Abstract
Objective:
To compare health care medical resource utilization in low-risk nulliparous pregnancies according to BMI categories.
Methods:
This is a secondary analysis of a multi-center randomized controlled trial of induction of labor at 390/7-394/7 weeks compared with expectant management in low-risk nulliparous pregnant people, defined as those without standard obstetric indications for delivery at 39 weeks. BMI at randomization was categorized into 4 groups (<25, 25–29, 30–39, and ≥40 kg/m2). The primary outcome of this analysis was time spent in labor and delivery (L&D) from admission to delivery. Secondary outcomes included length of stay post-delivery, total hospital length of stay, and antepartum, intrapartum, and postpartum resource utilization, which were defined a priori. Multivariable generalized linear modeling and logistic regressions were performed, and 99% confidence intervals were calculated.
Results:
A total of 6,058 pregnant people were included in the analysis with 640 (10.6%) with BMI <25 kg/m2, 2,222 (36.7%) with BMI 25–29 kg/m2, 2,577 (42.5%) with BMI 30–39 kg/m2, and 619 (10.2%) with BMI ≥40 kg/m2. Time spent in labor and delivery increased from 15.1±9.2 hours for people with BMI < 25 kg/m2 to 23.5±13.6 hours for people with BMI ≥40 kg/m2, and every 5-unit increase in BMI was associated with an average 9.8% increase in time spent in labor and delivery (adjusted estimate per 5-unit increase in BMI 1.10, 99% CI 1.08, 1.11). Increasing BMI was not associated with an increase in antepartum resource utilization except for blood tests and urinalysis. However, increasing BMI was associated with higher odds of intrapartum resource utilization, longer total hospital length of stay, and postpartum resource utilization. For example, every 5-unit increase in BMI was associated with an increase of 26.1% in the odds of antibiotic administration, 57.6% in placement of intrauterine pressure catheter, 5.1% in total inpatient length of stay, 31.0 in postpartum emergency department visit, and 23.9% in postpartum hospital admission.
Conclusions:
Among low-risk nulliparous people, higher BMI was associated with longer time from admission to delivery, total hospital length of stay, and more frequent utilization of intrapartum and postpartum resources.
Clinical Trial Registration:
Précis:
Among nulliparous low-risk people, higher body mass index was associated with longer time from admission to delivery, total hospital length of stay, and more frequent utilization of intrapartum and postpartum health care resources.
Introduction
Obesity, defined as body mass index (BMI) ≥30 kg/m2, is a public health problem in the United States and prevalent among pregnant people.1–3 Between 2016 and 2019, the rate of pre-pregnancy obesity rose from 26.1% to 29%, and this rise was pervasive across groups with varied demographic characteristics.1 Compared with normal-weight pregnant people, those with obesity have an increased risk of pregnancy complications including preeclampsia and other hypertensive disorders, pre-gestational and gestational diabetes mellitus, cesarean delivery, stillbirth, large for gestational age birth, and infant death.3,4 Prior studies have shown an increase in total direct healthcare costs, with more hospitalizations, higher rates of prescription drug use, and more outpatient visits in pregnant women with obesity. However, many of these studies relied on administrative data or their outcomes were not prospectively collected.5–8 Moreover, none of these studies were specifically in low-risk nulliparous people, and it remains uncertain as to whether increasing BMI independently affects resource utilization.3
Following the publication of the ARRIVE trial9 in 2018, the American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine (SMFM) stated that it is “reasonable for obstetricians and health-care facilities to offer elective induction of labor to low-risk nulliparous people at 39 weeks’ gestation.”2 While induction of labor was associated with longer durations in labor and delivery, it led to significantly fewer antepartum visits, tests, and treatments, and shorter maternal and neonatal hospital durations post-delivery.10 For health care systems that are considering or have adopted a policy of routinely offering labor induction, and in view of the epidemic of obesity in the U.S., it is important to better understand the relationship between obesity and resource utilization. Therefore, the purpose of this analysis was to compare health care resource utilization in the antepartum, intrapartum, and postpartum periods among low-risk nulliparous people according to their BMI categories.
