Cesarean delivery is the most common surgery performed in women worldwide. In the United States alone, there were more than 1.2 million cesarean deliveries performed in 2014, representing more than 32% of all births.1 Over the past 50 years, the maternal and neonatal short-term risks associated with cesarean deliveries have decreased despite the rise in cesarean delivery rates from an estimated 5% in the 1960s to more than 30% in 2015.2 In light of the ever evolving safety profile of a cesarean delivery, women are now increasingly undergoing elective cesarean deliveries on the basis of maternal requests.3
Although there have been many studies evaluating the immediate and short-term complications of cesarean deliveries,4–6 few studies have evaluated the long-term morbidity associated with this delivery modality.7 Given the appreciable degree of adhesion formation that is often associated with cesarean deliveries,8 it is not unlikely that a cesarean delivery may result in a long-term elevated risk of small bowel obstruction. Small bowel obstruction is a long-term complication of abdominal surgery that results in significant morbidity and mortality. To date, there has been no large study evaluating the long-term surgical risks associated with cesarean deliveries, particularly as it pertains to small bowel obstruction. The purpose of our study was to evaluate the association of cesarean delivery with the incidence of small bowel obstruction.
MATERIALS AND METHODS
We performed a population-based cohort study using the U.K. Clinical Practice Research Datalink and the linked U.K. Hospital Episode Statistics repository from 1998 to 2015. The Clinical Practice Research Datalink is a governmental research service that provides high-quality data on outpatient primary care services for public health research and has been available since 1987. Individual patients cared for by general practitioners registered within the Clinical Practice Research Datalink were linked to the Hospital Episode Statistics data sets to provide data on admission to the hospital, creating an outpatient and inpatient follow-up for purposes of this study.
Cohort entry was restricted to live births occurring during the 10-year period between January 1, 1998, and December 31, 2007. First, patients who had a unique delivery code within the Hospital Episode Statistics were identified. Only women who were identified as having had a first live birth during the study period were included in the cohort using the date of the first live birth as the point of cohort entry. The Clinical Practice Research Datalink was then used to identify and exclude women who had a birth before the study start date or who had a small bowel obstruction before cohort entry. Among the remaining women, only those with at least 1 year of “up-to-standard” data before the birth and 6 months of follow-up were included in the study. “Up-to-standard” refers to practices that submit data to the Clinical Practice Research Datalink and undergo regular audits meeting internal quality standards. We elected to restrict women with a first live birth to the study for two reasons: first, women with a stillbirth are very seldom at risk of cesarean delivery and, thus, would not be representative of the risk of the exposure, and second, those who are beyond their first birth are at an inherently different risk of cesarean delivery than a first birth.
We then used the Hospital Episode Statistics to classify exposure status as having had or not had a first cesarean delivery. Person-time was considered unexposed until the time that a cesarean delivery occurred. Our main outcome was the occurrence of a small bowel obstruction occurring beyond 30 days after delivery. The outcome was identified from several sources: 1) Clinical Practice Research Datalink data—using bowel obstruction medical codes and 2) Hospital Episode Statistics data—using the International Classification of Disease, 10th Revision codes and the Classification and Procedures version 4 from admission records. Total follow-up time was measured from the time of first live birth until the earliest of the following: 1) exit from the Clinical Practice Research Datalink, 2) death by any cause, 3) January 26, 2015 (end of follow-up), or 4) cesarean delivery. Women were considered unexposed if they had a vaginal birth at first delivery and exposed if they had a cesarean delivery at the first delivery. For those women who initially had a vaginal birth but underwent a cesarean delivery for a subsequent pregnancy, follow-up time was considered to be unexposed from the time of the vaginal birth to the cesarean delivery at which point subsequent follow-up time was considered exposed.
Analysis was carried out in three steps. First, descriptive statistics were used to compare baseline clinical characteristics of women who had a first vaginal birth with women who had a first cesarean delivery. Characteristics were obtained both from clinical outpatient practice (Clinical Practice Research Datalink ) and from the hospital admission records (Hospital Episode Statistics). Obesity was defined as body mass index (calculated as weight (kg)/[height (m)]2) greater than 30 or presence of the coding variable “obesity” within the Clinical Practice Research Datalink clinical record. Second, Kaplan-Meier analysis was carried out to compare the time-to-event from vaginal delivery with that of a cesarean delivery on the occurrence of a first small bowel obstruction. Third, Cox proportional hazard analyses were run to evaluate the overall association of cesarean delivery with the occurrence of a first small bowel obstruction, the association of multiple cesarean deliveries with a first small bowel obstruction, and the association of cesarean delivery with operative and nonoperative small bowel obstruction. Finally, we conducted a Poisson regression analysis to evaluate the association of cesarean delivery with the number of small bowel obstruction episodes during the total follow-up time. All analyses were adjusted for baseline characteristics (Table 1).
Statistical analyses were carried out using the SAS Enterprise Guide 6.1. The Independent Scientific Advisory Committee of the Clinical Practice Research Datalink (Protocol No. 13_181RA) and the Research Ethics Board of the Jewish General Hospital, Montreal, Quebec, Canada, approved our study (Protocol No. 14-007).
