Cesarean Delivery Complicated by Peripartum Infection and Risk of Uterine Rupture During Subsequent Trial of Labor

Abstract

Background:

Uterine rupture is a rare, but dangerous obstetric complication that can occur during trial of vaginal birth.

Methods:

The aim of this study was to evaluate the relationship between peripartum infection at the first caesarean delivery to uterine dehiscence or rupture at the subsequent delivery. We conducted a retrospective case-control study from March 2014 to October 2020 at a single academic medical center. The study group included women with a prior caesarean delivery and proven dehiscence or uterine rupture diagnosed during their subsequent delivery. The control group included women who went through a successful vaginal birth after cesarean section (VBAC) without evidence of dehiscence or uterine rupture. We compared the rate of peripartum infection during the first cesarean delivery (CD) and other relevant variables, between the two groups.

Results:

A total of 168 women were included, 71 with uterine rupture or dehiscence and 97 with successful VBAC as the control group. The rate of peripartum infection at the first caesarean delivery was significantly higher in the study group compared to the control group (22.2% vs. 8.2%, p = 0.013). Multivariate logistic regression analysis found that peripartum infection remained an independent risk factor for uterine rupture at the subsequent trial of labor after CD (95% confidence interval, p = 0.034).

Conclusion:

Peripartum infection in the first caesarean delivery, may be an independent risk-factor for uterine rupture in a subsequent delivery.

Introduction

Cesarean delivery (CD) is one of the most common surgeries worldwide, performed on one in three women in the United States and even four in five women in other regions of the world.¹ Once delivered by CD, the rate of a subsequent vaginal birth after cesarean section (VBAC) is low, and estimated to be 13.3% in the United States.² Based on previously published data, the implications of repeated CD on morbidity are profound and include higher rates of blood transfusions, surgical injuries, and hysterectomies, with increased rates of complications as the number of CD increases.^3,4 Therefore, many health care organizations try to increase VBAC rates. When consulting a woman regarding the mode of delivery, the risks and benefits of trial of labor after cesarean (TOLAC) should be discussed as well.⁵

Each woman should be advised according to her own obstetric and demographic characteristics, as well as her personal risk-factors for uterine rupture.⁶ Known risk factors, such as previous arrest of descent, vertical uterine scar, and so on⁷ are well established, but infection of surgical site in the first CD is a possible meaningful risk-factor that has not been investigated in depth. The few studies that have been published are small and most did not isolate the infection as an independent risk factor.⁸ Previous peripartum infections are not included into consultations regarding the preferred mode of delivery after a CD.

Peripartum infections may influence scar tissue healing and consequently affect uterine scar strength and ability to remain intact during subsequent pregnancies and deliveries. Peripartum infections include chorioamnionitis during labor, postpartum endometritis, and other surgical site infections. Surgical site infection tends to occur within 30 days of the operation⁹ and may affect superficial or deep tissues. Pathogens commonly associated with obstetrical infections are gram-negative bacilli, enterococci, Group B streptococci, and anaerobes.¹⁰

It is well-established that surgical scar infection is a risk-factor for impaired healing.¹¹ The pathogens interfere with leukocyte function and angiogenesis. Free radicals and cytotoxic enzymes are released from neutrophils,^12,13 which may interfere with the healing process and leave the scar weaker and more vulnerable. Although it is not commonly considered a risk-factor for uterine rupture during labor, a surgical scar with previous Surgical site infection (SSI) may have a higher tendency to rupture during a subsequent trial of labor.

The objective of this study was to investigate whether peripartum infection in the first CD is an independent risk-factor for uterine rupture in a subsequent delivery.

Materials and Methods

This retrospective case-control study was performed in a single academic medical center from March 2014 to October 2020. Women with uterine rupture or dehiscence were included in the study group and compared to women who had a successful TOLAC (the control group). Demographic and obstetrical data as well as any record of infection during or within a month after a prior CD were investigated.

Rupture of the uterus was defined as complete disruption of the previous cesarean scar with communication between the uterine and abdominal cavities. Dehiscence of the scar was defined as an incomplete disruption that included ruptured uterine muscle with intact serosa as the only layer keeping the fetus in the uterus. All cases of rupture and dehiscence were identified during CD.

