Full-dilatation cesarean section and risk of cesarean scar diverticulum: a multi-institutional cohort study

Between January 2018 and January 2023, a total of 156,356 women had labored at the three participating hospitals. Among them, 145,224 cases were not eligible for this study based on the inclusion and exclusion criteria. Eventually, a total of 11,132 women were included in this study. Among them, 5091 women (45.73%) underwent cesarean section before labor (group A), 3603 women (32.37%) had cesarean delivery during the first stage of labor (group B), and the remaining 2438 women (21.90%) underwent second-stage cesarean delivery (group C). Figure  1 presents the process of case selection. Fig. 1 The process of cases inclusion The process of cases inclusion As for the baseline data characteristics of the study cohort, no statistical difference was identified in most baseline data among the groups, including the maternal age at birth ( P  = 0.213), the maternal BMI at birth ( P  = 0.750), the smoking status at pregnancy booking ( P  = 0.363), medical comorbidities (for chronic hypertension: P  = 0.703. for pregnancy-related hypertension: P  = 0.413. for pre-pregnancy diabetes: P  = 0.604. for gestational diabetes mellitus: P  = 0.301), the proportion of no prenatal care ( P  = 0.615), the birth weight of the baby ( P  = 0.315), and the gestational age of birth ( P  = 0.756). In terms of the proportion of induction of labor, the numbers for Groups A, B, and C were 7.5%, 17.0%, and 21.3%, respectively, and the differences among the three groups were statistically significant ( P  = 0.013). Characteristics of the study cohort, stratified by the timing of cesarean section, are presented in Table  1 . Table 1 Characteristics of the study cohort Cesarean section before labor (group A = 5091) the first-stage cesarean section (group B = 3603) the second-stage cesarean section (group C = 2438) P The maternal age at birth 0.213 Maternal age (years) 26.5 ± 4.3 27.2 ± 4.4 26.8 ± 4.6 Maternal age categories 0.097 18–35 years 4327 (85.0%) 2882 (80.0%) 1877 (77.0%) > 35–45 years 764 (15.0%) 721 (20.0%) 561 (23.0%) Maternal BMI at birth 22.7 ± 4.8 22.3 ± 4.4 24.0 ± 5.1 0.750 Maternal BMI categories 0.191 Underweight 153 (3.0%) 180 (5.0%) 122 (5.0%) Normal weight 3564 (70.0%) 2342 (65.0%) 1463 (60.0%) Overweight and obese 1374 (27.0%) 1081 (30%) 853 (35.0%) Smoking status at pregnancy booking 0.363 Current smoker 301 (5.9%) 217 (6.0%) 178 (7.3%) Non-smoker 4790 (94.1%) 3386 (94.0%) 2260 (92.7%) Medical comorbidities Chronic hypertension 224 (4.4%) 227 (6.3%) 139 (5.7%) 0.703 Pregnancy-related hypertension 687 (13.5%) 544 (15.1%) 285 (11.7%) 0.413 Pre-pregnancy diabetes 305 (6.0%) 173 (4.8%) 139 (5.7%) 0.604 Gestational diabetes mellitus 632 (12.4%) 490 (13.6%) 288 (11.8%) 0.301 No prenatal care 484 (9.5%) 299 (8.3%) 244 (10.0%) 0.615 Induction of labor 382 (7.5%) 613 (17.0%) 519 (21.3%) 0.013 Birth weight (g) 3051.5 ± 464 3080.8 ± 572 3245.6 ± 501 0.315 Gestational age at delivery (weeks) 38.3 ± 1.4 38.8 ± 1.6 38.5 ± 1.5 0.756 BMI body mass index Values are presented as mean ± standard deviation, median (minimum–maximum), or as number (percentage) Characteristics of the study cohort BMI body mass index Values are presented as mean ± standard deviation, median (minimum–maximum), or as number (percentage) As for the reasons for cesarean delivery, in Group A and Group B, more than half of the cesarean sections were due to maternal factors, with proportions of 68.0% and 55.0%, respectively. while in Group C, this proportion was 29.5%. Table  2 shows the proportions of indications for cesarean section. Table 2 Indications for cesarean delivery Group A ( N  = 5091) Group B ( N  = 3603) Group C ( N  = 2438) P Indications 0.025 Maternal factors 3462 (68.0%) 1982 (55.0%) 719 (29.5%) Fetal factors 1629 (32.0%) 1621 (45.0%) 1719 (70.5%) Values are presented as a number (percentage) Indications for cesarean delivery Values are presented as a number (percentage)

