Cesarean delivery has become more common, with primary cesarean delivery rates increasing from 21.2% in 1996 to 27.1% in 2003.1 Over the same interval, vaginal birth after cesarean birth has decreased from 28.3% to 10.6%.1 This decrease may be attributed to concerns regarding the risks during trials of labor, such as uterine rupture, estimated to occur in 0.3–4.0% of pregnancies with history of cesarean delivery.2–7 The risk for poor obstetric outcome in a subsequent pregnancy has been shown to be related to surgical technique, with classical cesarean delivery having the highest risk for rupture and lower segment incisions having a lower risk.8
Currently, a low-transverse incision is the preferred method of hysterotomy during cesarean delivery. This incision has traditionally been repaired with a two-layer closure, although a one-layer closure has been shown to be as effective for surgical closure and hemostasis,9 with the advantages of shorter operative time, decreased blood loss, lower rates of endometritis, and shorter hospital stay.10 A one-layer closure usually involves a single continuous, locking layer of absorbable suture. A two-layer closure typically adds an imbricating layer of absorbable suture. Substantial research effort has attempted to estimate whether one- or two-layer closure provides better outcomes and a lower risk for uterine rupture in subsequent pregnancies.10–13 Several studies have shown an increase in uterine scar disruption for one-layer closure10,13 while others have shown no effect.11,12 Importantly these research efforts have been limited by their retrospective nature and the biases inherent in studying surgical technique with these methods.
To assess the risk of uterine rupture in a subsequent pregnancy, researchers have used ultrasonography in the evaluation of the uterine scar in the second14 and third trimesters15–18 as well as the postpartum period.19 It has generally been found that the thicker the uterine scar the lower the rate of complications.14–16,18,19 One may postulate that a thicker scar is stronger, and thus performs better, than a thinner one. Whether thickness of the uterine scar varies with the technique used at the time of hysterotomy closure is an unexplored question.
Our hypothesis is that uterine scar thickness is unaffected by the type of hysterotomy closure (one- versus two-layer). We performed a randomized trial of hysterotomy closure at primary cesarean delivery using one- versus two-layer technique, with a prospective blinded ultrasonographic follow-up of the uterine scar.
MATERIALS AND METHODS
Patients were identified as potential study subjects if they were to have a term primary cesarean delivery at term at Yale-New Haven Hospital between January 2005 and June 2006. Patients were approached for participation if their delivery was a scheduled primary cesarean delivery or a nonemergent cesarean delivery in labor. Exclusion criteria were multiple gestations, a pain score greater than 4 out of 10, abnormalities of fetal heart rate (variable or late decelerations), prior uterine surgery, hydramnios, uterine malformation, maternal diabetes, connective tissue disorders, and non-English language. Before initiation of the study, random allocation was performed using premade allocation cards (N=30), specifying “one-layer” or “two-layer,” sorted in blocks of 10 and placed into sequentially numbered, sealed, opaque envelopes. After obtaining written consent and confirming entry into the study, each patient was assigned a treatment group by selection of the next consecutive envelope. The group allocation was revealed to the surgeon during the surgery just before the hysterotomy repair.
The Yale Human Investigation Committee approved the research protocol and consent process.
Patients randomly assigned to one-layer closure had their hysterotomy closed in one layer with a running, locking suture of 0-Polysorb (coated, braided glycolide/lactide polymer; U.S. Surgical, Norwalk, CT).
Patients randomly assigned to a two-layer closure had an initial closure identical to the one-layer closure as above. An additional layer of 0-Polysorb suture was used to imbricate the first layer in a continuous, nonlocking suture.
All closures were performed by one of the investigators according to a study protocol. For each patient, additional hemostatic sutures were placed at the discretion of the operating physician, and the number of additional sutures was recorded. Also recorded prospectively were labor at time of cesarean delivery, clinical chorioamnionitis or endometritis as assessed by the treating physician, estimated blood loss at surgery, hematocrit change (preoperative minus 24 hours postoperative), duration of surgery, birth weight, sex of newborn, and whether the mother was breastfeeding. All patients received a single dose of a first-generation cephalosporin antibiotic (or equivalent if penicillin allergic) at umbilical cord clamping per hospital protocol.
