A systematic review and metaanalysis were performed through electronic database searches to estimate the effect of uterine closure at cesarean on the risk of adverse maternal outcome and on uterine scar evaluated by ultrasound. Randomized controlled trials, which compared single vs double layers and locking vs unlocking sutures for uterine closure of low transverse cesarean, were included. Outcomes were short-term complications (endometritis, wound infection, maternal infectious morbidity, blood transfusion, duration of surgical procedure, length of hospital stay, mean blood loss), uterine rupture or dehiscence at next pregnancy, and uterine scar evaluation by ultrasound. Twenty of 1278 citations were included in the analysis. We found that all types of closure were comparable for short-term maternal outcomes, except for single-layer closure, which had shorter operative time (–6.1 minutes; 95% confidence interval [CI], –8.7 to –3.4; P < .001) than double-layer closure. Single layer (–2.6 mm; 95% CI, –3.1 to –2.1; P < .001) and locked first layer (mean difference, –2.5 mm; 95% CI, –3.2 to –1.8; P < .001) were associated with lower residual myometrial thickness. Two studies reported no significant difference between single- vs double-layer closure for uterine dehiscence (relative risk, 1.86; 95% CI, 0.44–7.90; P = .40) or uterine rupture (no case). In conclusion, current evidence based on randomized trials does not support a specific type of uterine closure for optimal maternal outcomes and is insufficient to conclude about the risk of uterine rupture. Single-layer closure and locked first layer are possibly coupled with thinner residual myometrium thickness.
Cesarean is one of the most frequent surgical procedures around the world, constituting the delivery method in up to 30% of births. While it allows safe delivery in many situations, it remains associated with a risk of adverse outcomes. Short-term complications of cesarean include hemorrhage, wound disruption, infection, and venous thromboembolism. In subsequent pregnancies, a previous cesarean significantly increases the risk of 2 major obstetrical complications: uterine rupture and placenta accreta. Moreover, long-term adverse outcomes include pain, abnormal uterine bleeding, intraperitoneal adhesions, infertility, and additional risk of complications from future abdominal surgeries, including cesareans and hysterectomies.
A growing body of evidence suggests that the surgical technique for uterus closure influences uterine scar healing, but there is still no consensus about optimal uterus closure. Some techniques seem to have the potential to decrease the risk of short-term complications, while others have long-term benefits, such as reduced risk of uterine rupture. Recently, interest has grown in the impact of single- vs double-layer closure of hysterotomy on the risk of uterine rupture. Our recent metaanalysis of observational studies, including randomized and nonrandomized trials, suggested that locked, single-layer closure was associated with a significantly increased risk of uterine rupture in the next pregnancy compared to double-layer closure (odds ratio, 4.96; 95% confidence interval [CI], 2.58–9.52; P < .001), but unlocked single-layer closure was not (odds ratio, 0.49; 95% CI, 0.21–1.16; P = .10). However, most studies were retrospective, and the metaanalyses included only 160 cases of uterine rupture. Therefore, we believe that additional evidence-based literature is required before recommending one technique over another. With growing interest in uterine scar evaluation by ultrasound, and numerous publications showing a relationship among uterine scar defects, adverse gynecological outcomes, and uterine rupture, this review and metaanalysis could provide additional evidence regarding optimal uterine closure technique for short-term outcomes and scar defects in addition to long-term outcomes, such as uterine rupture.
Our objective was to compare the effect of single- vs double-layer and locked vs unlocked closure of low transverse cesarean on the risk of adverse maternal outcomes, including uterine rupture and uterine scar defect evaluated by ultrasound.
Source
A systematic review and metaanalyses were performed. Only randomized controlled trials (RCTs) that compared number of layers (single vs double) and the use of locking vs unlocking sutures for uterus closure at the time of low transverse cesarean were included. However, there was no other restriction about type of skin incision, abdominal opening, or closure of other layers (fascia, peritoneum, or skin). A list of key words and medical subject headings were combined to search the electronic databases PubMed, Web of Science, Embase, and Cochrane central register: “uterus,” “uterine,” “dehiscence,” “rupture,” “separation,” “scar,” “VBAC,” “vaginal birth after cesarean,” “closure,” and “cesarean” from 1980 through September 2013. References from other selected articles, reviews or metaanalysis were searched for additional relevant articles. Titles, abstracts and full texts were screened by 2 independent reviewers (S.R., E.B.) for inclusion. No language restriction was applied. Disagreement was resolved by discussion with a third reviewer (S.D.). The quality of studies was evaluated by Cochrane handbook criteria for judging risk of bias, and studies with high risk of bias were evaluated by sensitivity analysis. The quality and integrity of this review were validated with PRISMA: preferred reporting items for systematic reviews and metaanalyses.