Three Chinese centers were getting involved (Longquanyi District of Chengdu Maternal and Child Health Hospital, The Second Affiliated Hospital of Chengdu Medical College, and West China Second University Hospital). Our study is a multi-institutional retrospective cohort study. Due to the retrospective design of this study and the fact that no identifiable personal information was included in this study, the ethics review committee of the participating centers has approved it and exempted this study from ethical review. By searching the electronic medical records system, we identified pregnant women who gave birth at the participating hospitals between January 2018 and January 2023. Women were eligible for this study if they met the following criteria: (1) had a single pregnancy, (2) were between 18 and 45 years old when they gave birth, (3) the gestational age at delivery was between 37 and 40 + 6 weeks, (4) the fetal presentation was cephalic, (5) gave birth by cesarean section with transverse incision of the lower uterine segment, (6) had live birth, and (7) underwent pelvic ultrasound examination at the participating hospitals within two years after delivery. Cases were excluded from our study if they (1) had multiple pregnancies, (2) had premature deliveries, (3) had placenta previa, (4) had a history of cesarean section or myomectomy, (5) had a history of uterine adenomyosis, or (6) had a history of uterine perforation or rupture. Meanwhile, those cases that contain incomplete data of interest would not be included in this study either. Based on the timing of the cesarean section, the eligible cases were divided into three groups: the cesarean section before labor group (group A), the first-stage cesarean section group (group B), and the second-stage cesarean section group (group C). In our study, regular uterine contractions accompanied by dilation of the cervix were regarded as the beginning of labor; full dilation of the cervix was regarded as a sign of the second stage of labor. We collected the data of interest by searching the electronic medical records of the eligible cases. The data of interest were as follows: the maternal age at birth, the maternal body mass index (BMI) at birth, smoke history, the specific medical conditions (including chronic hypertension, pregnancy-related hypertension, pre-pregnancy diabetes, and gestational diabetes mellitus), whether the eligible cases underwent prenatal care as recommended during pregnancy, whether the included cases underwent induction of labor, the neonatal weight, the gestational age at delivery, and the indications for cesarean delivery. In our study, maternal age was categorized as 18–35 years and > 35–45 years. Maternal BMI at pregnancy booking was categorized according to the World Health Organization definitions as underweight (< 18.5 kg/m 2 ), normal weight (18.5–24.9 kg/m 2 ), or overweight/obese (≥ 25 kg/m 2 ). The maternal smoking status at pregnancy booking was categorized as a current smoker or a non-smoker. In our study, the medical comorbidities of interest were defined according to the corresponding clinical practice guidelines. For the convenience of comparison among groups, in our study, the indications for cesarean section were classified as maternal factors and fetal factors. The former included pelvic stenosis, placental abruption, gestational hypertension, heart disease during pregnancy, and acute fatty liver during pregnancy, etc. The latter includes abnormal fetal position, umbilical cord prolapse, nuchal cord, fetal distress, and macrosomia. In our study, the primary outcome of interest was the rate of CSD after cesarean section. The time cut-off point for follow-up was January 1, 2025. The secondary outcome of interest in our study was the rate of symptomatic CSD after cesarean section. In our study, the method for diagnosing CSD was transvaginal ultrasound examination, which was performed by two experienced ultrasound physicians from the participating institutions. In this study, we defined CSD as an anechoic defect within the myometrium of the lower uterine segment at least 1.0 mm in depth [ 21 ]. The symptomatic CSD refers to abnormal uterine bleeding caused by CSD. In our study, the included cases underwent transvaginal ultrasound examination due to abnormal uterine bleeding, routine check-ups, or other gynecological reasons. In our study, abnormal uterine bleeding was defined according to the International Federation of Gynecology and Obstetrics systems [ 22 ]. We carried out tests for skewness/kurtosis to examine the normality of quantitative variables. For non-normally distributed data, the Kruskal-Wallis test was carried out to make comparisons among groups. For normally distributed data, the analysis of variance was employed to make comparisons among groups. For categorical variables, we employed Fisher’s exact test and the Pearson