Methods
This was a secondary analysis of the ARRIVE trial, a multicenter open-label randomized trial of elective induction of labor at 390/7 to 394/7 weeks versus expectant management in low-risk nulliparous people.9 Details of the trial have been previously reported. Briefly, singleton, nulliparous people were screened at 34–38 completed weeks of gestation at 41 U.S. hospitals participating in the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) Maternal–Fetal Medicine Units Network. Consenting people who met eligibility criteria were randomized at 380/7 to 386/7 weeks to either planned induction of labor at 390/7 weeks to 394/7 weeks or to expectant management until at least 405/7 weeks but no later than 422/7 weeks. There were no other trial-specific requirements for antepartum or intrapartum care. Antepartum and intrapartum management of patients in the parent trial was according to standard of care at clinical centers. Inclusion criteria for the trial included a well-dated pregnancy with a live, singleton fetus in cephalic presentation, with no contraindication to vaginal delivery and no plan for cesarean delivery or labor induction at the time of randomization. People were excluded if, at randomization, there was any maternal or fetal indication for delivery prior to 40 5/7 weeks (e.g., hypertension, diabetes mellitus, fetal growth restriction, placental abruption).9 For this analysis, we additionally excluded people with missing BMI or maternal length of stay in labor and delivery. Trained and certified research staff abstracted information from medical records, including demographic information, medical, obstetrical, and social history, and outcome data. Institutional review board approval was available at each of the participating sites and written informed consent was obtained from participants before randomization. This analysis was deemed exempt from further review by the Office of Responsible Research Practices at the Ohio State University.
The primary outcome for this analysis is total maternal inpatient time spent in labor and delivery (L&D). This was calculated during the inpatient admission for delivery as the difference between the time of admission to the time of delivery (hours), and was chosen because time in L&D is a major determinant of patient flow and unit staffing, as well as a surrogate for cost. Secondary outcomes included antepartum, intrapartum, and postpartum utilization of resources as defined a priori in the parent trial, with details previously reported.9,10 After randomization, study participants were followed up – both by review of medical records as well as by interviews performed by research personnel – to collect data on antepartum resources, which included whether, prior to their admission for delivery, they had any outpatient visits or inpatient admissions. The outpatient visits were also classified according to whether they were routine visits for prenatal care or unscheduled visits that were not for routine prenatal care (i.e., ambulatory clinic visits for routine and non-routine prenatal care, respectively). Unscheduled outpatient visits were further categorized according to their location of occurrence (i.e., ambulatory clinic, urgent care center, or hospital-based setting, such as an emergency department or obstetrical triage). Data on evaluations or interventions (e.g., imaging studies, laboratory tests, medications given, antenatal testing, etc) during outpatient visits or inpatient admission were collected. During the admission for delivery, interventions that were utilized for the participant were abstracted from the medical record. These included interventions such as cervical ripening agents, oxytocin infusions, use of fetal scalp electrodes or intrauterine pressure catheters, antibiotics, uterotonics, and magnesium sulfate seizure prophylaxis. Women were interviewed between 4 and 8 weeks after hospital discharge and asked whether they had attended any additional outpatient visits or inpatient admissions. This interview was augmented for all women by medical record abstraction of documented health care through 8 weeks after discharge.
Resource utilization after randomization was categorized a priori according to when resources were used: antepartum, intrapartum during the delivery admission, or postpartum. Our use of the terms “race” and “ethnicity” recognizes these terms as social constructs and does not presuppose a biologic construct.11,12 Race and ethnicity were self-reported by patients. All patients who identified as Hispanic ethnicity, regardless of race, were categorized as Hispanic. Non-Hispanic patients were categorized as Black, White, or None of the above. The “none of the above” race category, which includes non-Hispanic Asian, Native Hawaiian or Pacific Islander, American Indian/Alaskan Native, unknown, or more than one race was collapsed to make comparisons between groups with low frequency. BMI is the weight in kilograms divided by the square of the height in meters, and the weight and height information were obtained from medical records. BMI at randomization was analyzed as a categorical variable (normal < 25 kg/m2, overweight 25–29 kg/m2, obese 30–39 kg/m2, and morbidly obese ≥40 kg/m2) for descriptive purposes. Maternal characteristics were compared using the Cochran-Armitage trend test for binary variables, score test from multinomial logistic regression for multinomial variables, or Jonckheere-Terpstra trend test for continuous variables to assess trends across BMI categories. For each type of resource that was significantly associated with BMI, differences in the frequency of use per 1,000 participants of obese and normal BMI were calculated.