The electronic medical records, which are filled thoroughly and prospectively were reviewed for demographic characteristics (age, BMI, gravity, parity) and for medical information (timing of first CD regarding phase of delivery, surgical technique, interdelivery interval), and neonatal birth weight. Infection was defined as either intrapartum with maternal temperature above 38°C during or up to 24 hours after delivery. Postpartum fever was defined as maternal temperature above 38°C, 24–48 hours after delivery, and SSI was defined according to the 2018 ACOG criteria and included endometritis, infected hematoma or infection of surgical wound up to 30 days from the previous CD.¹⁴ To set the diagnosis of SSI, we used not only the ICD9 headlines but also analyzed laboratory data of wound cultures, nonprophylactic antibiotics prescriptions, as well as clinical impression as written by the medical team in the electronic files.

Based on a uniform protocol, all patients undergoing CD received standard infection-prevention measures including preoperative intravenous (IV) antibiotic prophylaxis with one to two grams of cefamezine based on maternal weight, surgical length, and blood loss. For patients allergic to penicillin, clindamycin and gentamicin were administered.¹⁵

Intrapartum fever was treated with a combination of ampicillin and gentamicin, while patients with intrapartum fever undergoing CD also received IV metronidazole. Postpartum fever and endometritis were treated with IV Augmentin up to 48 hours from their last fever, according to protocols suggested by the ACOG.¹⁶ Patients with large infected abdominal hematomas or any other severe infection were treated with a triad of ampicillin gentamicin and IV metronidazole until improvement was noted or positive cultures allowed more specific treatment according to the institutional guidelines.

We compared three groups and subgroups: (1) All patients with rupture or dehiscence compared to controls (no rupture or dehiscence), (2) Rupture versus dehiscence, and (3) Rupture versus controls. We evaluated each group in terms of the type of infection as a risk-factor for rupture or dehiscence: intrapartum chorioamnionitis, postpartum endometritis, or SSI.

The Institutional Review Board of Meir Medical Center approved the study, which was performed in accordance with the Ethical Standards of the Declaration of Helsinki and its later amendments. Informed consent was waived due to the retrospective design of the study.

Statistical analysis was performed using the SPSS, version 21.0 software (IBM Corp., Armonk, NY). Continuous variables were compared by Mann–Whitney U-test. Chi-square or Fisher's exact test was used for categorical data, as appropriate. Demographics and medical data were compared between the study and control group using univariate analysis. Potential confounders and significant variables in the univariate analysis were entered into multivariable binary regression analysis to establish their effect on the risk for uterine rupture. A probability value <0.05 was considered significant.

Results

A total of 71 women were included in the study group, of which 40 experienced a complete thickness uterine rupture and 31 had uterine scar dehiscence. They were compared to 97 cases of successful VBAC by maternal age, BMI, and gestational age and neonatal birth weights of the first and second deliveries.

The basic characteristics of the study and the control groups are shown in Table 1. Maternal age, gravidity, and parity were higher in the study group, as expected by the study design. The study group included women with dehiscence or asymptomatic rupture in a repeat CD, whereas the control group included only women after their first VBAC, which means they only had one prior CD. Although the study group was statistically older than the study group (27.8 ± 4.6 years vs. 30.2 ± 4.8 years, p = 0), this was not clinically significant. There were no significant differences between the groups regarding neonatal birth weight at first delivery and a statistical but not clinically significant difference in birth weight at the second delivery. These potential confounders were included in the logistic regression (Tables 2 and 3). Infection in the previous delivery remained statistically significant with odds ratio of 3.17.

Table 1.

