Abstract
Objective
To evaluate whether elective induction of labor between 39 through 41 weeks, as compared to expectant management, is associated with reduced cesarean delivery and other adverse outcomes among obese women and their offspring.
Methods
We conducted a retrospective cohort study using the 2007–2011 California Linked Patient Discharge Data–Birth Cohort File of 165,975 singleton, cephalic, non-anomalous deliveries to obese women. For each gestational week (39–41), we used multivariable logistic regression models, stratified by parity, to assess whether elective induction of labor or expectant management was associated with lower odds of cesarean delivery and other adverse outcomes.
Results
At 39 and 40 weeks, cesarean delivery was less common in obese nulliparous women who were electively induced versus those who were expectantly managed (at 39 weeks, frequencies were 35.9% versus 41.0%, respectively [p<0.05]; adjusted odds ratio [OR] [95% Confidence Interval (CI)]: 0.82 [0.77, 0.88]). Severe maternal morbidity was less frequent among electively induced obese nulliparous patients (at 39 weeks, 5.6% versus 7.6% [p<0.05]; adjusted OR: 0.75 [95% CI 0.65, 0.87]). Neonatal intensive care unit (NICU) admission was less common among electively induced obese nulliparous women (at 39 weeks, 7.9% versus 10.1% [p<0.05]; adjusted OR: 0.79 [95% CI 0.70, 0.89]). Patterns were similar among obese parous women at 39 weeks (crude frequencies and adjusted ORs [95% CIs] were as follows: for cesarean delivery, 7.0% versus 8.7% [p<0.05], and 0.79 [0.73, 0.86]; for severe maternal morbidity, 3.3% versus 4.0% [p<0.05], and 0.83 [0.74, 0.94]; for NICU admission: 5.3% versus 7.4% [p<0.05], and 0.75 [0.68, 0.82]). Similarly, elective induction at 40 weeks was associated with reduced odds of cesarean delivery, maternal morbidity, and NICU admission among both obese nulliparous and parous patients.
Conclusion
Elective labor induction after 39 weeks was associated with reduced maternal and infant morbidity among obese women. Further prospective investigation is necessary.
INTRODUCTION
Maternal obesity (pre-pregnancy body mass index ≥30 kg/m2) increases the risks of adverse obstetric, fetal, and infant outcomes (1–3). These risks persist even in the absence of other chronic diseases (3). Despite the high U.S. prevalence of pre-pregnancy obesity (24.8%) (4) and the myriad of complications associated with this condition, a uniform standard of care regarding delivery timing and method does not exist specifically for obese gravid patients.
Obese women and their infants are at elevated risk of preeclampsia, macrosomia, shoulder dystocia, brachial plexus injury, meconium aspiration syndrome, and stillbirth (1,2,5,6), all of which increase with advancing gestational age (7–10). It is plausible that these adverse outcomes could be prevented through elective induction of labor and earlier delivery (11–13). However, the potential negative side effects of elective labor induction and earlier delivery must also be considered (14,15). The risks and benefits of elective labor induction, as compared to expectant management, at different gestational ages have not been thoroughly evaluated among obese gravid patients. Most previous research has focused on the general population (12), although some recent analyses have evaluated obese women specifically (7,16). Notably, although elective labor induction is not the only method to effect earlier delivery, it may be associated with fewer complications and lower costs than elective cesarean delivery (12,17,18).
Our objective was to assess whether elective induction of labor at each gestational week from 39—41 weeks, as compared to expectant management, was associated with reduced cesarean delivery and other adverse outcomes among obese gravid patients without chronic disease.
MATERIALS AND METHODS
In this retrospective cohort study, we used the 2007–2011 California Linked Patient Discharge Data–Birth Cohort File (19). To create this dataset, the California Office for Statewide Health Planning and Development linked vital records with maternal and infant hospital discharge data for deliveries in California, plus out-of-state deliveries to California residents. Over 95% of deliveries were successfully linked (19). Women could potentially be included more than once in our dataset, as deliveries occurring to the same woman in different years were not linked. We used hospital discharge data from the delivery visit, which included ICD-9-CM diagnostic and procedure codes. Medical diagnoses and procedures (including induction of labor, most study outcomes, and pregnancy characteristics) were coded as present if detected in either vital records or discharge data. This approach improves the sensitivity of detecting pregnancy complications while negligibly impacting specificity (20–22). Infant birthweight was taken from vital records data only.
Obese women (pre-pregnancy body mass index ≥30 kg/m2) with singleton, ≥39-week deliveries in cephalic presentation were included if they did not have preexisting medical complications (including chronic hypertension, preexisting diabetes, gestational diabetes, preexisting cardiac disease, preexisting renal disease, preexisting liver or biliary tract disorder, placenta previa, vasa previa, or isoimmunization), a prior cesarean delivery, or an infant with a major congenital anomaly that was likely to have affected clinical management (many of these anomalies would have been diagnosed prenatally). Observations were excluded due to missing data on study eligibility criteria, the exposure variable, or other covariates (observations with missing data on outcome variables remained eligible for inclusion in our study sample). Gestational age was defined by best obstetric estimate. Parity was defined as the number of previous pregnancies reaching ≥20 weeks of gestation. Pre-pregnancy body mass index was derived using vital records data.
We evaluated pregnancy outcomes at each gestational week (39–41) by comparing exposed women (those who underwent induction without medical indication in the given week) with unexposed women (expectantly managed women who delivered at a later gestational week). The expectant management group consisted of all women who delivered in later weeks, regardless of delivery method or labor onset type. Medical indications for labor induction (used to classify inductions as elective versus non-elective) were defined using recent Joint Commission guidelines (Appendix 1, available onlineat http://links.lww.com/AOG/B41, contains a list of conditions in this study) (7,23).