chi-square test to analyze the differences among groups. To estimate the risk of CSD after full-dilation cesarean delivery compared to women who had a cesarean delivery before labor or a first-stage cesarean delivery, we carried out a multiple logistic regression. To select the potential confounders for the multivariable modeling, we conducted the bivariate analyses. Variables that had a P value of less than 0.2 in the bivariate analyses were considered eligible for multiple logistic regression analyses [ 23 ]. A backward stepwise elimination strategy was used to reach a parsimonious model. Any variable that affected the effect size between exposure status and outcome by at least 15% was retained in the model as a confounder [ 24 , 25 ]. The following variables were included in the model: the maternal age at birth, maternal BMI at birth, smoking status at pregnancy booking, medical comorbidities, prenatal care, induction of labor, and neonatal weight. Using logistic regression, the odds ratio (OR), the adjusted odds ratio (aOR), and their 95% confidence interval (95% CI) were estimated. In this study, we employed STATA 17.0 for Windows (STATA Corporation, College Station, TX) to conduct all data analyses. All hypothesis tests were two-sided, and P values less than 0.05 were considered statistically significant.

In summary, our study finds that the risk of CSD is highly associated with the timing of cesarean section. Compared with the cesarean section that occurs before labor and during the first-stage of labor, the second-stage cesarean section places women at a higher risk of developing CSD. Given that this study is a retrospective cohort study, our findings still need to be supported by more prospectively designed studies. Also, additional studies are required to investigate the underlying mechanism.

With three high-volume centers involved and 11,132 women included, our study found that the risk of CSD after cesarean section was highly associated with the timing of cesarean section. Compared with cesarean section before labor and during the first stage of labor, the second-stage cesarean section resulted in more CSD. As the rate of cesarean section increases, the incidence of CSD also rises. Poor healing of the uterine incision after cesarean section causes the formation of CSD, presenting as thinning of the muscle layer at the incision of the uterus and manifesting as abnormal uterine bleeding [ 26 , 27 ]. CSD is a long-term complication of cesarean section, severely affecting the quality of life of women and increasing the risk of adverse outcomes in women with re-pregnancy [ 28 ]. With the increasing prevalence of cesarean section and the progress of diagnostic methods, the incidence of CSD is getting higher. Therefore, it is important to understand its formation and prevention. The aetiology of CSD is not yet clear, and it is mainly considered to be associated with the location of the incision in the uterus, the techniques of suture, the number of cesarean sections, and infection [ 15 ]. Especially, inappropriate suturing methods are highly correlated with the formation of CSD. At present, it is believed that single-layer sutures are more prone to incision defects, while overly dense sutures will reduce the blood supply to the incision, thereby leading to poor incision healing [ 15 , 29 ]. Currently, there are few studies on this topic. A systematic review conducted by Bij de Vaate AJ and others revealed that the risk factors for CSD are as follows: single-layer myometrium closure, multiple cesarean sections, and uterine retroflexion [ 30 ]. Also, some researchers even created a prognostic model on niche development after a first caesarean section; they found that many factors could be associated with the development of CSD, including gestational age, twin pregnancy, smoking, closure, and surgical experience [ 31 ]. We found that the formation of CSD is highly correlated with the timing of cesarean section. Compared with the cesarean section before labor and during the first stage of labor, the cesarean section during the second stage of labor greatly increases the possibility of CSD. Its underlying mechanism remains unclear. Someone believes that the cesarean section in the second stage of labor may cause trauma to the cervix, which may be related to the formation of CSD [ 32 ]. Due to the anatomical changes in the lower segment of the uterus during the second stage of labor, the