Obesity is a major health issue among women of childbearing age. Class III obesity is defined as a body mass index (BMI, calculated as weight in kilograms divided by height in meters squared) of 40 or higher, based on the World Health Organization guidelines.1 In the 2011–2012 National Health and Nutrition Examination Survey, 7.7% of women between the ages of 20–39 years had a BMI of 40 or higher. Within this age group, non-Hispanic black women had the highest prevalence of class III obesity (17.5%) compared with non-Hispanic white women (6.8%), Asian women (1.1%), and Hispanic women (5.8%).2 A higher BMI is associated with hypertensive disorders of pregnancy, gestational diabetes, and stillbirth that result in a higher likelihood of induction of labor.3–5 Although one prior study of women with a BMI of 40 or higher suggested no increase in risk of cesarean delivery after undergoing induction of labor, multiple others have found this group of women to be at increased risk of cesarean delivery and associated morbidity.4,6,7
Various studies have sought to develop practical models to estimate risk of cesarean delivery in women who are undergoing induction of labor.8–11 In these studies, BMI is an independent predictor of cesarean delivery. In addition to BMI, maternal comorbidities such as chronic hypertension and diabetes, maternal age, height, parity, gestational weight gain, and cervical dilation have been identified as risk factors for cesarean delivery after induction of labor. These risk factors have been used to create models that predict the rate of cesarean delivery in women who are undergoing induction of labor.9,11 Risk models have also been use to predict morbidity in the setting of cesarean delivery. In women who underwent induction of labor with an unfavorable cervix, Hamm et al9 found that a higher predicted risk of cesarean delivery using a previously validated model was associated with increased maternal and neonatal morbidity.
Women with class III obesity who are undergoing induction of labor have a high failure rate of induction and significant resultant morbidity. However, risk factors for failed induction specifically among morbidly obese women have not fully been assessed. Furthermore, risk assessment models have not been fully evaluated in this cohort of women. Therefore, we sought to identify risk factors associated with cesarean delivery in women with class III obesity who were undergoing an induction. We then evaluated the performance of a previously derived and validated cesarean delivery risk calculator for labor induction in our cohort of women with class III obesity.8
We conducted a retrospective cohort study of women with class III obesity who underwent induction of labor at the Hospital of the University of Pennsylvania and Pennsylvania Hospital from January 1, 2013, through December 31, 2015. The study was approved by the Institutional Review Board at the University of Pennsylvania (#824496).
All women with a BMI of 40 or higher at the time of delivery and a singleton pregnancy at 34 weeks or greater of gestation were included. Body mass index was calculated based on maternal weight at time of delivery or last prenatal visit within 2 weeks of delivery, if weight at delivery was not available. Height was recorded in inches based on self-report from prenatal data at the time of initial prenatal visit. Pregnancy dating was confirmed by last menstrual period consistent with first- or second-trimester ultrasound scan or ultrasound scan alone, per current guidelines.12 Women with one prior cesarean delivery were included. Pregnancies with major fetal anomalies, preexisting fetal genetic disorder, multiple gestations, women in spontaneous labor and those with a contraindication to induction of labor were excluded.
Obese patients were identified using our institution's electronic delivery record. Individual chart review for prespecified inclusion and exclusion criteria, including eligibility for induction of labor, was performed by two of the investigators (C.P.T., L.R.), who also reviewed data after entry for accuracy. Demographic variables, prenatal data, induction parameters and delivery information were abstracted and entered into a REDCap electronic database.13
Induction of labor was defined as 1) absence of regular contractions (fewer than four in 10 minutes) at the time of admission; 2) cervical dilation 3 cm or less; and 3) use of one or a combination of ripening or induction agents, including pharmacologic (Misoprostol, Dinoprostone, Pitocin), mechanical (intracervical balloon catheter), or amniotomy. Active labor was defined as 6 cm dilation or greater with regular contractions (four or more in 10 minutes). Variables evaluated as risk factors for cesarean delivery included BMI category, maternal age, parity, height, pregestational or gestational diabetes, indication for induction, cervical dilation at the start of induction, modified Bishop score (cervical dilation, effacement, station) in addition to the individual components of the modified Bishop score at start of induction and gestational age at delivery.