Our primary outcome was maternal mode of delivery (cesarean delivery, operative vaginal delivery [forceps or vacuum], and non-operative vaginal delivery). Secondary study outcomes included severe maternal morbidity (a composite outcome of postpartum hemorrhage, third- or fourth-degree perineal lacerations, unplanned surgical procedure, uterine rupture, maternal intensive care unit admission, maternal sepsis, and endometritis), infant mortality (death in first year of life), neonatal intensive care unit (NICU) admission (NICU stay >24 hours [for infants who died ≤24 hours after birth, any NICU admission], neonatal transfer, or infant hospital stay exceeding maternal postpartum stay (8)), macrosomia (birthweight ≥4500 grams), clinical chorioamnionitis, meconium aspiration syndrome, respiratory distress syndrome (RDS), shoulder dystocia, and brachial plexus injury.
We used χ2 and Fisher’s Exact tests to evaluate whether the distributions of maternal sociodemographic characteristics and pregnancy outcomes differed between electively induced and expectantly managed women. We modeled associations between elective labor induction and pregnancy outcomes using logistic regression. In multivariable analyses, we adjusted for maternal characteristics (age, race and ethnicity, education, obesity class, payment source, first-trimester prenatal care initiation), birth year (before 2009 versus 2009 or later), and hospital type (community or teaching). These covariates were selected a priori based on evidence of their associations with elective labor induction and the outcomes (7,8); we did not adjust for potential intermediates of the associations between elective labor induction and pregnancy outcomes. With the exceptions of hospital type, payment source, and birth year, covariates were classified using vital records.
We calculated crude and adjusted odds ratios (ORs) comparing elective labor induction in each individual week (39–41) to expectant management, stratifying by parity (nulliparous, parous). Specifically, each model compared electively induced deliveries during the given week to all deliveries in later weeks. In each model, spontaneous and medically indicated deliveries that occurred during the index week were excluded. We assessed effect measure modification by obesity class using χ2 tests of the interaction terms (p<0.20). In supplemental analyses, we calculated adjusted ORs stratified by both obesity class (1 to 3) and parity. In sensitivity analyses, we revised our list of indications for labor induction (Appendix 1, available online at http://links.lww.com/AOG/B41). We added certain intrapartum complications to the list of indications (coagulation deficiency hemorrhage, amniotic infection, and fetal distress or fetal heart rate abnormalities with unspecified time of onset) and removed others (placental abruption and fetal-maternal hemorrhage). These intrapartum complications could either be medical indications for labor induction or consequences of labor induction, depending on their timing.
We used SAS Version 9.4 (Cary, NC) for data analysis. This study was approved by the Emory University Institutional Review Board, the California Committee for the Protection of Human Subjects, and the California Office for Statewide Health Planning and Development. Informed consent was not necessary due to the de-identified nature of the dataset.
RESULTS
Out of 2,622,927 California deliveries occurring between 2007–2011, we excluded 2,456,952 ineligible births, mostly due to maternal pre-pregnancy BMI <30 kg/m2 (Figure 1). A total of 165,975 deliveries occurring at 39 weeks of gestation or later to obese women remained eligible for analysis.
Figure 1.
The top of this flow chart shows the number of women excluded from our sample due to study ineligibility (eg, BMI less than 30, pre-existing maternal conditions, multiple gestations) or missing data. Numbers of excluded observations do not overlap. The bottom of this flowchart represents the number of obese women analyzed at each gestational week (39 weeks through 41 weeks), stratified by parity. Some of these numbers overlap, because obese women who were expectantly managed at one gestational week could be electively induced at a later week. In each comparison (eg, elective induction at 39 weeks of gestation compared with expectant management), spontaneous and medically indicated deliveries that occurred during the index week were excluded. As a consequence, week-specific counts may not sum to the total number of observations.*Pre-existing medical conditions include chronic hypertension, pre-existing or gestational diabetes, pre-existing cardiac disease, pre-existing renal disease, pre-existing liver or biliary tract disorder, vasa or placental previa, or isoimmunization.
The frequencies of most maternal sociodemographic characteristics varied by exposure status and gestational age (Table 1). Electively induced obese women were more likely than expectantly managed obese women to be ≥25 years of age and to have initiated prenatal care in the first trimester. In addition, obese women who were electively induced at 39 or 40 weeks were more likely than expectantly managed obese women to be parous and to deliver in a community hospital.
Table 1.
Frequencies of Maternal Characteristics among Electively Induced and Expectantly Managed Obese Women
| 39 weeks | 40 weeks | 41 weeks | ||||
| Maternal characteristic | Elective Induction (n=13,568) | Expectant Management (n=95,094) | Elective Induction (n=17,809) | Expectant Management (n=25,279) | Elective Induction (n=9,909) | Expectant Management (n=1,928) |
| Parity | ||||||
| Nulliparous (%) | 29.1* | 42.8 | 39.6* | 49.3 | 50.1 | 50.2 |
| Obesity class | ||||||
| 1 (30.0– <35.0 kg/m2) (%) | 63.9 | 64.0 | 63.7* | 62.2 | 61.6 | 61.1 |
| 2 (35.0– <40.0 kg/m2) (%) | 23.7 | 24.0 | 24.1 | 24.7 | 25.2 | 24.2 |
| 3 (≥40.0 kg/m2) (%) | 12.4 | 12.0 | 12.2 | 13.1 | 13.3 | 14.7 |
| Maternal age, years | ||||||
| ≤24 (%) | 31.9* | 36.5 | 36.1* | 37.6 | 37.0* | 40.4 |
| 25–34 (%) | 54.9 | 52.2 | 52.1 | 51.3 | 52.3 | 49.7 |
| ≥35 (%) | 13.3 | 11.4 | 11.8 | 11.1 | 10.8 | 9.9 |
| Race and ethnicity | ||||||
| Hispanic (all races) (%) | 58.8* | 61.8 | 61.8* | 56.8 | 55.4 | 54.6 |
| Non-Hispanic Black (%) | 5.3 | 6.8 | 6.0 | 7.4 | 7.4 | 8.3 |
| Non-Hispanic White (%) | 29.4 | 25.0 | 26.2 | 29.3 | 31.3 | 30.4 |
| Other, non-Hispanic (%) | 6.5 | 6.4 | 6.0 | 6.5 | 5.9 | 6.7 |
| First trimester prenatal care initiation | ||||||
| Yes (%) | 84.0* | 79.3 | 80.7* | 75.9 | 77.5* | 68.0 |
| Year of delivery | ||||||
| 2007 or 2008 (%) | 38.3* | 39.9 | 40.7 | 40.5 | 39.6* | 45.3 |
| Delivery at teaching hospital | ||||||
| Yes (%) | 4.4* | 9.9 | 6.0* | 11.7 | 10.4 | 11.2 |
Table data are percentages (%) within exposure categories. No observations in our analytic sample were missing data on the exposure or covariates.
p<0.05, based on χ2 test examining whether covariate frequencies varied by exposure category.