Generalized linear models assuming a gamma distribution and log link function were used for continuous outcomes and logistic regression models were used for categorical outcomes. Models for antepartum resources were adjusted for maternal age, parent-trial randomization arm, and hospital as a random effect. Models for intrapartum resources, length of stay variables, and postpartum resources were adjusted for maternal age, parent-trial randomization arm, hospital as a random effect, and modified Bishop score at randomization (< 5 or ≥5). For the postdelivery outcomes, we considered mode of delivery as a mediator so it was not included as a covariate in the multivariable model. We additionally evaluated whether there were any interactions between primary trial assignment groups (induction vs. expectant management) and BMI for the primary and secondary outcomes.
Due to multiple comparisons, statistical significance was defined as p<0.01 and 99% confidence intervals were presented. Statistical analyses were performed using SAS statistical software (SAS Institute, Inc, Cary, NC) version 9.4. All tests were two tailed and no imputation for missing values was performed.
Results
Of 6,106 people who were randomized, 6,058 (99.2%) had data available and were eligible for inclusion in this analysis. Of those, 640 (10.6%) were of normal BMI (23.5±1.2 kg/m2), 2,222 (36.7%) of overweight BMI (27.6±1.4 kg/m2), 2,577 (42.5%) of obese BMI (33.8±2.7 kg/m2), and 619 (10.2%) of morbidly obese BMI (44.7±4.5 kg/m2) at randomization.
Baseline and demographic characteristics of study participants are presented in Table 1. Tests of trend were not significant for differences in gestational age at randomization or in primary trial group assignments across BMI groups. However, people with obese and morbidly obese BMI were less likely to be married or have private insurance, and were more likely to self-identify as non-Hispanic Black or Hispanic, smoke during pregnancy, have an unfavorable cervix (Bishop score < 5) at randomization, and undergo cesarean delivery compared with participants with normal BMI.
Table 1:
Characteristics of the study participants
| BMI <25 kg/m2 (n=640) |
BMI 25–29 kg/m2 (n=2,222) |
BMI 30–39 kg/m2 (n=2,577) |
BMI ≥40 kg/m2 (n=619) |
P for trend | |
|---|---|---|---|---|---|
| Maternal age, years | 24.2±5.2 | 24.8±5.2 | 24.6±5.0 | 24.4±4.9 | 1.00 |
| AMA (≥35 years) | 19 (3.0) | 100 (4.5) | 108 (4.2) | 22 (3.6) | 0.84 |
| Ethnicity and Race Hispanic Non-Hispanic Black Non-Hispanic White None of the above |
133 (20.8) 148 (23.1) 293 (45.8) 66 (10.3) |
603 (27.1) 412 (18.5) 1077 (48.5) 130 (5.9) |
759 (29.5) 620 (24.1) 1088 (42.2) 110 (4.3) |
166 (26.8) 217 (35.1) 206 (33.3) 30 (4.8) |
<0.001 |
| Married | 385 (60.2) | 1406 (63.3) | 1485 (57.6) | 308 (49.8) | <0.001 |
| Private insurance | 309 (48.3) | 1111 (50.0) | 1086 (42.1) | 210 (33.9) | <0.001 |
| Smoked during pregnancy | 42 (6.6) | 130 (5.9) | 228 (8.8) | 60 (9.7) | <0.001 |
| Modified Bishop score at randomization <5 | 350 (54.7) | 1350 (60.8) | 1692 (65.7) | 445 (71.9) | <0.001 |
| Assigned to IOL group | 327 (51.1) | 1087 (48.9) | 1326 (51.5) | 303 (48.9) | 0.75 |
| Cesarean delivery | 79 (12.3) | 314 (14.1) | 618 (24.0) | 225 (36.3) | <0.001 |
| Admitting practitioner specialty MFM OB/GYN (general/other subspecialty) or Family Practice Midwife |
73 (11.4) 537 (83.9) 30 (4.7) |
248 (11.2) 1849 (83.2) 125 (5.6) |
345 (13.4) 2057 (79.8) 175 (6.8) |
119 (19.2) 463 (74.8) 37 (6.0) |
<0.001 |
Data presented as mean ± standard deviation or n (%), unless otherwise specified.