We collected information on the following outcomes: (1) maternal infectious morbidity defined as combination of wound infection, endometritis, and postoperative febrile morbidity or equivalent; (2) endometritis; (3) wound infection; (4) blood transfusion; (5) mean blood loss; (6) duration of surgical procedure; and (7) length of hospital stay. We collected data regarding long-term adverse outcomes during subsequent pregnancies, such as uterine rupture or dehiscence. Uterine rupture is defined as complete separation of the uterine scar with visceral peritoneum disruption or bladder rupture, necessitating an emergency intervention (or equivalent definition). Uterine dehiscence is defined as partial opening of the uterus with intact visceral peritoneum (or an equivalent definition). We collected all information regarding postcesarean ultrasonographic evaluation of uterine scar, including the presence of uterine scar defect (defined as the observation of myometrial loss or deformity at the cesarean scar site), residual myometrial thickness (in mm), or residual myometrial thickness less than a specific cutoff determined by the author. In case of multiple ultrasound measures in time, available data or those who were close to 6-month evaluation were analyzed.
Risk ratios (RRs) with 95% CI compared binary outcomes (uterine rupture, infectious morbidity, endometritis, wound infection, blood transfusion, scar defect) between the 2 closure types, and mean difference with 95% CI compared continuous outcomes (duration of procedure, length of hospital stay, mean blood loss, residual myometrial thickness). Heterogeneity between studies was assessed according to I 2 criteria of Higgins et al. Pooled RR were calculated with fixed effects or with DerSimonian and Laird random effects in the absence and presence of heterogeneity (I 2 >50%), respectively. Sensitivity analysis was conducted to investigate robustness of the findings and to explain heterogeneity between studies, comparing suture type (locked or unlocked), thread type (synthetic or chromic), region (North America or other), and primary cesarean (yes or no).
Results
Our systematic search identified 1278 articles that were first screened by title and abstract, including 176 that were kept for further evaluation, 39 that were considered potentially eligible, and 20 (13,086 women) meeting all inclusion criteria ( Figure 1 ). Of them, 16 (41%) studies reported postoperative outcomes, including maternal infectious morbidity, the result encountered most frequently (in 11 trials). Six randomized trials detailed uterine scar evaluation with ultrasound, including 3 that reported the risk of uterine scar defect and 3 that observed residual myometrial thickness, but raw data could not be extracted from 1 of them and, consequently, could not be included in the analysis. Only 2 studies reported follow-up at the next pregnancy and the risk of uterine scar dehiscence or uterine rupture ( Table 1 ).
| Study | n | Inclusion criteria | Single layer | Double layer | Outcomes | Scar evaluation |
|---|---|---|---|---|---|---|
| CORONIS, 2013 | 8516 | Primary or secondary CS through planned transverse abdominal incision | Any method | Any method | Maternal infectious morbidity, endometritis, wound infection, blood transfusion, operative time, hospital stay | |
| El-Gharib and Awara, 2013 | 150 | Primary CS | Continuous locked | First layer continuous locked, second imbricating | RMT | Transabdominal US at 48 h, 2 wk, and 6 wk post-CS |
| Guyot-Cottrel, 2011 | 70 | ≥18 y, ≥37 wk, near 1 of 2 hospitals | Continuous unlocked | Continuous unlocked | Scar defect, blood transfusion | Transvaginal US at 6 wk post-CS |
| Yasmin et al, 2011 | 90 | Repeat CS in singleton pregnancy | Continuous locked | Continuous, locked (group B) or continuous unlocked (group C) | RMT, operative time, uterine dehiscence | Transabdominal US at 6 wk post-CS |
| CAESAR, 2010 | 2727 | Primary CS planned through LUS | Continuous, locked or unlocked | Continuous, locked or unlocked | Maternal infectious morbidity, endometritis, wound infection, blood transfusion, operative time, hospital stay | |
| Nabhan, 2008 | 600 | LUS CS | nr | nr | Maternal infectious morbidity, wound infection, blood transfusion | |
| Borowski et al, 2007 | 64 | Primary CS | nr | nr | Scar defect | Transvaginal US at 6 wk post-CS |
| Hamar et al, 2007 | 30 | Nonurgent primary CS | Continuous locked | First layer continuous locked, second imbricating | Operative time | US at 48 h, 2 wk, and 6 wk post-CS |
| Banad, 2006 | 100 | LUS CS | nr | nr | Maternal infectious morbidity, endometritis, wound infection | |
| Poonam et al, 2006 | 400 | CS >37 wk gestation | nr | nr | Maternal infectious morbidity, blood transfusion, operative time, hospital stay | |