The frequencies of adverse pregnancy outcomes also varied between electively induced and expectantly managed obese women (Table 2). Cesarean delivery was less common among electively induced, versus expectantly managed, obese women at 39 weeks of gestation (among nulliparous patients: 35.9% [95% Confidence Interval (CI) 34.1%, 37.8%] versus 41.0% [95% CI 40.4%, 41.6%], respectively); among parous patients: 7.0% [95% CI 6.4%, 7.6%] versus 8.7% [95% CI 8.4%, 9.0%], respectively; Table 2). Similarly, the frequency of cesarean delivery was lower among electively induced obese women at 40 weeks of gestation (e.g., among nulliparous women: 41.8% [95% CI 40.4%, 43.2%] versus 46.2% [95% CI 45.2%, 47.3%], respectively). In contrast, at 39 and 40 weeks, the risk of operative vaginal delivery was moderately increased among electively induced, as compared to expectantly managed, obese women (e.g., among nulliparous women at 39 weeks, 8.9% [95% CI 7.9%, 10.1%] versus 7.1% [95% CI 6.8%, 7.4%], respectively). Severe maternal morbidity occurred less frequently among obese women who were electively induced, versus expectantly managed, at 39 and 40 weeks of gestation (e.g., among nulliparous women at 39 weeks: 5.6% [95% CI 4.9%, 6.3%] versus 7.6% [95% CI 7.3%, 7.8%], respectively; among parous women at 39 weeks: 3.3% [95% CI 3.0%, 3.7%] versus 4.0% [95% CI 3.9%, 4.2%]). Infant death was rare, and it occurred at a similar frequency among electively induced and expectantly managed obese women at all gestational ages (e.g., among nulliparous women at 40 weeks, 0.1% [95% CI 0.0%, 0.2%] versus 0.1% [95% CI 0.1%, 0.2%]). NICU admission was less common among electively induced obese patients in all gestational weeks (e.g., among nulliparous patients at 39 weeks, 7.9% [95% CI 7.1%, 8.7%] versus 10.1% [95% CI 9.8%, 10.4%], respectively). The frequencies of several other adverse perinatal outcomes were also lower among electively induced obese patients (Table 2). Cesarean delivery, severe maternal morbidity, NICU admission, chorioamnionitis, RDS, and meconium aspiration syndrome increased between 2007 and 2011, while operative vaginal delivery, macrosomia, shoulder dystocia, and infant death decreased.
Table 2.
Distribution of Pregnancy Outcomes among Electively Induced and Expectantly Managed Obese Women and their Offspring
| Nulliparous women (n=65,333) | ||||||
| 39 weeks | 40 weeks | 41 weeks | ||||
| Outcome | Elective Induction (n=3,942) | Expectant Management (n=40,667) | Elective Induction (n=7,061) | Expectant Management (n=12,453) | Elective Induction (n=4,961) | Expectant Management (n=967) |
| Cesarean delivery, n (%) | 1,416 (35.9)* | 16,673 (41.0) | 2,948 (41.8)* | 5,757 (46.2) | 2,258 (45.5) | 478 (49.4) |
| Operative vaginal delivery (total), n (%) | 352 (8.9)* | 2,890 (7.1) | 554 (7.8)* | 848 (6.8) | 366 (7.4) | 59 (6.1) |
| Vacuum, n (%) | 327 (8.3) | 2,657 (6.5) | 525 (7.4) | 774 (6.2) | 327 (6.6) | 55 (5.7) |
| Forceps, n (%) | 18 (0.5) | 183 (0.5) | 22 (0.3) | 59 (0.5) | 31 (0.6) | 4 (0.4) |
| Unspecified or both (%) | 7 (0.2) | 50 (0.1) | 7 (0.1) | 15 (0.1) | 8 (0.2) | 0 (0.0) |
| Spontaneous vaginal births, n (%) | 2,174 (55.2)* | 21,104 (51.9)* | 3,559 (50.4)* | 5,848 (47.0) | 2,337 (47.1) | 430 (44.5) |
| Severe maternal morbidity, n (%)† | 221 (5.6)* | 3,071 (7.6) | 483 (6.8)* | 1,053 (8.5) | 408 (8.2) | 87 (9.0) |
| Infant death, n (%)‡ | 1 (0.0) | 48 (0.1) | 6 (0.1) | 15 (0.1) | 5 (0.1) | 0 (0.0) |
| NICU admission, n (%)§ | 311 (7.9)* | 4,104 (10.1) | 612 (8.7)* | 1,291 (10.4) | 449 (9.1)* | 137 (14.2) |
| Macrosomia, n (≥4500 g) (%) | 27 (0.7)* | 977 (2.4) | 144 (2.0)* | 430 (3.5) | 182 (3.7)* | 53 (5.5) |
| Chorioamnionitis, n (%) | 151 (3.8)* | 2,818 (6.9) | 354 (5.0)* | 1,037 (8.3) | 392 (7.9) | 91 (9.4) |
| Meconium aspiration syndrome, n (%)‡ | 20 (0.5)* | 398 (1.0) | 27 (0.4)* | 161 (1.3) | 45 (0.9)* | 19 (2.0) |
| Respiratory distress syndrome, n (%) | 74 (1.9) | 888 (2.2) | 134 (1.9)* | 296 (2.4) | 101 (2.0) | 26 (2.7) |
| Shoulder dystocia, n (%) | 62 (1.6) | 577 (1.4) | 115 (1.6) | 175 (1.4) | 90 (1.8) | 10 (1.0) |
| Brachial plexus injury, n (%)‡ | 3 (0.1) | 77 (0.2) | 15 (0.2) | 28 (0.2) | 12 (0.2) | 2 (0.2) |
| Parous women (n=100,642) | ||||||
| 39 weeks | 40 weeks | 41 weeks | ||||
