Objective
The purpose of this study was to examine predictors associated with cesarean delivery (CD) among extremely obese women undergoing a trial of labor (TOL).
Study design
Using a delivery database, we identified all pregnant women delivering at our institution from Jan. 1, 2008, through July 31, 2010, weighing >275 lb at the time of delivery who attempted a TOL with a singleton gestation >34 weeks’ gestation. Demographic and obstetrical factors were compared for those having a successful vaginal delivery to those having a CD.
Results
During the study period, there were 357 pregnant women who weighed >275 lb (all with body mass index [BMI] >40 kg/m 2 ), and among these, 248 (69.5%) attempted a TOL. Women having a CD had a greater BMI (51.6 vs 49.9 kg/m 2 , P = .038), were less likely to be parous (32.2% vs 65.8%, P < .0001), and were more likely to be induced (80.5% vs 57.8%) compared to those having a vaginal delivery. Using a multivariable logistic regression model, among nulliparous women, maternal age, parity, and cervical dilation at time of admission were independent predictors for CD. Furthermore, an increase in BMI of 10 kg/m 2 was associated with a 3.5 increased odds ( P = .002) for CD.
Conclusion
Among nulliparous extremely obese women attempting a TOL, BMI was an independent predictor of CD, with the rate of CD increasing further with increasing BMI. The underlying mechanisms for failed TOL in the setting of maternal obesity remain largely unknown.
In the United States, approximately one third of women aged 20-39 years are obese and more than half are classified as overweight. A body mass index (BMI) of ≥30 kg/m 2 is categorized as obese and obesity is divided into 3 classes, according to the World Health Organization: class I (BMI 30-34.9 kg/m 2 ), class II (35-39.9 kg/m 2 ), and class III (≥40 kg/m 2 ). Approximately 8% of reproductive-aged women fall into the class III category, also referred to as extremely obese or morbidly obese. Although there is yet a consensus definition for obesity beyond BMI ≥40 kg/m 2 , some authors have described women with a BMI >50 kg/m 2 as super or massive obesity.
Obesity is a risk factor for cesarean delivery (CD). Emerging evidence suggests that the process of labor may differ in obese women in comparison to women of normal weight. Moderate to strong associations have been described of increasing maternal weight among women undergoing labor induction with increased labor duration, oxytocin requirements, and risk of failed induction. In a large population-based cohort study, Wolfe et al observed that rates of labor induction increased with increasing BMI where 34% of women with BMI >40 kg/m 2 underwent labor induction and these women had higher rates of failed induction of labor compared to normal-weight women. In a secondary analysis of a labor induction trial, obese women were less likely to achieve active labor, required greater oxytocin during labor induction, and experienced greater time to transition to active labor than women with a normal BMI.
Obese women who undergo CD are at increased risk for infectious morbidity and surgical morbidity is further increased with CD following a failed trial of labor (TOL). Total intraoperative time is also increased significantly for women with a BMI >40 kg/m 2 who undergo elective CD.
It is unclear whether obesity is the driving force for the observed failure of labor progression in induced labor, or whether other factors, such as baseline cervical condition or not allowing for adequate time to achieve active labor account for the differences in labor course for the obese woman. Evidence suggest that the pattern of labor progression differs for women with electively induced labor compared to those with spontaneous labor and that elective induction in nulliparous women with an unfavorable cervix have a high rate of labor arrest and CD. In neither of these studies nor the study on the natural history of the normal first stage of labor was there a control for obesity. The aim of this study was to determine factors associated with a failed TOL among women with class III obesity.
Materials and Methods
The protocol was approved by the Duke University Medical Center Institutional Review Board. Using a delivery database, we identified all pregnant women delivering at our institution from Jan. 1, 2008, through July 31, 2010, weighing >275 lb at the time of delivery admission. Those women weighing >275 lb with a singleton gestation, who were >34 weeks’, and undergoing spontaneous or induced labor were included. Women with a multiple gestation, delivered <34 weeks, or whose pregnancies were complicated by congenital abnormalities or intrauterine fetal death were excluded. The primary outcome of interest was failed TOL defined by CD.