actual location of the transverse incision may be quite different from that of cesarean section before the onset of labor and during the first stage of labor [ 33 – 35 ]. In addition, the extension of the incision during surgery and the inadvertent tacking up of the cervical tissue into the closure may be correlated with the formation of CSD [ 33 – 35 ]. Its underlying mechanism still requires more research to be explored. Also, to prevent CSD, we need to reduce the incidence of cesarean section during the second stage of labor. Attention should be particularly paid to the management of the second stage of labor [ 35 , 36 ]. As far as we know, there are currently few relevant studies on this topic. Our study had three high-volume centers involved, and more than 10,000 subjects were included. However, this study still suffers from some limitations. First, our study suffers from some inherent bias because of the retrospective nature of the study design, including information bias, selection bias, etc. During the implementation stage of the study, in order to minimize the impact of these biases as much as possible, we strictly followed the implementation guidelines of cohort studies [ 37 ]. Second, due to the absence of relevant content in the original medical records, data on cesarean section suture techniques were not extracted in this study, which, to some extent, undermined the consistency of our study findings. Third, our study only included women from southwest China; this may not present the general population very well. Fourth, this study is a retrospective study, whether a case has CSD was mainly determined by the patient’s ultrasound examination report. In clinical practice, different ultrasound doctors have certain differences in their diagnoses of CSD. Lastly, due to the limited resources, some important datawere not collected, such as the position of the uterus, the thickness of the myometrium at the diverticulum, and the location of the diverticulum, etc. Despite these limitations, our study provides important knowledge for this clinical topic.

Worldwide, cesarean section rates are high and rising [ 1 – 5 ]. According to published reports, approximately one in five births occurs through cesarean section [ 1 , 2 ]. There are significant regional variations in population cesarean section rates, which range from nearly 4% in West and Central Africa to nearly 45% in Latin America and the Mediterranean region [ 2 , 6 , 7 ]. Cesarean section is a double-edged sword. On the one hand, rational use of cesarean section can reduce the rate of maternal and fetal morbidity and mortality [ 1 , 6 ]. On the other hand, overuse of cesarean section is not only unbeneficial but even harmful [ 3 , 4 , 8 ]. Rational use of cesarean section is a major health issue worldwide. As a surgical delivery method, a cesarean section may cause short-term and long-term complications for women, including but not limited to the following: infection, abnormal placentation, uterine rupture, chronic pelvic pain, pelvic adhesions, fertility issues, and irregular bleeding [ 9 – 14 ]. Among them, irregular bleeding is common and often thought to be closely related to a condition called cesarean scar diverticulum (CSD). CSD is caused by the incision healing defect in the lower section of the uterus after cesarean section [ 15 , 16 ]. The incidence of CSD among women with irregular vaginal bleeding after cesarean section varies according to the approaches of detection, about 60% when assessed by transvaginal ultrasonography and about 90% when assessed by hysteroscopy [ 17 ]. Also, the incidence of CSD is associated with the number of cesarean section [ 18 ]. CSD mainly presents as abnormal uterine bleeding, which seriously deteriorates women’s quality of life [ 19 ]. Among the published studies, it is widely believed that CSD is mainly caused by the deficiency of uterine closure techniques [ 16 , 20 ]. However, few studies have looked at the impact of cesarean section timing on the incidence of CSD. Considering the physiological and physical changes in the lower segment of the uterus during labor, we believe that the timing of cesarean section may be related to the risk of CSD. Our study, conducted at three high-volume centers in China, aimed to answer whether second-stage cesarean section significantly increases a woman’s risk of CSD compared to cesarean section before labor and first-stage cesarean section.