| Outcome | Elective Induction (n=9,626) | Expectant Management (n=54,427) | Elective Induction (n=10,748) | Expectant Management (n=12,826) | Elective Induction (n=4,948) | Expectant Management (n=961) |
| Cesarean delivery, n (%) | 673 (7.0)* | 4,722 (8.7) | 892 (8.3)* | 1,289 (10.0) | 473 (9.6) | 111 (11.6) |
| Operative vaginal delivery (total), n (%) | 426 (4.4)* | 2,152 (4.0) | 532 (4.9)* | 486 (3.8) | 199 (4.0) | 34 (3.5) |
| Vacuum, n (%) | 399 (4.2) | 2,040 (3.8) | 504 (4.7) | 464 (3.6) | 187 (3.8) | 33 (3.4) |
| Forceps, n (%) | 23 (0.2) | 80 (0.2) | 19 (0.2) | 18 (0.1) | 11 (0.2) | 1 (0.1) |
| Unspecified or both (%) | 4 (0.0) | 32 (0.1) | 9 (0.1) | 4 (0.0) | 1 (0.0) | 0 (0.0) |
| Spontaneous vaginal births, n (%) | 8,527 (88.6)* | 47,553 (87.4) | 9,324 (86.8)* | 11,051 (86.2) | 4,276 (86.4) | 816 (84.9) |
| Severe maternal morbidity, n (%)† | 319 (3.3)* | 2,200 (4.0) | 396 (3.7)* | 623 (4.9) | 261 (5.3) | 50 (5.2) |
| Infant death, n (%)‡ | 5 (0.1) | 59 (0.1) | 15 (0.1) | 16 (0.1) | 5 (0.1) | 2 (0.2) |
| NICU admission, n (%)§ | 514 (5.3)* | 4,039 (7.4) | 656 (6.1)* | 1,023 (8.0) | 325 (6.6) | 79 (8.2) |
| Macrosomia (≥4500 g), n (%) | 169 (1.8)* | 1,804 (3.3) | 299 (2.8)* | 638 (5.0) | 259 (5.2)* | 67 (7.0) |
| Chorioamnionitis, n (%) | 62 (0.6)* | 583 (1.1) | 89 (0.8)* | 191 (1.5) | 78 (1.6) | 10 (1.0) |
| Meconium aspiration syndrome, n (%)‡ | 15 (0.2)* | 221 (0.4) | 26 (0.2)* | 74 (0.6) | 20 (0.4)* | 11 (1.1) |
| Respiratory distress syndrome, n (%) | 109 (1.1) | 622 (1.1) | 118 (1.1) | 174 (1.4) | 55 (1.1) | 14 (1.5) |
| Shoulder dystocia, n (%) | 201 (2.1)* | 1,553 (2.9) | 296 (2.8)* | 465 (3.6) | 186 (3.8) | 41 (4.3) |
| Brachial plexus injury, n (%)‡ | 20 (0.2) | 120 (0.2) | 16 (0.2)* | 41 (0.3) | 12 (0.2)* | 7 (0.7) |
Table data are frequencies (n) and percentages (%) within exposure categories.
p<0.05. Unless otherwise noted, p-values come from a 1 df χ2 test of whether outcome frequencies varied by exposure category. P-values for cesarean delivery, operative vaginal delivery, and spontaneous vaginal births (i.e., non-operative vaginal delivery) come from a 2 df χ2 test examining whether mode of delivery varies by exposure category. Although forceps delivery and vacuum extraction were combined in all analyses, the breakdown of forceps delivery versus vacuum extraction is shown here.
Of 10,529 women, 56.8% had postpartum hemorrhage, 31.9% had a third-or-fourth degree perineal laceration, and 11.3% experienced multiple or rare complications.
Fisher exact test examining whether outcome frequencies varied by exposure category.
NICU stay >24 hours (or, for infants who died <24 hours after birth, any NICU admission), neonatal transfer, or infant hospital stay longer than maternal postpartum hospital stay. Number of observations with missing values on NICU admission, by parity (nulliparous, parous): 39 weeks (21; 5), 40 weeks (1; 30), 41 weeks (12; 5). When considering all deliveries together (i.e., all types of labor onset and delivery), the frequency of NICU admission was lower among deliveries at 39–40 weeks versus deliveries at ≥41 weeks (e.g., among obese nulliparous women, 9.7% versus 10.4%, p<0.05; among obese parous women: 7% versus 8%, p<0.05).
There was no statistically significant effect measure modification by obesity class. Crude ORs (Table 3) were similar in magnitude to adjusted ORs (Figure 2; Table 4). In adjusted models, elective induction of labor between 39 and 40 weeks was associated with reduced odds of cesarean delivery (among obese nulliparous women, adjusted OR at 39 weeks=0.82 [95% CI 0.77, 0.88], and adjusted OR at 40 weeks=0.85 [95% CI 0.80, 0.90]; among obese parous women, adjusted OR at 39 weeks=0.79 [95% CI 0.73, 0.86], and adjusted OR at 40 weeks=0.81 [95% CI 0.74, 0.89]; Figure 2a). The odds of operative vaginal delivery were slightly elevated among electively induced obese nulliparous women at 39 weeks (adjusted OR=1.16 [95% CI 1.03, 1.31]) and electively induced obese parous women at 40 weeks (adjusted OR=1.25 [95% CI 1.10, 1.41]; Figure 2b). Elective labor induction between 39 and 40 weeks was associated with reduced odds of severe maternal morbidity, with adjusted ORs (95% CIs) ranging from 0.75 (0.65, 0.87) to 0.84 (0.75, 0.94) among obese nulliparous women and from 0.75 (0.66, 0.85) to 0.83 (0.74, 0.94) in obese parous women (Figure 2c).
Table 3.