Patients had an evaluation of the lower uterine segment immediately preoperative (baseline) and then follow-up measurements at 48 hours, 2 weeks, and 6 weeks postoperatively according to a study protocol. Baseline measurements were obtained by measuring the myometrial thickness in the midsagittal plane by transabdominal ultrasonography at a point below the reflection of the bladder in a technique described by Buhimschi et al20 and illustrated in Figure 1. Previously published reports have shown that the lower uterine segment thickness measured by ultrasonography is not altered by labor status.20 Postoperative evaluation of the uterine incision involved identifying the uterine scar as described by Koutsougeras et al19 and measuring the scar in the midsagittal plane perpendicular to the uterine wall by transabdominal or transvaginal approach. The scar was identified by a discontinuity in the architecture of the uterus in the midsagittal plane and was manifest by either a hyperechoic or hypoechoic line perpendicular to the wall of the uterus. A transabdominal approach was attempted, but if the scar was poorly visualized, transvaginal ultrasonography was used. Measurements were done by one of the investigators (B.D.H., S.B.S., or C.S.B.) who were blinded to the allocation group for hysterotomy closure. Previous studies by our group have validated the interobserver and intraobserver variability of these measurements taken by this method.20
Measurements were carried out using GE Logiq 3 or GE Voluson 730 ultrasound machine (GE Healthcare, Chalfont St. Giles, United Kingdom) using a 3- to 5-MHz or 4- to 7-MHz transabdominal transducer or a 5- to 9-MHz transvaginal transducer as necessary.
A sample size calculation was performed to estimate the required number of participants. Based on available data,20 to detect a 10% difference in myometrial thickness between groups or between study periods with 80% power and α=0.05, it was anticipated that 30 subjects (15 per closure group) would be required.
Data were tested for normality using the Kolmogorov-Smirnov method and reported as either mean and 95% confidence interval (CI) (for normally distributed data) or as median and interquartile range (IQR) (for skewed data). Comparisons between two groups were performed using Student t tests or Mann-Whitney rank sum tests as appropriate. Two-tailed analysis was performed in all cases. Multiple comparison procedures were performed using repeated measures analysis of variance followed by Tukey or Dunn's post hoc comparisons as appropriate. Proportions were compared with χ2 or Fisher exact test as appropriate. Statistical analyses was performed using SPSS 11.0.4 (SPSS, Inc, Chicago, IL). Sample size and power analysis calculations were performed using JMP 5.0 (SAS Institute, Inc, Chicago, IL). Throughout our analyses, we considered a P value <.05 to indicate statistical significance, and all analyses were based on an intention-to-treat.
Thirty patients were recruited and randomly assigned to one- or two-layer closure (15 per group, Fig. 2). Demographic data revealed that the women in the one-layer group were significantly older, but that there were no other significant demographic differences between groups (Table 1).
All patients had the allocated closure. In 8 (53%) of the one-layer closure cases, the surgeon deemed it necessary to use additional hemostatic sutures compared with 4 (27%) of the two-layer closure group (P=.14). There were no significant differences seen in estimated blood loss, hematocrit change, operative time, or other intraoperative data (Table 2). There were no intraoperative complications in either group.
In all, 90% of follow-up visits were kept and the uterine scar was visualized in 99% of the visits (representative image in Figure 3). Repeated measures analysis of variance, showed significant variation across time points compared with either lower segment thickness at baseline (P<.001) or with scar thickness at 48 hours postoperatively (P<.001), but that this variation did not depend on closure type (P=.79 for all visits and P=.81 beginning with 48 hour postoperative time point) (Fig. 4). Within each measurement set, the thickness of the lower segment and scar showed no differences between groups at baseline (median [95% CI], one-layer: 4.9 [3.9–6.0] mm compared with two-layer: 4.9 [4.5–5.4] mm, P=.97), at 48 hours (one-layer: 25.4 [21.2–29.7] mm compared with two-layer: 30.3 [27.5–33.0] mm, P=.09), at 2 weeks (one-layer: 17.1 [15.3–19.0] mm compared with two-layer: 16.6 [14.8–18.4] mm, P=.75), or at 6 weeks (one-layer: 13.0 [10.8–15.3] mm compared with two-layer: 14.1 [11.4–16.8] mm, P=.61). For both closure techniques, there was an initial 5- to 6-fold increase in scar thickness at 48 hours postpartum (P<.001) compared with the preoperative myometrial thickness. This initial increase was followed by a gradual reduction in thickness of the scar site. However, we found that, even at 6 weeks post partum, the uterine wall at the site of the scar was significantly thicker than before surgery, irrespective of the closure technique.