The primary outcome was cesarean delivery for any indication. Based on the 2012 Eunice Kennedy Shriver National Institute of Child Health and Human Development guidelines, failed induction was defined as “failure to generate regular contractions and cervical change after at least 24 hours of oxytocin with artificial rupture of membranes if feasible.”14 Arrest disorders were also defined based on criteria outlined in the Eunice Kennedy Shriver National Institute of Child Health and Human Development guidelines.14 All data were analyzed by BMI category (40–49.9, 50–59.9, and 60 or more). Secondary outcomes were analyzed as composites of maternal and neonatal morbidity. The maternal composite included postpartum hemorrhage (1,000 mL estimated blood loss or greater), need for transfusion of blood products (packed red blood cells, fresh frozen plasma or platelets), operative injury (uterine extension, uterine artery laceration requiring O'Leary or compression stitch), bladder or bowel injury requiring repair, and hysterectomy. The maternal composite also included infectious complications of postpartum endometritis (defined as temperature greater than 38°C with fundal tenderness), culture-proven urinary tract infection with more than 100,000 colony forming units, wound cellulitis treated with antibiotics and culture proven sepsis. Postpartum ileus or bowel obstruction, wound complication (breakdown or dehiscence), venous thromboembolism diagnosed on imaging, maternal intensive care unit admission or death were also included in the maternal composite. The neonatal composite included 5-minute Apgar score less than 7, arterial cord pH 7.15 or less, and neonatal intensive care unit admission longer than 48 hours. Prolonged maternal length of stay (more than 2 days for vaginal delivery and 4 days or more for cesarean delivery) was also reviewed and reported.
For this study, we included all women who met the inclusion criteria over the course of 2 years at our institution. We chose this time period to ensure there were no contemporary practice pattern changes for labor induction. Our sample size included 428 women who met inclusion criteria in our study. We performed a post hoc power analysis. Using our fixed sample size, we had approximately 80% power to identify a clinically significant odds ratio (OR) of 1.4 or higher for the association between nulliparity and cesarean delivery in women with a BMI of 40 or higher who underwent induction of labor.
Continuous variables were analyzed with t test and Mann-Whitney U test. Categorical data were analyzed using Pearson χ2 and Fisher exact tests, where appropriate. Multivariable logistic regression models were then used to assess the relationship between maternal and neonatal clinical characteristics and the primary and secondary outcomes. All factors with P<0.2 were considered for inclusion in multivariable modeling. Confounders were retained in the model if their exclusion altered the association between the primary risk factors and outcome by more than 10%. All data analyses were performed using STATA 14.2. Statistical significance was set a P<.05.
Once risk factors were verified, we sought to validate the performance of a previously derived calculator to predict risk of cesarean delivery in our population of women with class III obesity.8 Clinical characteristics and calculator variables from the initial study (nulliparity, gestational age at induction 40 weeks or more, BMI category at delivery, modified Bishop score, and maternal height) were compared with the current cohort. In an effort to replicate the same inclusion and exclusion criteria used in the original cohort from which the induction of labor calculator was derived and subsequently validated, we excluded women with a prior cesarean delivery and those delivering preterm at less than 37 weeks of gestation (n=73). A total of 355 women were included in this analysis. All women were categorized as having a BMI of 40 or higher for the purposes of using the original calculator. The area under the receiver operating characteristic (ROC) curve was used as a measure of the ability of the calculator to discriminate the likelihood of cesarean delivery in women with a BMI of 40 or higher who underwent an induction of labor.
A total of 485 women with a BMI of 40 or higher underwent an induction of labor during the study period from January 1, 2013, through December 31, 2015, and 428 women met inclusion criteria for final analysis (Fig. 1). Clinical characteristics of women in the cohort by mode of delivery are presented in Table 1. More than three quarters of the women were black and 55% were nulliparous. Of those included, 81.8% (n=350/428) had a BMI of 40–49.9, 14.5% (n=62/428) had a BMI of 50–59.9, and 3.7% (n=16/428) had a BMI of 60 or higher. The overall cesarean delivery rate for the cohort was 49.1% (210/428). A higher BMI was associated with a higher likelihood of cesarean delivery. More specifically, in women undergoing induction, cesarean delivery was performed in 46% of women with a BMI of 40–50, 63% with a BMI of 50–60, and 69% with a BMI higher than 60 (P=.012). More than two thirds of women with a BMI of 50 or higher underwent cesarean delivery. On univariate analysis, nulliparity, BMI category, a shorter maternal height, chronic hypertension, more unfavorable initial cervical examination, preterm delivery, and modified Bishop score were associated with a significantly higher rate of cesarean delivery. Almost 50% of cesarean deliveries were performed for nonreassuring fetal status, with the remaining 50% due to maternal indication. There was no difference in the indication for induction among women who underwent cesarean delivery compared with those who underwent vaginal delivery.