Crude Odds Ratios for Elective Induction of Labor Compared With Expectant Management and Pregnancy Outcomes Among Obese Women
| Nulliparous women (n=65,333) | |||
| Outcome | Elective Induction, 39 weeks (n=3,942) vs. Expectant Management (n=40,667) | Elective Induction, 40 weeks (n=7,061) vs. Expectant Management (n=12,453) | Elective Induction, 41 weeks (n=4,961) vs. Expectant Management (n=967) |
| Cesarean delivery | 0.82 (0.77—0.88) | 0.84 (0.79—0.89) | 0.87 (0.75—1.00) |
| Operative vaginal delivery | 1.18 (1.05—1.33) | 1.07 (0.96—1.20) | 1.14 (0.85—1.53) |
| Severe maternal morbidity* | 0.73 (0.63—0.84) | 0.80 (0.71—0.89) | 0.91 (0.71—1.16) |
| Infant death | 0.21 (0.03—1.56) | 0.71 (0.27—1.82) | Did not converge |
| NICU admission† | 0.76 (0.68—0.86) | 0.82 (0.74—0.91) | 0.60 (0.49—0.74) |
| Macrosomia (≥4500 g) | 0.28 (0.19—0.41) | 0.58 (0.48—0.70) | 0.66 (0.48—0.90) |
| Chorioamnionitis | 0.54 (0.45—0.63) | 0.58 (0.51—0.66) | 0.83 (0.65—1.05) |
| Meconium aspiration syndrome | 0.52 (0.33—0.81) | 0.29 (0.19—0.44) | 0.46 (0.27—0.78) |
| Respiratory distress syndrome | 0.86 (0.67—1.09) | 0.79 (0.65—0.98) | 0.75 (0.49—1.16) |
| Shoulder dystocia | 1.11 (0.85—1.45) | 1.16 (0.92—1.47) | 1.77 (0.92—3.41) |
| Brachial plexus injury | 0.40 (0.13—1.27) | 0.94 (0.50—1.77) | 1.17 (0.26—5.24) |
| Parous women (n=100,642) | |||
| Outcome | Elective Induction, 39 weeks (n=9,626) vs. Expectant Management (n=54,427) | Elective Induction, 40 weeks (n=10,748) vs. Expectant Management (n=12,826) | Elective Induction, 41 weeks (n=4,948) vs. Expectant Management (n=961) |
| Cesarean delivery | 0.79 (0.73—0.86) | 0.82 (0.75—0.90) | 0.81 (0.65—1.01) |
| Operative vaginal delivery | 1.10 (0.99—1.23) | 1.30 (1.14—1.47) | 1.12 (0.77—1.62) |
| Severe maternal morbidity* | 0.81 (0.72—0.92) | 0.75 (0.66—0.85) | 1.01 (0.74—1.38) |
| Infant death | 0.48 (0.19—1.19) | 1.12 (0.55—2.26) | 0.49 (0.09—2.50) |
| NICU admission† | 0.70 (0.64—0.77) | 0.75 (0.68—0.83) | 0.78 (0.61—1.01) |
| Macrosomia (≥4500 g) | 0.52 (0.44—0.61) | 0.55 (0.48—0.63) | 0.74 (0.56—0.97) |
| Chorioamnionitis | 0.60 (0.46—0.78) | 0.55 (0.43—0.71) | 1.52 (0.79—2.95) |
| Meconium aspiration syndrome | 0.38 (0.23—0.65) | 0.42 (0.27—0.65) | 0.35 (0.17—0.73) |
| Respiratory distress syndrome | 0.99 (0.81—1.22) | 0.81 (0.64—1.02) | 0.76 (0.42—1.37) |
| Shoulder dystocia | 0.73 (0.63—0.84) | 0.75 (0.65—0.87) | 0.88 (0.62—1.24) |
| Brachial plexus injury | 0.94 (0.59—1.51) | 0.46 (0.26—0.83) | 0.33 (0.13—0.84) |
Table data are crude odds ratios (95% confidence intervals). Mode of delivery was modeled using multinomial logistic regression. Other outcomes were modeled using logistic regression.
Composite outcome including postpartum hemorrhage, third-or-fourth degree perineal laceration, unplanned surgical procedure, uterine rupture, maternal intensive care unit admission, maternal sepsis, and endometritis.
NICU stay >24 hours (or, for infants who died <24 hours after birth, any NICU admission) documented in vital records, neonatal transfer documented in vital records, or infant hospital stay longer than maternal postpartum hospital stay (documented in hospital discharge data). Number of observations with missing values on NICU admission, by parity (nulliparous, parous): 39 weeks (21; 5), 40 weeks (1; 30), 41 weeks (12; 5).
Figure 2.
Adjusted odds ratios for elective labor induction, as compared to expectant management, and pregnancy outcomes among obese women. A–K display adjusted odds ratios, stratified by parity, for elective induction of labor (compared with expectant management) and pregnancy outcomes among obese women and their offspring. Models were adjusted for maternal age, education, and race and ethnicity; first-trimester prenatal care initiation; payment source for delivery; birth year; obesity class; and delivery at a teaching hospital. Cesarean delivery (A), operative vaginal delivery* (B), severe maternal morbidity† (C), infant death‡ (D), neonatal intensive care unit (NICU) admission§ (E), macrosomia (≥4,500 g) (F), chorioamnionitis (G), meconium aspiration syndrome (H), respiratory distress syndrome (I), shoulder dystocia (J), and brachial plexus injury‖ (K). *Mode of delivery was a three category outcome modeled using multinomial logistic regression. †Includes postpartum hemorrhage, third-or-fourth degree perineal lacerations, unplanned surgical procedure, uterine rupture, maternal intensive care unit admission, sepsis, and endometritis. ‡Models for infant death did not converge at 41 weeks of gestation. §Number of observations with missing values on NICU admission, by parity (nulliparous; parous): 39 weeks of gestation (21; 5), 40 weeks of gestation (1; 30), 41 weeks of gestation (12; 5). ‖Models for brachial plexus injury did not converge among nulliparous patients at 41 weeks of gestation.
Table 4.