Our results demonstrate that, in the immediate postoperative period, it is possible to follow the progression of scar morphology by ultrasonography. In women who have primary cesarean delivery, there is an immediate increase in the thickness of the uterine scar compared with the preoperative myometrial thickness, and then a gradual decrease as the scar is remodeled. These changes in the scar thickness are independent of the method of closure (one- compared with two-layer) and confirm our original hypothesis.
Prior efforts with ultrasound evaluation of the uterine scar have focused on antepartum assessment, and less on postpartum evaluation of the hysterotomy incision repair stratified by closure technique. For example, investigators have elucidated the natural history of scar thickness in women with a prior uterine scar21 and found a correlation between ultrasonographic and clinically determined thickness at cesarean delivery.22 Other investigators have found that antepartum uterine scar thickness inversely correlates with risk of intrapartum rupture,16 and that antepartum assessment can predict term intrapartum uterine rupture with a high degree of accuracy.18 The technique of antepartum evaluation of the myometrium20 and postpartum evaluation of the uterine scar14,19 have been described. Our study was designed to determine if uterine closure technique has an effect on subsequent measurement of the uterine scar by ultrasonography.
Initial reports with one- compared with two-layer closures focused on intraoperative and immediate postoperative outcomes and found them to be equivalent.9 A follow-up study found equivalent outcomes in the subset of women who had a subsequent pregnancy and a trial of labor.12 Durnwald and Mercer10 performed a retrospective analysis of women with a prior primary cesarean delivery. No difference was found in rates of uterine rupture although there was an increase in “uterine windows” at cesarean delivery in women with a prior one-layer closure. Bujold et al13 performed a chart review of women undergoing a trial of labor at their institution with a history of a prior cesarean delivery. A significant increase in the odds ratio for uterine rupture was seen in women with a prior single-layer closure. These studies provide important information but nonetheless are limited by their retrospective design and unknown confounders affecting the decisions for initial closure and subsequent trials of labor,10,12,13 the interpregnancy interval on healing, and recent changes in practice patterns.13 Additionally, confounders such as operator technique and other intraoperative factors could not be taken into account. In contrast, we randomized allocation to scar closure and evaluated the effect postoperatively. Confounding variables were thus minimized by randomization.
In the context of prior studies and our current understanding of the relationship between scar thickness and wound strength, our data suggest that it may not matter which closure type is used at the time of closure of a low-transverse uterine incision after cesarean delivery. Further work is needed to determine if our findings of equivalent scar thickness persist and whether the postoperative scar thickness is predictive of future uterine disruption.
Post hoc sample size calculations were performed to determine the number of patients needed to achieve statistical significance for the differences we observed in our trial. We used the same parameters used in the sample size calculation for this study (power=80%, α=0.05). Total sample sizes (both one- and two-layer groups) varied from 1,696 patients for the 2-week measurements to 9,522 patients for the 48-hour measurements.
Although our sample size was small, significant results were obtained regarding the thickness of the scar over the study interval. No significant differences were seen between closure groups, but post hoc power analysis revealed that large sample sizes would need to be obtained to show significant differences between closure types. It is unclear how long the uterus takes to complete its remodeling, and it is possible that significant differences between closure types could appear at longer time intervals. A prior study using magnetic resonance imaging suggests that the remodeling may not be complete up to 6 months after surgery.24 In our study, the closure groups appeared to follow the same path in the change in thickness over the follow-up interval, and it is unlikely that differences would appear between closure types at a later time point. Additional studies focused on the correlation between closure technique, scar thickness, the interpregnancy interval, and future long-term obstetric performance are warranted.
In conclusion, it is possible to evaluate, in a longitudinal fashion, the uterine scar by ultrasonography in the postoperative period. Significant variation was seen over time in the scar thickness, but this was unrelated to uterine closure type. This study provides a longitudinal description of myometrial wound remodeling using ultrasonography during the first 6 weeks post partum and establishes the effect of uterine closure on myometrial scar morphology. Our observations enhance myometrial wound healing knowledge and may open novel areas of investigation for potential therapeutic intervention in humans to reduce the risk of rupture during trial of labor after prior cesarean delivery.
1. Menacker F. Trends in cesarean rates for first births and repeat cesarean rates for low-risk women: United States, 1990-2003. Natl Vital Stat Rep 2005;54:1–8.
2. Nielsen TF, Ljungblad U, Hagberg H. Rupture and dehiscence of cesarean section scar during pregnancy and delivery. Am J Obstet Gynecol 1989;160:569–73.