Characteristics of the Groups

First CD	Study group: rupture and dehiscence (n = 71)	Control group: successful VBAC (n = 97)	p	Study group: complete uterine rupture (n = 40)	P
Maternal age, years ± SD	30.2 ± 4.8	27.8 ± 4.6	0.00	29.9 ± 4.54	0.03
Gravidity	3.69	3	0.01	3.4	0.16
Parity	1.96	1.5	0.03	1.8	0.22
Timing of the CD, N (%)
Elective	24 (35)	27 (28)	0.34	9 (24)	0.69
Emergent before labor	6 (9)	10 (10)	0.73	4 (11)	0.93
Emergent at first stage of labor	33 (48)	42 (43)	0.57	21 (57)	0.17
Emergent at second stage of labor	6 (9)	18 (19)	0.06	3 (8)	0.09
Postpartum hemorrhage, n (%)	4 (6)	4 (4)	0.6	4 (11)	0.24
Neonatal birth weight, grams	2933 ± 737.8	3024.03 ± 652.2	0.42	2730 ± 798	0.05
Second delivery
Maternal age, years ± SD	33.46 ± 4.9	29.94 ± 4.63	0.00	33.1 ± 4.43	0.00
BMI	29.12 ± 5.07	29.07 ± 4.37	0.97	28.35 ± 5.44	0.62
Obesity rate (%)^a	21	33.3	0.12	0	<0.01
Months since CD	38.7 ± 24.4	28.67 ± 10.7	0.00	41.34 ± 25.4	0.00
Induction, n (%)	0	2 (2)	0.16	0	0.37
Gestational age, weeks	36.97 ± 3.41	38.95 ± 1.39	0.00	36.6 ± 3.7	0.00
Augmentation (Pitocin), n (%)	1 (1)	1 (1)	0.81	1 (3)	0.92
Neonatal birth weight, G ± SD	2993.3 ± 669	3196.4 ± 434	0.03	2916 ± 709	0.03

Obesity—BMI >30.

BMI, body mass index; CD, cesarean delivery; SD, standard deviation; VBAC, vaginal birth after cesarean section.

Table 2.

Table of Multiple Logistic Regression Including Possible Confounders to Demonstrate the Independent Association Between Infection and Uterine Rupture and Dehiscence

Variable	OR	95% CI	p
Infection	3.17	1.08–9.24	0.034
Interval between deliveries	1.03	1.00–1.05	0.019
Gravity	1.15	0.93–1.42	0.185
Maternal age	1.11	1.02–1.20	0.010
Neonatal birth weight	0.99	0.99–0.99	0.045

CI, confidence interval; OR, odds ratio.

Table 3.

Table of Multiple Logistic Regression Including Possible Confounders to Demonstrate the Independent Association Between Infection and Uterine Rupture Only

Variable	OR	95% CI	p
Infection	4.07	1.16–14.24	0.028
Interval between deliveries	1.05	1.01–1.08	0.004
Gravity	1.01	0.76–1.34	0.907
Maternal age	1.09	0.98–1.20	0.082
Neonatal birth weight	0.99	0.99–0.99	0.004

Sixteen of 71 (22.5%) women with rupture or dehiscence of the uterus during the current birth, had an infection in proximity to their previous CD, compared to 8 of 97 (8.2%) in the control group (p = 0.01).

We also compared the group with complete uterine rupture to the control group. Among 40 women with uterine rupture, 10 (25%) had an infection at their first CD, compared to 8 of 97 (8.2%) of the controls (p = 0.008). This result remained statistically significant in the logistic regression (Table 3). There was no difference regarding the infection rate between the uterine rupture group and the dehiscence group (25% vs. 16%, p = 0.27).

We divided the type of infection into three groups: intrapartum fever, postpartum fever, and SSI. SSI and, in particular, endometritis, had the strongest correlation to uterine rupture at subsequent delivery. None of the control group patients had endometritis, but 7 of 71 (9.8%) in the study group and 5 of 40 (12.8%) in the complete rupture group did. No significant difference was found between groups in terms of intrapartum fever (Table 4).

Table 4.

Infection Type

Variable	Control group (n = 97)	Study group (n = 71)		Complete rupture (n = 40)
Intrapartum fever	8 (8.2%)	3 (4.2%)	p = 0.63	1 (2.5%)	p = 0.42
Postpartum fever	0	5 (10%)	p = 0.02	2 (5%)	p = 0.16
Endometritis	0	7 (9.8%)	p = 0.002	5 (12.5%)	p = 0.02

Comment

This study found that women who experienced an infection at their previous CD had a significantly higher chance of uterine rupture at their next delivery. The type of infection found to be most correlated with uterine rupture was SSI, with endometritis being the most common type, and the least correlated was chorioamnionitis (intrapartum fever), meaning that the timing of infection may have a decisive effect on the risk of rupture or dehiscence.