For the purposes of this study, all women who received oxytocin either for induction or augmentation of labor were classified as having oxytocin exposure. Misoprostol, oxytocin, amniotomy, and Foley catheter placement were the modalities used for labor induction during the study period. Cervical dilation on admission was used in lieu of a complete Bishop score or other cervical scoring systems because Bishop score was not consistently recorded for the majority of our subjects. The choice for method of induction or cervical ripening was left to the discretion of the admitting clinician. Cervical ripening was with either 25 μg of misoprostol every 4 hours until an adequate contraction pattern was achieved or a transcervical Foley catheter with gentle traction. Subjects with a favorable cervical examination result were typically managed either with amniotomy or oxytocin infusion according to an institutional protocol. Indications for CD were for standard definitions. Attending physicians, residents, and midwives at Duke University Hospital provided the intrapartum management throughout the study period.
Baseline demographic variables including BMI, race, and induction status were compared between those who had a failed TOL (CD) and those having a vaginal delivery. Obstetrical variables including birthweight, medical comorbidities, and rates of chorioamnionitis and shoulder dystocia were also compared. BMI was analyzed both as a continuous variable and as 3 groups (40-50, 50-60, and >60 kg/m 2 ). Secondary outcomes that were potentially related to our primary outcome were also compared: chorioamnionitis, postpartum hemorrhage, shoulder dystocia, and operative assistance. Failed induction of labor was defined as a subject not in active labor despite labor augmentation. Active labor was defined as a cervical examination of at least 4-6 cm in the setting of regular uterine contractions.
Continuous variables were compared using a Student t test and categorical variables compared using χ 2 . A multivariable logistic regression model for the outcome CD was created using predictors that were significantly associated with mode of delivery or were thought to be clinically relevant to mode of delivery. Univariate odds ratios (OR) with 95% confidence intervals (CI) were calculated for each of the predictors for the outcome CD. The interaction between BMI and parity was found to be significant so the multivariate logistic regression model was run separately for nulliparous and parous women. Adjusted ORs were then reported for each predictor, while controlling for all other predictors in the model.
To determine if extremely obese women (BMI ≥40 kg/m 2 ) undergoing a TOL have different rates of CD compared to non-extremely obese women (BMI <40 kg/m 2 ) undergoing a TOL, a cohort of women with a BMI <40 kg/m 2 were identified and compared to the extremely obese group. The non-extremely obese women were identified from the Duke University Hospital delivery database as the first 248 consecutive women who underwent a TOL and delivered beginning July 2009, with a BMI <40 kg/m 2 . The mean age, BMI, gestational age of delivery, and mode of delivery was determined for this group. We used χ 2 to compare the odds of CD between the extremely obese and non-extremely obese women undergoing a TOL, stratified by parity.
Statistical analysis was performed using software (JMP for Macintosh, version 9.0; SAS Institute Inc, Cary, NC). A P value < .05 was deemed significant.
Results
During the study period, there were 8197 total deliveries at Duke University Hospital, of which 3143 (38.3%) were by cesarean. Among these deliveries, there were 357 obese subjects weighing >275 lb who carried a singleton gestation ≥34 weeks’ gestation. Of these, 248 (69.5%) attempted a TOL and 109 (30.5%) had a CD without a TOL. Of the women who did not attempt a TOL, the indications for CD were as follows: repeat elective CD or prior myomectomy (n = 76, 69.7%), malpresentation (n = 10, 9.2%), nonreassuring fetal testing (n = 8, 7.3%), elective due to obesity and inability to monitor fetus (n = 5, 4.6%), placenta previa/accreta (n = 3, 2.7%), macrosomia (n = 3, 2.7%), maternal medical indications (n = 2, 1.8%), and fetal congenital anomalies (n = 2, 1.8%).
Obese women having a TOL were younger (31.0 vs 34.2 years, P = .001), had a lower mean BMI (50.5 vs 53.4 kg/m 2 , P = .003), had lower mean maternal weight (313 vs 323 lb, P = .05), and delivered at a later gestational age (38.6 vs 37.9 weeks, P = .01) compared to obese women proceeding directly to CD and not attempting a TOL. There were no differences in mean birthweight (3299 vs 3347 g, P = .55) between those women attempting a TOL compared to those not attempting a TOL. None of the subjects attempting a TOL had a prior CD, therefore, no subjects attempted a TOL after CD. Of the 248 women attempting a TOL, there were 136 women (54.8%) with a BMI of 40-50 kg/m 2 , 93 women (37.5%) with a BMI of >50-60 kg/m 2 , and 19 women (7.7%) with a BMI >60 kg/m 2 .