Adjusted Odds Ratios for Elective Induction of Labor Compared With Expectant Management and Pregnancy Outcomes Among Obese Women
| Nulliparous women (n=65,333) | |||
| Outcome | Elective Induction, 39 weeks (n=3,942) vs. Expectant Management (n=40,667) | Elective Induction, 40 weeks (n=7,061) vs. Expectant Management (n=12,453) | Elective Induction, 41 weeks (n=4,961) vs. Expectant Management (n=967) |
| Cesarean delivery | 0.82 (0.77—0.88) | 0.85 (0.80—0.90) | 0.87 (0.75—1.00) |
| Operative vaginal delivery | 1.16 (1.03—1.31) | 1.07 (0.95—1.20) | 1.14 (0.85—1.53) |
| Severe maternal morbidity* | 0.75 (0.65—0.87) | 0.84 (0.75—0.94) | 0.89 (0.69—1.13) |
| Infant death | 0.23 (0.03—1.68) | 0.75 (0.29—1.95) | Did not converge |
| NICU admission† | 0.79 (0.70—0.89) | 0.84 (0.76—0.94) | 0.60 (0.49—0.74) |
| Macrosomia (≥4500 g) | 0.28 (0.19—0.41) | 0.60 (0.50—0.73) | 0.63 (0.46—0.86) |
| Chorioamnionitis | 0.56 (0.47—0.66) | 0.61 (0.54—0.69) | 0.77 (0.61—0.98) |
| Meconium aspiration syndrome | 0.55 (0.35—0.87) | 0.31 (0.20—0.46) | 0.49 (0.28—0.84) |
| Respiratory distress syndrome | 0.89 (0.70—1.13) | 0.82 (0.67—1.01) | 0.79 (0.51—1.23) |
| Shoulder dystocia | 1.14 (0.87—1.48) | 1.22 (0.96—1.56) | 1.75 (0.91—3.40) |
| Brachial plexus injury | 0.41 (0.13—1.29) | 0.90 (0.48—1.70) | Did not converge |
| Parous women (n=100,642) | |||
| Outcome | Elective Induction, 39 weeks (n=9,626) vs. Expectant Management (n=54,427) | Elective Induction, 40 weeks (n=10,748) vs. Expectant Management (n=12,826) | Elective Induction, 41 weeks (n=4,948) vs. Expectant Management (n=961) |
| Cesarean delivery | 0.79 (0.73—0.86) | 0.81 (0.74—0.89) | 0.81 (0.64—1.01) |
| Operative vaginal delivery | 1.07 (0.97—1.20) | 1.25 (1.10—1.41) | 1.13 (0.77—1.64) |
| Severe maternal morbidity* | 0.83 (0.74—0.94) | 0.75 (0.66—0.85) | 0.95 (0.69—1.30) |
| Infant death | 0.45 (0.18—1.13) | 1.10 (0.54—2.23) | Did not converge |
| NICU admission† | 0.75 (0.68—0.82) | 0.77 (0.70—0.86) | 0.79 (0.61—1.03) |
| Macrosomia (≥4500 g) | 0.50 (0.43—0.59) | 0.55 (0.48—0.63) | 0.74 (0.56—0.99) |
| Chorioamnionitis | 0.60 (0.46—0.79) | 0.57 (0.44—0.73) | 1.35 (0.69—2.64) |
| Meconium aspiration syndrome | 0.41 (0.24—0.69) | 0.45 (0.28—0.70) | 0.41 (0.19—0.87) |
| Respiratory distress syndrome | 1.03 (0.84—1.26) | 0.81 (0.64—1.03) | 0.84 (0.46—1.54) |
| Shoulder dystocia | 0.72 (0.62—0.83) | 0.77 (0.66—0.89) | 0.87 (0.62—1.24) |
| Brachial plexus injury | 0.96 (0.59—1.54) | 0.47 (0.26—0.83) | 0.34 (0.13—0.89) |
Table data are adjusted odds ratios (95% confidence intervals). These data are also presented graphically in Figure 2. Mode of delivery was modeled using multivariable, multinomial logistic regression. Other outcomes were modeled using multivariable logistic regression. Models were adjusted for maternal age, maternal education, maternal race and ethnicity, initiation of prenatal care in the first trimester, principal source of payment for delivery, birth year, obesity class, and delivery at a teaching hospital.
Composite outcome including postpartum hemorrhage, third-or-fourth degree perineal laceration, unplanned surgical procedure, uterine rupture, maternal intensive care unit admission, maternal sepsis, and endometritis.
NICU stay >24 hours (or, for infants who died <24 hours after birth, any NICU admission) documented in vital records, neonatal transfer documented in vital records, or infant hospital stay longer than maternal postpartum hospital stay (documented in hospital discharge data). Number of observations with missing value on NICU admission, by parity (nulliparous, parous): 39 weeks (21; 5), 40 weeks (1; 30), 41 weeks (12; 5).
Elective induction of labor was not associated with infant death in adjusted models; however, models at 41 weeks did not converge (Figure 2d). Elective induction of labor at 39 and 40 weeks was associated with reduced odds of NICU admission (e.g., adjusted ORs [95% CIs] at 39 weeks were 0.79 [0.70, 0.89] among obese nulliparous women and 0.75 [0.68, 0.82] among obese parous women; Figure 2e).The adjusted odds of other neonatal complications were generally lower in offspring of electively induced versus expectantly managed obese women (Figures 2f through 2k), with some exceptions (e.g., there was no association between elective labor induction and RDS).
After stratifying by obesity class, precision decreased, and some findings were no longer statistically significant (Appendices 2–4, available online at http://links.lww.com/AOG/B41). In addition, various models for rare outcomes did not converge. Most stratified associations were in the same direction as in the main analyses. Point estimates were fairly similar by obesity class, and 95% CIs overlapped by obesity category for nearly all outcomes. Two new associations emerged in stratified analyses: elective induction at 40 weeks was associated with reduced odds of RDS among class 2 obese nulliparous women, while elective induction at 41 weeks was associated with increased odds of severe maternal morbidity among class 2 obese parous women (Appendix 3, available online at http://links.lww.com/AOG/B41). After altering our assumptions about the timing of intrapartum complications, associations between elective labor induction and cesarean appeared stronger (Appendix 5, available online at http://links.lww.com/AOG/B41). Most other maternal and infant outcomes were relatively unaffected.