Infection as a risk-factor for uterine rupture has not been studied in detail in the past. To the best of our knowledge, our cohort of 71 cases of uterine rupture is the largest to explore this subject, including 40 cases of complete rupture and 31 cases of dehiscence, as well as 97 controls.

Lieberman et al.⁷ found a correlation between infection and rupture. The rate of postpartum fever of any kind was 38% (8/21) in the uterine rupture group compared to 15% (13/84) in the control group (p = 0.03).⁸ Like our study, no statistical difference in the intrapartum fever group was found. They reviewed a database of patients from 1984 to 1996, with 70%–75% receiving antibiotic prophylaxis. Our results are based on deliveries from 2014 to 2022, with 100% prophylaxis treatment according to the ACOG guidelines and a very organized, prospective data collection system, based on the ICD9 headlines and on detailed review of the electronic medical records.

Gabbay-Benziv et al. found a correlation between complications at first CD (infection, PPH) and uterine rupture during TOLAC (p = 0.042).¹⁷ They had 3 cases of infection, all of which were postpartum fever documented during postpartum admission (4 days). In our study, 24 cases of infections were documented, 13 of which were diagnosed after readmission due to symptoms of infection.

Vilchez et al. used a large database of patients, and found that chorioamnionitis during the index delivery is a risk-factor for uterine rupture (odds ratio: 5.7).¹⁸ They concluded that this is probably due to the association with protracted labor. Our study explored the impact of the infection itself on the uterine scar, evaluated at the next delivery.

Since our data come from a single tertiary center, the treatment and surgical protocols were relatively uniform. Only 1% of the study group and 3% of the control group were treated with Pitocin for augmentation during TOLAC (p = 0.81). One patient in the control group was induced using an Atad catheter, compared to none in the study group. All uterine closures of the first CD were done in two layers. The time interval between the CD and next delivery was 38.7 ± 24.4 months in the study group and 28.6 ± 10.7 months in the control group (p = 0.00). However, there were large variations within the groups. It can be assumed that the experience of a traumatic labor (such as one accompanied by infection or rehospitalization) may affect the time interval between deliveries, but there are many other impact factors and this topic requires further research.

We analyzed two groups: one including women with dehiscence or uterine rupture (n = 71) and a subgroup of women with complete uterine rupture (n = 40). We did so because we believe that uterine rupture and dehiscence are one sequence of the same entity with the same etiology. Thus, it is important to include cases of dehiscence in the analysis. On the contrary, we understand that including them may pose a confounder because some patients in the control group may have had a subclinical dehiscence during their VBAC, and since we do not perform a routine scar test after successful VBAC, these cases could be misdiagnosed. Therefore, we chose to present the groups together and separately and show that in both, infection in the first delivery was found to be a risk-factor for uterine rupture.

A limitation of this study is its retrospective nature, as well as the small number of patients. Including the dehiscence cases caused the groups to not be homogeneous in terms of gravity, parity, and maternal age. However, none of these variables was significant in the regression analysis, showing that infection is well-correlated with uterine rupture. The average gestational age was younger in the study group compared to the controls and birth weights were smaller. We believe this may be due to symptoms of rupture that developed earlier and led to an emergent CD.

This study investigated a possible new and original risk-factor for a disastrous event. Further studies are needed to establish this finding, but this is a manageable risk and its incidence can be reduced.

Knowing that infection in general and endometritis specifically may increase the risk for uterine rupture in the next delivery, should inform the treating staff on the following issues: when consulting with a patient on repeat CD versus TOLAC, endometritis after the previous CD should be included in the considerations, along with other traditional risk factors. In addition, our results emphasize the importance of adhering to the guidelines for perioperative preventive treatment of infection, and finally, fast and accurate treatment of diagnosed infection in general and endometritis, in particular.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

References

Antoine

, Young

. Cesarean section one hundred years 1920–2020: The Good, the Bad and the Ugly. J Perinat Med, 2020; 49(1):5–16; doi: 10.1515/jpm-2020-0305

Osterman

MJK

. Recent trends in vaginal birth after cesarean delivery: United States, 2016–2018. NCHS Data Brief. 2020; 359:1–8.