Table 1 provides demographic information by mode of delivery among the 248 subjects who attempted a TOL ( Table 1 ). There were no differences in age, race, or gestational age of delivery by mode of delivery. Women having a CD were less likely to be parous than those having a vaginal delivery. Maternal weight did not differ by mode of delivery, but women having a CD had a significantly greater BMI than those having a vaginal delivery, suggesting that women having a CD were shorter than those having a vaginal delivery. The overall CD rate for those attempting a TOL was 35.1% and when stratified by BMI category (40-50, >50-60, >60 kg/m 2 ) was 31.6%, 34.4%, and 63.2%, respectively ( P = .031). The effect of BMI on mode of delivery was even more pronounced among nulliparous women. Of nulliparous women, the CD rate for those attempting a TOL when stratified by BMI category (40-50, >50-60, >60 kg/m 2 ) was 39.7%, 65.0%, and 77.8%, respectively ( P = .010).
Characteristic | Vaginal delivery (n = 161) | CD (n = 87) | P value |
|---|---|---|---|
| Age, y a | 29.8 ± 6.0 | 30.6 ± 6.5 | .335 |
| Race, n (%) | .580 | ||
| Caucasian | 41 (25.5) | 16 (18.4) | |
| African American | 111 (68.9) | 66 (75.8) | |
| Hispanic | 5 (3.1) | 2 (2.3) | |
| Other | 4 (2.5) | 3 (3.4) | |
| Gestational age, wk a | 38.5 ± 2.1 | 38.7 ± 2.1 | .412 |
| Maternal weight, lb a | 312 ± 29 | 315 ± 42 | .548 |
| Maternal BMI, kg/m 2 a | 49.9 ± 5.6 | 51.6 ± 6.7 | .038 |
| Parous, n (%) | 106 (65.8) | 28 (32.2) | < .0001 |
Labor was induced in 163 (66%) of the subjects attempting a TOL. Indications for labor induction were: chronic hypertension (n = 52, 31.9%), preeclampsia (n = 48, 29.4%), pregestational or gestational diabetes (n = 19, 11.6%), postdates pregnancy (n = 14, 8.6%), other maternal medical conditions (lupus, coagulation disorder, cholestasis of pregnancy, maternal cardiac conditions; n = 5, 3.1%), fetal indications (fetal growth restriction, oligohydramnios, polyhydramnios; n = 18, 11.0%), elective (n = 4, 2.4%), and rupture of membranes (n = 3, 1.8%).
Women having a CD had a smaller median [quartile] cervical dilation at admission (1 [0, 2] vs 2 [1, 4] cm, P < .0001) compared to those having a vaginal delivery. There were no differences in birthweight or rates of chorioamnionitis between the 2 groups. Shoulder dystocia occurred in 6.8% of those obese women having a vaginal delivery and postpartum hemorrhage was more common among those women having a CD ( Table 2 ). The most common indication for CD was arrest of labor (n = 31, 35.6%) followed by failed induction of labor (n = 25, 28.7%) and nonreassuring (abnormal) fetal heart rate status (n = 25, 28.7%). Of the 31 women whose CD was performed for an arrest disorder, 9 women had a cervix dilation of 10 cm and their CD was performed for arrest of descent. Among those whose indication for CD was arrest of labor, the median (interquartile) cervical dilation at time of CD was 5.5 (4.25, 10) cm.
Characteristic | Vaginal delivery (n = 161) | CD (n = 87) | P value |
|---|---|---|---|
| Induced labor, n (%) | 93 (57.8) | 70 (80.5) | .0002 |
| Cervical dilation upon admission, cm a | 2 (1, 4) | 1 (0, 2) | < .0001 |
| Birthweight, g b | 3295 ± 579 | 3306 ± 694 | .894 |
| Chorioamnionitis, n (%) | 5 (3.1) | 4 (4.6) | .561 |
| Shoulder dystocia, n (%) | 11 (6.8) | 0 (0) | .002 |
| Postpartum hemorrhage, n (%) c | 8 (5.0) | 17 (19.5) | .0004 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