DISCUSSION
In this study, elective induction of labor at 39 and 40 weeks of gestation was associated with reduced odds of cesarean delivery, severe maternal morbidity, and neonatal morbidity, with no change in the odds of infant mortality. Elective induction at 41 weeks was also associated with reduced odds of some neonatal complications. With the exception of operative vaginal delivery at 39 weeks (nulliparous women) and 40 weeks (parous women), elective induction was not associated with increased pregnancy complications. There were no significant differences by obesity class.
This is an extension of the analysis conducted by Lee et al., who examined obese women in the 2007 California Linked dataset (7). Similarly to our analysis, Lee et al. found that elective labor induction from 39–41 weeks was associated with reduced odds of cesarean delivery, macrosomia, and chorioamnionitis among obese women (7). With five years’ data (2007–2011), we were able to newly document several significant associations between elective labor induction between 39 and 41 weeks of gestation and reduced odds of other major complications (e.g., NICU admission, meconium aspiration syndrome, and others).
Similarly to our study, Schuster et al. reported that a clinical protocol to induce obese women by their estimated due date reduced the rate of cesarean delivery, as compared to rates of cesarean delivery before the protocol was initiated (24). Unlike our study, Schuster et al.’s clinical protocol was associated with a slight increase in NICU admission, although these findings were not specific to obese women (24). This clinical protocol was tested in a single healthcare system, and analysis was limited to proxy indicators of neonatal morbidity (e.g., NICU admission) (24). In a hospital-based retrospective cohort study, Wolfe et al. found that elective labor induction at 39 or 40 weeks, as compared to expectant management ≥39 weeks, was associated with increased risk of cesarean delivery and NICU admission among obese nulliparous patients with an unfavorable cervix (16). Although Wolfe et al. accessed medical records, they did not include parous women, adjust for covariates, or stratify analyses by gestational week at induction (16). Our method may more closely represent “real-time” obstetric-decision making.
Our study has many strengths. We included a large sample of over 165,000 deliveries. This allowed us to examine rare outcomes, such as brachial plexus injury. Our analysis also produced more precise estimates than prior investigations. We used expectant management as the comparison group (rather than spontaneous labor), which is a valid clinical alternative to labor induction (12). Another strength is our comparison of elective labor induction to expectant management for each week between 39–41 weeks. In contrast to earlier studies, we assessed elective labor induction at 41 weeks of gestation and stratified by obesity class. Our dataset is diverse and population-based with high rates of record linkage (19). Finally, we tested the robustness of our assumptions in sensitivity analyses. Limitations in this study included inability to evaluate stillbirth, as all stillbirths were excluded due to preexisting maternal conditions or missing data on study eligibility criteria. As stillbirth at subsequent gestational ages is prevented with induction at an earlier gestational age, this may have impacted our perinatal mortality estimates. Due to low numbers of events, we could not evaluate neonatal mortality separately from total infant mortality. Our analyses for cesarean delivery could be biased down and away from the null, as the risk of cesarean increases with gestational age (25). There may be residual confounding in our study, as our dataset did not contain information on maternal discomfort, provider preferences, or cervical status (15,26). Women with a favorable cervix or those who are healthy may be more likely to be electively induced and less likely to deliver via cesarean section (8,26). This could bias our results in favor of elective induction. Our results may only be generalizable to obese Californian women without preexisting disease who delivered between 2007 and 2011. Due to changing outcome frequencies over time, the patterns and associations in our study (years 2007–2011) may not represent those in 2012 or later. Finally, although medical complications may be underreported in administrative data, linked datasets are accurate for many complications and procedures (20–22).
Additional research using larger sample sizes of morbidly obese women may help determine whether a uniform policy on term elective induction is appropriate for all obese women. In addition, future studies should consider utilizing a randomized, controlled trial design to reduce unobserved confounding. Future analyses of stillbirth and neonatal mortality are also essential. In conclusion, elective labor induction between 39 0/7 and 40 6/7 weeks of gestation may be associated with reduced maternal and neonatal morbidity among obese women and their offspring.
Supplementary Material
Acknowledgments
Supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health (Grant 5T32HD052460-10, Emory University), Maternal and Child Health Bureau, Health Resources and Services Administration (Grant T03MC07651, Emory University), and Emory University Laney Graduate School. The funders had no role in study design; in data collection, analysis, or interpretation; in the writing of the manuscript; or in the decision to submit the article for publication.
Footnotes
Presented at the 29th Annual Meeting of the Society for Pediatric and Perinatal Epidemiologic Research, Miami, FL, June 20–21, 2016.
Financial Disclosure
The authors did not report any potential conflicts of interest.
Each author has indicated that he or she has met the journal’s requirements for authorship.