Cook

, Jarvis

, Knight

, et al. Multiple repeat caesarean section in the UK: Incidence and consequences to mother and child. A National, Prospective, Cohort Study. BJOG, 2013; 120(1):85–91; doi: 10.1111/1471-0528.12010

Marshall

, Fu

, Guise

. Impact of multiple cesarean deliveries on maternal morbidity: A systematic review. Am J Obstet Gynecol, 2011; 205(3):262.e1–8 doi: 10.1016/j.ajog.2011.06.035

Joseph

, Shiliang

, Muraca

, et al. Mode of delivery after a previous cesarean birth, and associated maternal and neonatal morbidity. CMAJ, 2018; 190(18):E556–E564; doi: 10.1503/cmaj.170371

Grobman

, Lai

, Landon

, et al. Development of a nomogram for prediction of vaginal birth after cesarean delivery. Obstet Gynecol, 2007; 109(4):806–812; doi: 10.1097/01.AOG.0000259312.36053.02

Ronel

, Wiznitzer

, Sergienko

, et al. Trends, risk factors and pregnancy outcome in women with uterine rupture. Arch Gynecol Obstet, 2012; 285(2):317–321; doi: 10.1007/s00404-011-1977-8

Shipp

, Zelop

, Cohen

, et al. Post-cesarean delivery fever and uterine rupture in a subsequent trial of labor. Obstet Gynecol, 2003; 101(1):136–139; doi: 10.1097/00006250-200301000-00026

Gur

, Duggal

, Rongpharpi

, et al. Post caesarean surgical site infections. Arch Clin Microbiol, 2015; 6(1):1–6; doi: 10.1016/j.ajic.2008.04.112

10.

Owens

, Stoessel

. Surgical site infections: epidemiology, microbiology and prevention. J Hosp Infect, 2008; 70 Suppl 2:3–10 doi:10.1016/S0195-6701(08)60017-1

11.

Wolcott

, Rhoads

, Dowd

. Biofilms and chronic wound inflammation. J Wound Care, 2008; 17(8):333–341; doi: 10.12968/jowc.2008.17.8.30796

12.

Davis

, Ricotti

, Cazzaniga

, et al. Microscopic and physiologic evidence for biofilm-associated wound colonization in vivo. Wound Repair Regen, 2008; 16(1):23–29; doi: 10.1111/j.1524-475X.2007.00303.x

13.

Black

, Costerton

. Current concepts regarding the effect of wound microbial ecology and biofilms on wound healing. Surg Clin North Am, 2010; 90(6):1147–1160; doi: 10.1016/j.suc.2010.08.009

14.

Infections

. ACOG Practice Bulletin No. 195 Summary: Prevention of infection after gynecologic procedures. Obstet Gynecol, 2018; 131(6):1177–1179; doi: 10.1097/AOG.0000000000002672

15.

Antibiotics

. ACOG Practice Bulletin No. 199: Use of Prophylactic Antibiotics in Labor and Delivery. Obstet Gynecol, 2018; 132(3):e103–e119; doi: 10.1097/AOG.0000000000002833

16.

Gynecologists T American College of Obstetricians and Intrapartum Management of Intraamniotic Infection. Obstet Gynecol. 2017; 130(2):e95–e101.

17.

Salman

, Hiersch

, Shmueli

, et al. Complicated primary cesarean delivery increases the risk for uterine rupture at subsequent trial of labor after cesarean. Arch Gynecol Obstet, 2018; 298(2):273–277; doi: 10.1007/s00404-018-4801-x

18.

Vilchez

Gustavo

, Nazeer

, Kumar

, et al. Contemporary epidemiology and novel predictors of uterine rupture: A nationwide population-based study. Arch Gynecol Obstet, 2017; 296(5):869–875; doi: 10.1007/s00404-017-4508-4