For the entire study cohort, a total of 1134 (10.2%) women had CSD after cesarean section. The numbers of CSD for Groups A, B, and C were 397 (7.8%), 313 (8.7%), and 424 (17.4%), respectively. The differences among groups were statistically significant ( P  < 0.001). In Group A, a total of 285 (5.6%) women suffered from symptomatic CSD, accounting for 71.8% of women with CSD. The number of women who suffered from symptomatic CSD in Groups B and C is 227 (6.3%) and 295 (12.1%), respectively. The differences among groups were statistically significant ( P  < 0.001). Table  3 shows the incidence of CSD and symptomatic CSD in Groups A, B, and C. Table 3 Rate of uterine scar diverticulum according to the timing of cesarean section Group A ( N  = 5091) Group B ( N  = 3603) Group C ( N  = 2438) P uterine scar diverticulum 397 (7.8%) 313 (8.7%) 424 (17.4%) < 0.001 uterine scar diverticulum (with AUB) 285 (5.6%) 227 (6.3%) 295 (12.1%) < 0.001 AUB abnormal uterine bleeding Values are presented as a number (percentage) Rate of uterine scar diverticulum according to the timing of cesarean section AUB abnormal uterine bleeding Values are presented as a number (percentage) According to the results of statistical analyses, compared with the cesarean section before labor, cesarean section in the first stage of labor did not increase the risk of CSD (OR: 1.13, 95% CI: 0.57–2.13, P  = 0.435) and the risk of symptomatic CSD (OR: 1.41, 95% CI: 0.58–2.04, P  = 0.443). However, compared with women who underwent cesarean section before labor, women who underwent cesarean section during the second stage of labor had a significantly increased risk of CSD (OR: 2.59, 95% CI: 1.23–5.52, P  < 0.001) and symptomatic CSD (OR: 2.53, 95% CI: 1.30–5.31, P  < 0.001). To adjust for the influence of potential confounding factors on the results of bivariate comparison, we included the following factors for a multivariate analysis: maternal age at birth, maternal BMI at birth, smoking status of the included women, and medical comorbidities. The multivariate analysis shows that compared with the cesarean section before labor, the cesarean section after full dilation of the cervix significantly increases the risk of CSD (aOR: 2.67, 95% CI: 1.31–4.99, P  = 0.015) and symptomatic CSD (aOR: 2.45, 95% CI: 1.21–4.67, P  = 0.012). Table  4 shows the results of bivariate and multivariate analysis. Table 4 Bivariable and multivariable logistic regression of uterine scar diverticulum among women with cesarean section (reference group is group A) uterine scar diverticulum uterine scar diverticulum (with AUB) OR, 95% CI P aOR, 95% CI P OR, 95% CI P aOR, 95% CI P B versus A 1.13 (0.57–2.13) 0.435 1.15 (0.71–1.92) 0.287 1.41 (0.58–2.04) 0.443 1.32 (0.57–1.97) 0.614 C versus A 2.59 (1.23–5.52) < 0.001 2.67 (1.31–4.99) 0.015 2.53 (1.30–5.31) < 0.001 2.45 (1.21–4.67) 0.012 AUB abnormal uterine bleeding, aOR odds ratio, adjusted, CI confidence interval, OR odds ratio A: cesarean section before labor B: the first-stage cesarean section C: the second-stage cesarean section Bivariable and multivariable logistic regression of uterine scar diverticulum among women with cesarean section (reference group is group A) AUB abnormal uterine bleeding, aOR odds ratio, adjusted, CI confidence interval, OR odds ratio A: cesarean section before labor B: the first-stage cesarean section C: the second-stage cesarean section