As shown in Figure 1, there were 158,135 unique patients identified between January 1, 1998, and December 31, 2007. After excluding patients who had a first stillbirth or a live birth before the cohort start date, or had a prior small bowel obstruction, 121,256 women remained. Of these, 81,480 women met the remaining follow-up criteria. Among these women, 20,689 (25.4%) had a first cesarean delivery and 60,791 (74.6%) had a first vaginal delivery during the study period. A total of 575 new small bowel obstructions were identified within our cohort (incidence 9.1/10,000 person-years), which included a total follow-up time of 634,012 person-years, ranging from 6 months to 16.6 years.
At baseline, women with a first cesarean delivery were older with a slightly greater proportion having had abdominal surgery. No differences in race were noted (Table 1). Kaplan-Meier survival analysis indicates that women who have undergone a cesarean delivery are at increased risk of developing a small bowel obstruction as compared with those who have had a vaginal delivery (Fig. 2).
There were 280 cases of small bowel obstruction among women with a first cesarean delivery (1.35%) compared with 295 cases of small bowel obstruction among women with a first vaginal delivery (0.49%). The Cox proportional hazard models demonstrate that women are at significantly increased risk of having a small bowel obstruction if they have had a cesarean delivery compared with those who have never had a cesarean delivery (hazard ratio [HR] 2.54, 95% CI 2.15–3.00). Increasing numbers of cesarean deliveries were found to be associated with a greater risk of small bowel obstruction (HR 1.61, 95% CI 1.46–1.78) (Table 2). Although not in the table, this association remained in analyses restricted to women with no prior surgery. Schoenfeld residuals and proportionality testing were as well carried out and demonstrated a constant risk of small bowel obstruction over time.
Although small bowel obstructions treated with or without surgery were more common among women with a cesarean delivery, the requirement for a surgical intervention was even more common than was a small bowel obstruction without a surgical intervention (HR 2.72, 95% CI 2.28–3.25 and HR 1.82, 95% CI 1.21–2.74, respectively) (Table 3).
Repeated small bowel obstruction events were more common in women with a prior cesarean delivery than in women with no prior cesarean deliveries (Table 4). The number of distinct small bowel obstructions was greater among women with a cesarean delivery than among those with vaginal deliveries. When restricting our analyses to women who had any small bowel obstruction, women with a cesarean delivery had more small bowel obstructions than women who had a small bowel obstruction after a vaginal delivery (relative risk 2.62, 95% CI 2.30–2.97).
Cesarean delivery is the most common abdominal surgery performed on women worldwide. Our results suggest that cesarean deliveries are associated with an increased risk of both medically and surgically treated small bowel obstructions and that this risk rises with increasing number of cesarean deliveries.
It is known that major abdominal surgeries are associated with the formation of adhesions.8–11 Although adhesions in general are recognized as risk factors for small bowel obstruction,12–15 there have only been a few studies suggesting that cesarean deliveries in particular lead to small bowel obstruction. First, a study by Al-Took et al12 evaluated the association between gynecologic surgeries and small bowel obstruction. In their cohort study on 262 women with adhesion-related small bowel obstruction, the authors concluded that abdominal hysterectomy was the most common gynecologic surgery associated with small bowel obstruction and that, likely, cesarean delivery was also associated. Another study using the Swedish Medical Birth Registry used a case–control approach to evaluate the association of cesarean delivery with subsequent intestinal obstruction and found a twofold increase in risk of intestinal obstruction.16
Small bowel obstruction and associated morbidities have an important effect both at an individual patient level as well as at a hospital and health care level. In a recent meta-analysis, adhesions resulted in a considerable requirement for readmissions to the hospital, multiple surgeries for adhesiolysis, bowel injury, and reduced pregnancy rates.17 In a review of hospital effect, small bowel obstruction resulting from adhesions had an important effect on hospital workload and hospital bed use, with an estimated equivalent effect of one full-time surgical bed per year and more than 3% of all laparotomies performed in the hospital.18 With rising cesarean delivery rates, it is likely that these problems will become an important public health concern.
Our primary analysis was a time-to-event analysis adjusting for baseline characteristics. Initially, our research protocol planned for marginal structural models with the intention to adjust for time-dependent variables such as subsequent laparotomies and changing obesity rates, which may have been time-dependent confounders as well as intermediates. On conducting our initial analyses, it became evident that those events were too rare to provide a material contribution to our understanding of the cesarean delivery—small bowel obstruction relationship and, thus, marginal structural model analysis was not carried out as initially planned.