Table 2 presents the unadjusted and adjusted ORs from the multivariable logistic regression models. All variables noted to have significance in the univariate analyses or based on biological plausibility or previous work were assessed for inclusion in the multivariable logistic regression model. Our covariates for the adjusted model included BMI category, parity, height, cervical dilation, indication for induction and modified Bishop score. Those risk factors identified as significant in the unadjusted analysis remained significant in the multivariable model. A patient with a BMI of 50–59.9 had two times greater odds of cesarean delivery compared with those with a BMI of 40–49.9 (adjusted OR 2.08; 95% CI 1.11–3.93). Nulliparity was associated with 4.8 times greater odds of cesarean delivery compared with multiparous women with a history of vaginal delivery (adjusted OR 4.08; 95% CI 3.02–7.59). For each 1-inch increase in patient height, their odds of undergoing cesarean delivery decreased by 15%.
Secondary outcomes including the maternal and neonatal composite morbidities by mode of delivery are listed in Table 3. Overall maternal morbidity was significantly higher in obese women who underwent cesarean delivery, compared with those who delivered vaginally (51% vs 5%), and remained higher after adjustment for the same covariates as the primary outcome (adjusted OR 19.66; 95% CI 9.31–41.51, P<.001). In the cohort, 44% (n=91) of women undergoing cesarean delivery experienced a postpartum hemorrhage, and 6% (n=12) required blood transfusion compared with a postpartum hemorrhage rate of only 4% (n=8) and a 1% (n=3) rate of transfusion in the vaginal delivery group (P<.001 and P=.017). Neonatal morbidity was also significantly higher in women who underwent cesarean delivery compared with those who had a successful vaginal delivery (30% vs 16%) and also remained significant after adjustment (adjusted OR 2.58; 95% CI 1.50–4.41, P<.001).
We then sought to evaluate a previously validated cesarean delivery risk calculator in this large population of class III obese pregnant women. As demonstrated in Figure 2, when applying this calculator to our entire cohort, the area under the ROC curve was 75% (95% CI 0.70–0.79).8 The analysis was then repeated after excluding those with a prior cesarean delivery and those who delivered at less than 37 weeks of gestation. Using the inclusion criteria from the original study, the area under the ROC curve was 77% (95% CI 0.72–0.82).
In this retrospective cohort study of class III obese women who underwent induction of labor, there was a significant increase in rate of cesarean delivery that correlates with maternal and neonatal morbidity compared with those delivering vaginally. The rate of cesarean delivery in this cohort approaches 50% with an increase to 66% in women with a BMI of 50 or greater. In our study, a BMI of 50 or higher, nulliparity, height and cervical dilation were the most significant risk factors for cesarean delivery. Importantly, maternal and neonatal morbidity were both increased in those undergoing cesarean delivery after induction compared with those delivering vaginally. The composite maternal morbidity in the cesarean delivery group was 51% compared with 5% in the vaginal group (P=.001). Furthermore, 30% of neonates delivered by cesarean had associated morbidity compared with only 16% of neonates that delivered vaginally (P=.001).