| Batioglu et al, 1998 | 118 | LTCS | Continuous unlocked | Continuous locked for first layer | Maternal infectious morbidity, endometritis, operative time, hospital stay | |
| Sood, 2005 | 208 | LTCS | Continuous unlocked | Continuous unlocked | Maternal infectious morbidity, endometritis, wound infection, operative time, hospital stay | |
| Xavier et al, 2005 | 146 | LTCS | Continuous | Continuous | Maternal infectious morbidity, endometritis, wound infection, operative time, hospital stay | |
| Chitra et al, 2004 | 200 | Primary CS | Continuous locked | nr | Duration of CS | |
| Ferrari et al, 2001 | 158 | Primary CS >30 wk gestation | Continuous locked | Continuous | Duration of CS, hospital stay | |
| Bjorklund et al, 2000 | 339 | CS >37 wk gestation | nr | nr | Maternal infectious morbidity, wound infection, operative time, hospital stay | |
| Wallin and Fall, 1999 | 72 | Elective CS without prior abdominal surgery | Interrupted | Interrupted | Maternal infectious morbidity, endometritis, wound infection, blood transfusion | |
| Lal and Tsomo, 1988 | 100 | LUS CS | Interrupted | Interrupted | Scar defect | Hysterography at 3 mo post-CS |
| Hauth et al, 1992 Chapman et al, 1997 | 906 145 | LTCS | Continuous locked | First layer continuous locked, second imbricating | Maternal infectious morbidity, endometritis, blood transfusion, operative time, hospital stay, uterine dehiscence |
| Locked sutures | Unlocked sutures | |||||
|---|---|---|---|---|---|---|
| Ceci et al, 2012 | 60 | Nonurgent LUS CS in singleton pregnancy at ≥38 wk | Locked continuous excluding decidua | Interrupted suture excluding decidua | Scar defect | Transvaginal US at 6, 12, and 24 mo post-CS |
| Yasmin et al, 2011 | 90 | Repeat CS in singleton pregnancy | Continuous locked | Continuous unlocked | RMT, operative time, uterine dehiscence | Transabdominal US at 6 wk post-CS |
We undertook sensitivity analysis for maternal infectious morbidity that was examined in 11 trials. We found no significant difference in any subgroup analysis. The number of studies that evaluated uterine scar defects and uterine rupture was too small to allow sensitivity analyses.
Moderate-to-high heterogeneity was observed for most outcomes so random effect analysis was undertaken ( Table 1 ). Accurate evaluation of study heterogeneity was not possible for uterine scar defect and uterine rupture because of their small number. Two studies were considered at high risk of bias, one for the quality of randomization, and the other for unclear reporting of some outcomes ( Figure 2 ). Removing both of them from the analysis did not significantly change the results so they were conserved for analyses.
Single- vs double-layer closure
No difference was observed between single- and double-layer closure for the presence of maternal infectious morbidity, endometritis, wound infection, blood transfusion, and hospital stay ( Table 2 ). Nonetheless, single-layer closure was associated with shorter operative time (4722 patients; –6.1 minutes; 95% CI, –8.7 to –3.4; P < .001).
| Outcome | No. of trials | No. of participants | Prevalence | RR (95% CI) | P value | I 2 | |
|---|---|---|---|---|---|---|---|
| Single layer (%) | Double layer (%) | ||||||
| Maternal infectious morbidity | 10 | 5868 | 416/2937 (14.2) | 425/2931 (14.5) | 0.92 (0.74–1.15) | .48 | 37% |
| Endometritis | 8 | 13,815 | 196/6907 (2.8) | 183/6908 (2.6) | 1.04 (0.81–1.34) | .76 | 26% |
| Wound infection | 8 | 13,730 | 566/6856 (8.3) | 612/6874 (8.9) | 0.93 (0.83–1.04) | .18 | 15% |
| Blood transfusion | 7 | 14,313 | 141/7149 (2.0) | 164/7164 (2.3) | 0.86 (0.69–1.08) | .19 | 23% |
| Means | Mean difference (95% CI) | ||||||
|---|---|---|---|---|---|---|---|
| Mean blood loss, mL | 6 | 1025 | 473 | 513 | –40 (–110 to 29) | .26 | 95% |
| Duration of cesarean, min | 9 | 4722 | 33.5 | 39.7 | –6.1 (–8.7 to –3.4) | < .0001 | 94% |
| Hospital stay, d | 5 | 4063 | 4.2 | 4.6 | –0.3 (–0.7 to 0.0) | .05 | 85% |
Single-layer closure was not linked with a significant risk of uterine scar dehiscence (187 patients; RR, 2.38; 95% CI, 0.63–8.96; P = .20) or uterine rupture (no case) compared to double-layer closure ( Table 3 ). No trials reported the impact of uterus closure on pelvic adhesion or long-term adverse gynecologic outcomes.
| Outcome | No. of trials | No. of participants | Prevalence | RR (95% CI) | P value | I 2 | |
|---|---|---|---|---|---|---|---|
| Single layer (%) | Double layer (%) | ||||||
| Scar defect evaluated by US | 3 | 193 | 25/100 (25.0) | 57/93 (61.3) | 0.53 (0.24–1.17) | .12 | 67% |
| Uterine rupture or dehiscence | 2 | 187 | 4/83 (4.8) | 3/104 (2.9) | 2.38 (0.63–8.96) | .20 | 0% |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