3. Meehan FP, Burke G, Kehoe JT, Magani IM. True rupture/scar dehiscence in delivery following prior section. Int J Gynaecol Obstet 1990;31:249–55.
4. Leung AS, Leung EK, Paul RH. Uterine rupture after previous cesarean delivery: maternal and fetal consequences. Am J Obstet Gynecol 1993;169:945–50.
5. Chazotte C, Cohen WR. Catastrophic complications of previous cesarean section. Am J Obstet Gynecol 1990;163:738–42.
6. Phelan JP, Clark SL, Diaz F, Paul RH. Vaginal birth after cesarean. Am J Obstet Gynecol 1987;157:1510–5.
7. Flamm BL, Lim OW, Jones C, Fallon D, Newman LA, Mantis JK. Vaginal birth after cesarean section: results of a multicenter study. Am J Obstet Gynecol 1988;158:1079–84.
8. Halperin ME, Moore DC, Hannah WJ. Classical versus low-segment transverse incision for preterm caesarean section: maternal complications and outcome of subsequent pregnancies. Br J Obstet Gynaecol 1988;95:990–6.
9. Hauth JC, Owen J, Davis RO. Transverse uterine incision closure: one versus two layers. Am J Obstet Gynecol 1992;167 (4 Pt 1):1108–11.
10. Durnwald C, Mercer B. Uterine rupture, perioperative and perinatal morbidity after single-layer and double-layer closure at cesarean delivery. Am J Obstet Gynecol 2003;189:925–9.
11. Tucker JM, Hauth JC, Hodgkins P, Owen J, Winkler CL. Trial of labor after a one- or two-layer closure of a low transverse uterine incision. Am J Obstet Gynecol 1993;168:545–6.
12. Chapman SJ, Owen J, Hauth JC. One- versus two-layer closure of a low transverse cesarean: the next pregnancy. Obstet Gynecol 1997;89:16–8.
13. 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 2002;186:1326–30.
14. Sambaziotis H, Conway C, Figueroa R, Elimian A, Garry D. Second-trimester sonographic comparison of the lower uterine segment in pregnant women with and without a previous cesarean delivery. J Ultrasound Med 23:907–11, 2004; quiz 913–4.
15. Michaels WH, Thompson HO, Boutt A, Schreiber FR, Michaels SL, Karo J. Ultrasound diagnosis of defects in the scarred lower uterine segment during pregnancy. Obstet Gynecol 1988;71:112–20.
16. Rozenberg P, Goffinet F, Phillippe HJ, Nisand I. Ultrasonographic measurement of lower uterine segment to assess risk of defects of scarred uterus. Lancet 1996;347:281–4.
17. Rozenberg P, Goffinet F, Philippe HJ, Nisand I. Thickness of the lower uterine segment: its influence in the management of patients with previous cesarean sections. Eur J Obstet Gynecol Reprod Biol 1999;87:39–45.
18. Asakura H, Nakai A, Ishikawa G, Suzuki S, Araki T. Prediction of uterine dehiscence by measuring lower uterine segment thickness prior to the onset of labor: evaluation by transvaginal ultrasonography. J Nippon Med Sch 2000;67:352–6.
19. Koutsougeras G, Karamanidis D, Chimonis G, Gottas N, Polydorou A, Elmazis C, et al. Evaluation during early puerperium of the low transverse incision after cesarean section through vaginal ultrasonography. Clin Exp Obstet Gynecol 2003;30:245–7.
20. Buhimschi CS, Buhimschi IA, Malinow AM, Weiner CP. Myometrial thickness during human labor and immediately post partum. Am J Obstet Gynecol 2003;188:553–9.
21. Gotoh H, Masuzaki H, Yoshida A, Yoshimura S, Miyamura T, Ishimaru T. Predicting incomplete uterine rupture with vaginal sonography during the late second trimester in women with prior cesarean. Obstet Gynecol 2000;95:596–600.
22. Tanik A, Ustun C, Cil E, Arslan A. Sonographic evaluation of the wall thickness of the lower uterine segment in patients with previous cesarean section. J Clin Ultrasound 1996;24:355–7.
23. Buhimschi CS, Buhimschi IA, Patel S, Malinow AM, Weiner CP. Rupture of the uterine scar during term labour: contractility or biochemistry? BJOG 2005;112:38–42.
© 2007 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
24. Dicle O, Kucukler C, Pirnar T, Erata Y, Posaci C. Magnetic resonance imaging evaluation of incision healing after cesarean sections. Eur Radiol 1997;7:31–4.