The cesarean delivery rate after induction in our cohort is comparable with the rates previously reported in women with a BMI of 40 or higher, which ranges from 30% to 63%.3,6,12 A multivariable analysis by Ronzoni et al4 showed that obese women were almost 2.5 times more likely to undergo cesarean delivery after induction of labor compared with nonobese women, even after controlling for maternal age, parity, fetal macrosomia, maternal weight gain, and cervical dilation. The risk factors associated with cesarean delivery identified in our study are similar to those previously described and include increasing BMI class, nulliparity, shorter height and an unfavorable cervix.4,7,15 The rate of cesarean delivery in nulliparous women in our cohort was 66% (n=156), which is also similar to previously reported rates of cesarean delivery (60–64%) after induction in nulliparous women with a BMI greater than 60.3,12 We found that nulliparous women with a BMI of 40 or higher who were undergoing induction with cervical dilation of 0–0.5 cm had a 70% risk of cesarean delivery in our study, whereas those with cervical dilation of 2–3 cm underwent cesarean delivery in 56% of cases in our cohort. This emphasizes the effect of multiple risk factors in women with class III obesity. Our study was not powered to assess the rate of cesarean delivery after induction in women without medical or fetal indication because only 15 women had an elective induction.
In obese women undergoing induction of labor, cesarean delivery is often associated with greater risk of operative complications, wound morbidity, postpartum infection, and neonatal morbidity compared with women who have a vaginal delivery.4 Subramaniam et al16 studied a group of women undergoing induction of labor with class III obesity and demonstrated a 3.7 times greater composite maternal morbidity in women delivering by cesarean after induction compared with those delivering vaginally. Our cohort demonstrated a higher rate of postpartum hemorrhage (44%) in women undergoing cesarean delivery after induction in women with class III obesity compared with prior studies.15,16 Our results also demonstrate increased neonatal morbidity after cesarean delivery compared with vaginal delivery, with 16% (n=34) of neonates delivered by cesarean requiring neonatal intensive care unit admission for more than 48 hours compared with only 7% (n=16) in the vaginal group (P=.004).7 In the cesarean delivery group, 17% (n=35) had an arterial cord pH of 7.15 or less compared with 8% (n=18) in the vaginal delivery group (P=.008).
Our study further emphasizes the importance of identifying risk factors that accurately predict cesarean delivery after induction. This information can be discussed with patients and families to facilitate informed discussions about delivery. Prediction models that use clinical risk factors can add to this important discussion for patients who are at risk of maternal and neonatal morbidity in the setting of a nonscheduled cesarean delivery. Rossi et al created a predictive model for cesarean delivery after analyzing risk factors in a population-based cohort study. The risk factors identified in the cohort included maternal weight, height and age, parity, gestational weight gain, prior vaginal or cesarean delivery and presence or absence of diabetes or hypertension. These factors were then used to assess risk of cesarean delivery. The study was a population-based cohort of women with a BMI of 30 or higher and included women who underwent induction of labor between 37 and 44 weeks of gestation. The overall cesarean delivery rate of women undergoing induction was 24.9%, with a rate of 31.1% in women with a BMI of 40 or higher.10 Although our study demonstrates similar risk factors, only women with class III obesity were included. Additionally, our study was able to include starting cervical examination and modified Bishop score, which were included in our analysis.
The clinical characteristics found to be most predictive of cesarean delivery in our study of women with class III obesity are similar to those identified in a recent study by Levine et al.8 In this study, a calculator reports the predicted probability of cesarean delivery based on risk factors in both normal, overweight and obese women. However, only 18.5% of the total cohort had a BMI of 40 or higher (n=91). Our group was able to further analyze this model in a large sample of women with class III obesity. After amending our cohort to include variables used in the original study, we were able to demonstrate that the cesarean delivery risk calculator accurately predicted the risk of cesarean delivery in women with class III obesity who were undergoing induction of labor to the same degree as was seen in the initial study. Although the risk calculator should not be used in isolation, it can be included in clinical discussions when counseling a woman about expectations at the time of induction. It is also important to note that there is no set value above which cesarean delivery without a trial of labor can be recommended. Further studies are needed to determine whether such a cutoff exists where morbidity associated with nonscheduled cesarean delivery is so great that an outright cesarean delivery would be preferable. The calculator can be found at http://www.uphs.upenn.edu/labor-induction-calculator/.8
The strengths of our study include a large number of women with class III obesity. This study included women from two teaching hospitals, in an urban area which increases generalizability. Individual chart review was performed and, therefore, abstraction did not rely on ICD coding. In contrast to the majority of studies, our findings were strengthened by only evaluating women with class III obesity in addition to using a strict definition for labor induction confirmed by individual chart review as evidenced by more than 90% of women in the cohort receiving Pitocin.15,16 Our findings provide additional data specifically for use when counseling women with class III obesity at the time of induction. Additionally, we were able to apply a user-friendly calculator for clinical use in this large subgroup of women with class III obesity who underwent induction of labor. Although an understudied population, our cohort included more than 75% non-Hispanic black women which does not necessarily translate to other ethnic subgroups. Limitations include the retrospective nature of our data. Additionally, height was a self-reported variable and was not directly measured. These factors carry the potential for self-reporting and misclassification bias. Misclassification bias could be seen in labeling of a patient as an “induction” when the patient may have actually been in early labor. This may result in an overestimation of vaginal deliveries after a presumed induction of labor, thereby leading us to report a lower cesarean delivery rate than the true rate in women with BMI of 40 or higher who were undergoing induction of labor. However, our stringent criteria for induction made the inclusion of women included in early labor unlikely. Further, our data are not able to account for provider practice variation that may occur when managing induction of labor in obese women. At the current time, further studies are needed to provide better counseling on the safest delivery mode in women with class III obesity.
1. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser 2000;894:1–253.
2. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011-2012. JAMA 2014;311:806–14.
3. Weiss JL, Malone FD, Emig D, Ball RH, Nyberg DA, Comstock CH, et al. Obesity, obstetric complications and cesarean delivery rate—a population-based screening study. Am J Obstet Gynecol 2004;190:1091–7.
4. Ronzoni S, Rosen H, Melamed N, Porat S, Farine D, Maxwell C. Maternal obesity class as a predictor of induction failure: a practical risk assessment tool. Am J Perinatol 2015;32:1298–304.
5. Trojner Bregar A, Tul N, Fabjan Vodusek V, Verdenik I, Lucovnik M, Janša V, et al. A dose-response relation exists between different classes of pre-gravid obesity and selected perinatal outcomes. Arch Gynecol Obstet 2017;296:465–8.
6. Kawakita T, Iqbal SN, Huang CC, Reddy UM. Nonmedically indicated induction in morbidly obese women is not associated with an increased risk of cesarean delivery. Am J Obstet Gynecol 2017;217:451.e1–8.
7. Wolfe KB, Rossi RA, Warshak CR. The effect of maternal obesity on the rate of failed induction of labor. Am J Obstet Gynecol 2011;205:128.e1–7.
8. Levine LD, Downes KL, Parry S, Elovitz MA, Sammel MD, Srinivas SK. A validated calculator to estimate risk of cesarean after an induction of labor with an unfavorable cervix. Am J Obstet Gynecol 2018;218:254.e1–7.
9. Hamm RF, Downes KL, Srinivas SK, Levine LD. Using the probability of cesarean from a validated cesarean prediction calculator to predict labor length and morbidity. Am J Perinatol 2019;36:561–6.
10. Rossi RM, Requarth EW, Warshak CR, Dufendach K, Hall ES, DeFranco EA. Predictive model for failed induction of labor among obese women. Obstet Gynecol 2019;134:485–93.
11. Tolcher MC, Holbert MR, Weaver AL, McGree ME, Olson JE, El-Nashar SA, et al. Predicting cesarean delivery after induction of labor among nulliparous women at term. Obstet Gynecol 2015;126:1059–68.
12. Ultrasound in pregnancy. Practice Bulletin No. 175. American College of Obstetricians and Gynecologists. Obstetrics Gynecol 2016;128:e241–56.
13. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42:377–81.
14. Spong CY, Berghella V, Wenstrom KD, Mercer BM, Saade GR. Preventing the first cesarean delivery: summary of a joint Eunice Kennedy Shriver National Institute of Child Health and Human Development, Society for Maternal-Fetal Medicine, and American College of Obstetricians and Gynecologists workshop. Obstet Gynecol 2012;120:1181–93.
15. Gunatilake RP, Smrtka MP, Harris B, Kraus DM, Small MJ, Grotegut CA, et al. Predictors of failed trial of labor among women with an extremely obese body mass index. Am J Obstet Gynecol 2013;209:562.e1–5.
16. Subramaniam A, Jauk VC, Goss AR, Alvarez MD, Reese C, Edwards RK. Mode of delivery in women with class III obesity: planned cesarean compared with induction of labor. Am J Obstet Gynecol 2014;211:700.e1–9.