate tracings prior to uterine rupture. Int J Gynaecol Obstet
11. Yap OW, Kim ES, Laros RK Jr. Maternal and neonatal outcomes after uterine rupture in labor. Am J Obstet Gynecol
12. Rageth JC, Juzi C, Grossenbacher H. Delivery after previous cesarean: a risk evaluation: Swiss Working Group of Obstetric and Gynecologic Institutions. Obstet Gynecol
© 2004 The American College of Obstetricians and Gynecologists
13. National Institute of Child Health and Human Development Research Planning Workshop. Electronic fetal heart rate monitoring: research guidelines for interpretation. Am J Obstet Gynecol