There are several limitations to our study. First, our study is a database study with inherent concerns as to the accuracy of reporting. However, given that our exposure of interest (mode of delivery) is one that is seldom improperly coded in administrative databases, it is unlikely that misclassification of our exposure would affect our results. In a study evaluating a birth registry in British Columbia, Frosst et al19 estimated a near 100% specificity in the validation of primary cesarean deliveries. As for the outcome of interest, our results remained consistent when restricting to surgical as well as nonsurgical small bowel obstruction, suggesting a consistency in association in both forms of small bowel obstruction. Second, our median follow-up of roughly 8 years limits the more long-term effect that may exist between cesarean delivery and small bowel obstruction. Although the trend suggests that the risk of small bowel obstruction increases with time, our study cannot evaluate the longer term morbidity that may occur with extended follow-up. Finally, although cohort inclusion was limited to women who had a first live birth during the index period, there is the possibility that this may not have been a woman's first birth. For instance, if a woman changed practices between deliveries, she would receive a new Clinical Practice Research Datalink user identification number and we would not be able to track her previous deliveries. Although this may have resulted in misclassification, it is unlikely to have resulted in a material difference in our overall findings because these women would likely still provide a “history of prior birth” limiting the extent of potential misclassification. Our study does have important strengths. First, our exposure is prospectively measured relative to the outcome, considerably reducing the possibility of information bias. Our study is a large cohort with sufficient power to evaluate the effect of cesarean delivery on small bowel obstruction, an established rare, but serious, event.
Cesarean delivery is an important risk factor for the development of surgical and nonsurgical small bowel obstruction. With rising cesarean delivery rates worldwide, small bowel obstruction-related morbidities may become a more important women's health concern.
1. Hamilton BE, Martin JA, Osterman MJK, Curtin SC, Mathews TJ. Births: final data for 2014. National vital statistics reports; vol 64, no 12. Hyattsville (MD): National Center for Health Statistics; 2015.
2. Hamilton BE, Martin JA, Osterman MJK, Driscoll AK, Mathews TJ. Births: final data for 2015. National vital statistics reports; vol 66, no 1. Hyattsville (MD): National Center for Health Statistics; 2017.
3. Coleman VH, Lawrence H, Schulkin J. Rising cesarean delivery rates: the impact of cesarean delivery on maternal request. Obstet Gynecol Surv 2009;64:115–9.
4. Hebert PR, Reed G, Entman SS, Mitchel EF Jr, Berg C, Griffin MR. Serious maternal morbidity after childbirth: prolonged hospital stays and readmissions. Obstet Gynecol 1999;94:942–7.
5. Liu S, Liston RM, Joseph KS, Heaman M, Sauve R, Kramer MS, et al. Maternal mortality and severe morbidity associated with low-risk planned cesarean delivery versus planned vaginal delivery at term. CMAJ 2007;176:455–60.
6. Nisenblat V, Barak S, Griness OB, Degani S, Ohel G, Gonen R. Maternal complications associated with multiple cesarean deliveries. Obstet Gynecol 2006;108:21–6.
7. Silver RM. Implications of the first cesarean: perinatal and future reproductive health and subsequent cesareans, placentation issues, uterine rupture risk, morbidity, and mortality. Semin Perinatol 2012;36:315–23.
8. Poole JH. Adhesions following cesarean delivery: a review of their occurrence, consequences and preventative management using adhesion barriers. Womens Health 2013;9:467–77.
9. Diamond MP, Daniell JF, Feste J, Surrey MW, McLaughlin DS, Friedman S, et al. Adhesion reformation and de novo adhesion formation after reproductive pelvic surgery. Fertil Steril 1987;47:864–6.
10. Szigetvari I, Feinman M, Barad D, Bartfai G, Kaali SG. Association of previous abdominal surgery and significant adhesions in laparoscopic sterilization patients. J Reprod Med 1989;34:465–6.
11. Davey AK, Maher PJ. Surgical adhesions: a timely update, a great challenge for the future. J Minim Invasive Gynecol 2007;14:15–22.
12. Al-Took S, Platt R, Tulandi T. Adhesion-related small-bowel obstruction after gynecologic operations. Am J Obstet Gynecol 1999;180:313–5.
13. Moran BJ. Adhesion-related small bowel obstruction. Colorectal Dis 2007;9(suppl 2):39–44.
14. Diamond MP, Freeman ML. Clinical implications of postsurgical adhesions. Hum Reprod Update 2001;7:567–76.
15. van Goor H. Consequences and complications of peritoneal adhesions. Colorectal Dis 2007;9(suppl 2):25–34.
16. Andolf E, Thorsell M, Kallen K. Cesarean delivery and risk for postoperative adhesions and intestinal obstruction: a nested case-control study of the Swedish Medical Birth Registry. Am J Obstet Gynecol 2010;203:406.e1–6.
17. ten Broek RP, Issa Y, van Santbrink EJ, Bouvy ND, Kruitwagen RF, Jeekel J, et al. Burden of adhesions in abdominal and pelvic surgery: systematic review and met-analysis. BMJ 2013;347:f5588.
18. Menzies D, Parker M, Hoare R, Knight A. Small bowel obstruction due to postoperative adhesions: treatment patterns and associated costs in 110 hospital admissions. Ann R Coll Surg Engl 2001;83:40–6.
19. Frosst G, Hutcheon J, Joseph KS, Kinniburgh B, Johnson C, Lee L. Validating the British Columbia Perinatal Data Registry: a chart re-abstraction study. BMC Pregnancy Childbirth 2015;15:123.