References
- 1.Aune D, Saugstad OD, Henriksen T, Tonstad S. Maternal body mass index and the risk of fetal death, stillbirth, and infant death: a systematic review and meta-analysis. JAMA. 2014;311(15):1536–46. doi: 10.1001/jama.2014.2269. [DOI] [PubMed] [Google Scholar]
- 2.Sebire NJ, Jolly M, Harris JP, et al. Maternal obesity and pregnancy outcome: a study of 287,213 pregnancies in London. Int J Obes Relat Metab Disord. 2001;25(8):1175–82. doi: 10.1038/sj.ijo.0801670. [DOI] [PubMed] [Google Scholar]
- 3.Kim S, Zhu Y, Grantz K, et al. Obstetric and Neonatal Risks Among Obese Women Without Chronic Disease. Obstet Gynecol. 2016;128(1):104–12. doi: 10.1097/AOG.0000000000001465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Branum A, Kirmeyer S, Gregory E. Prepregnancy Body Mass Index by Maternal Characteristics and State: Data From the Birth Certicate, 2014. Natl Vital Stat Rep. 2016;65(6):1–11. [PubMed] [Google Scholar]
- 5.Cedergren MI. Maternal morbid obesity and the risk of adverse pregnancy outcome. Obstet Gynecol. 2004;103(2):219–24. doi: 10.1097/01.AOG.0000107291.46159.00. [DOI] [PubMed] [Google Scholar]
- 6.Ecker JLGJ, Norwitz ER, Nadel AS, Repke JT. Birth weight as a predictor of brachial plexus injury. Obstet Gynecol. 1997;89(5 Part 1):643–7. doi: 10.1016/s0029-7844(97)00007-0. [DOI] [PubMed] [Google Scholar]
- 7.Lee VR, Darney BG, Snowden JM, et al. Term elective induction of labour and perinatal outcomes in obese women: retrospective cohort study. BJOG. 2016;123(2):271–278. doi: 10.1111/1471-0528.13807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Darney BG, Snowden JM, Cheng YW, et al. Elective induction of labor at term compared with expectant management: maternal and neonatal outcomes. Obstet Gynecol. 2013;122(4):761–9. doi: 10.1097/AOG.0b013e3182a6a4d0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhang X, Kramer MS. Variations in mortality and morbidity by gestational age among infants born at term. J Pediatr. 2009;154(3):358–62. doi: 10.1016/j.jpeds.2008.09.013. [DOI] [PubMed] [Google Scholar]
- 10.Arrowsmith S, Wray S, Quenby S. Maternal obesity and labour complications following induction of labour in prolonged pregnancy. BJOG. 2011;118(5):578–88. doi: 10.1111/j.1471-0528.2010.02889.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Gulmezoglu AM, Crowther CA, Middleton P, Heatley E. Induction of labour for improving birth outcomes for women at or beyond term. Cochrane Database Syst Rev. 2012;6:CD004945. doi: 10.1002/14651858.CD004945.pub3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Caughey AB, Sundaram V, Kaimal AJ, et al. Systematic review: elective induction of labor versus expectant management of pregnancy. Ann Intern Med. 2009;151(4):252–63. W53–63. doi: 10.7326/0003-4819-151-4-200908180-00007. [DOI] [PubMed] [Google Scholar]
- 13.Sanchez-Ramos L, Olivier F, Delke I, Kaunitz AM. Labor induction versus expectant management for postterm pregnancies: a systematic review with meta-analysis. Obstet Gynecol. 2003;101(6):1312–8. doi: 10.1016/s0029-7844(03)00342-9. [DOI] [PubMed] [Google Scholar]
- 14.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(2):128.e1–7. doi: 10.1016/j.ajog.2011.03.051. [DOI] [PubMed] [Google Scholar]
- 15.Induction of labor. ACOG Practice Bulletin No. 107: American College of Obstetricians and Gynecologists. Obstet Gynecol. 2009;114:386–97. doi: 10.1097/AOG.0b013e3181b48ef5. [DOI] [PubMed] [Google Scholar]
- 16.Wolfe H, Timofeev J, Tefera E, Desale S, Driggers RW. Risk of cesarean in obese nulliparous women with unfavorable cervix: elective induction vs expectant management at term. Am J Obstet Gynecol. 2014;211(1):53.e1–5. doi: 10.1016/j.ajog.2014.01.034. [DOI] [PubMed] [Google Scholar]
- 17.Kamath BD, Todd JK, Glazner JE, Lezotte D, AM L. Neonatal Outcomes After Elective Cesarean Delivery. Obstet Gynecol. 2009;113(6):1231–8. doi: 10.1097/AOG.0b013e3181a66d57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Petrou S, Khan K. An overview of the health economic implications of elective caesarean section. Appl Health Econ Health Policy. 2013;11(6):561–76. doi: 10.1007/s40258-013-0063-8. Review. [DOI] [PubMed] [Google Scholar]
- 19.Herrchen B, Gould JB, Nesbitt TS. Vital statistics linked birth/infant death and hospital discharge record linkage for epidemiological studies. Comp Biomed Res. 1997;30(4):290–305. doi: 10.1006/cbmr.1997.1448. [DOI] [PubMed] [Google Scholar]
- 20.Lydon-Rochelle MT, Holt VL, Nelson JC, et al. Accuracy of reporting maternal in-hospital diagnoses and intrapartum procedures in Washington State linked birth records. Paediatr Perinat Epidemiol. 2005;19(6):460–471. doi: 10.1111/j.1365-3016.2005.00682.x. [DOI] [PubMed] [Google Scholar]
- 21.Lydon-Rochelle MT, Holt VL, Cardenas V, et al. The reporting of pre-existing maternal medical conditions and complications of pregnancy on birth certificates and in hospital discharge data. Am J Obstet Gynecol. 2005;193(1):125–34. doi: 10.1016/j.ajog.2005.02.096. [DOI] [PubMed] [Google Scholar]
- 22.Yasmeen S, Romano PS, Schembri ME, Keyzer JM, Gilbert WM. Accuracy of obstetric diagnoses and procedures in hospital discharge data. Am J Obstet Gynecol. 2006;194(4):992–1001. doi: 10.1016/j.ajog.2005.08.058. [DOI] [PubMed] [Google Scholar]
- 23.The Joint Commission. Specifications manual for Joint Commission National Quality Measures (v2013A1), Table 11.07: conditions possibly justifying elective delivery prior to 39 weeks gestation. 2012 Available at: https://manual.jointcommission.org/releases/TJC2013A/AppendixATJC.html#Table_Number_11_07_Conditions_Po. Retrieved October 15, 2015.
- 24.Schuster M, Madueke-Laveaux O, Mackeen D, Feng W, Paglia M. The effect of the MFM obesity protocol on cesarean delivery rates. Am J Obstet Gynecol. 2016;215:492.e1–6. doi: 10.1016/j.ajog.2016.05.005. [DOI] [PubMed] [Google Scholar]
- 25.Glantz JC. Elective induction of labor at term compared with expectant management: maternal and neonatal outcomes. Obstet Gynecol. 2014;123(2 Pt 1):363. doi: 10.1097/AOG.0000000000000114. [DOI] [PubMed] [Google Scholar]
- 26.Pandis GK, Papageorghiou AT, Ramanathan VG, Thompson MO, Nicolaides KH. Preinduction sonographic measurement of cervical length in the prediction of successful induction of labor. Ultrasound Obstet Gynecol. 2001;18(6):623–628. doi: 10.1046/j.0960-7692.2001.00580.x. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.