AMA, advanced maternal age; BMI, body mass index; IOL, induction of labor.
P-value based on Cochran-Armitage trend test for binary variables, score test from multinomial logistic regression for multinomial variables, or Jonckheere-Terpstra trend test for continuous variables.
Number of missing values: insurance (n=1), modified Bishop score (n=2).
The primary outcome, time spent in L&D from admission until delivery, increased from 15.1±9.2 hours for participants with BMI < 25 kg/m2 to 23.5±13.6 hours for participants with BMI ≥40 kg/m2, and every 5-unit increase in BMI was associated with an average 9.8% increase in time spent in L&D (adjusted estimated per 5-unit increase in BMI 1.10, 99% Confidence Interval 1.08, 1.11). (Table 2) Moreover, during the delivery admission, BMI was associated with longer total hospital length of stay from admission to discharge, as well as a higher odds of using of intrapartum resources, including cervical ripening, use of oxytocin, placement of intrauterine pressure catheter and fetal scalp electrode, and administration of antibiotics. For example, every 5-unit increase in BMI was associated with a 26.1% odds increase in the administration of antibiotics, 57.6% increase in placement of intrauterine pressure catheter, and an average 5.1% increase in total inpatient LOS from admission to discharge. However, there were no differences in use of magnesium sulfate for seizure prophylaxis and uterotonics other than oxytocin. (Table 2)
Table 2:
Delivery admission use of resources, by body mass index
| BMI <25 kg/m2 (n=640) |
BMI 25–29 kg/m2 (n=2,222) |
BMI 30–39 kg/m2 (n=2,577) |
BMI ≥40 kg/m2 (n=619) |
Unadjusted estimate per 5-unit increase in BMI (99% CI) | Adjusted estimate per 5-unit increase in BMI (99% CI) | |
|---|---|---|---|---|---|---|
| Number of hours from admission to delivery | 15.1±9.2 | 16.8± 10.5 | 20.3±11.7 | 23.5±13.6 | 1.11 (1.09, 1.13) | 1.10 (1.08, 1.11) |
| Number of days from admission to discharge | 2.7±0.8 | 2.9±0.9 | 3.1±1.0 | 3.4±1.1 | 1.06 (1.05, 1.07) | 1.05 (1.04, 1.06) |
| Cervical ripening | 243 (38.0) | 902 (40.6) | 1278 (49.6) | 355 (57.4) | 1.21 (1.15, 1.28) | 1.20 (1.13, 1.28) |
| Oxytocin infusion | 442 (69.1) | 1673 (75.3) | 2128 (82.6) | 541 (87.4) | 1.33 (1.23, 1.44) | 1.32 (1.22, 1.43) |
| Intrauterine pressure catheter | 178 (27.8) | 670 (30.2) | 1143 (44.4) | 390 (63.0) | 1.54 (1.45, 1.63) | 1.58 (1.48, 1.68) |
| Fetal scalp electrode | 129 (20.2) | 417 (18.8) | 724 (28.1) | 288 (46.5) | 1.45 (1.36, 1.54) | 1.50 (1.41, 1.61) |
| Magnesium sulfate infusion | 13 (2.0) | 43 (1.9) | 73 (2.8) | 18 (2.9) | 1.19 (1.02, 1.38) | 1.14 (0.97, 1.34) |
| Epidural analgesia | 460 (71.9) | 1635 (73.6) | 1962 (76.1) | 475 (76.7) | 1.06 (0.99, 1.13) | 0.99 (0.91, 1.08) |
| Antibiotic infusion | 219 (34.2) | 885 (39.8) | 1219 (47.3) | 371 (59.9) | 1.28 (1.21, 1.35) | 1.26 (1.19, 1.33) |
| Uterotonic medication | 54 (8.4) | 217 (9.8) | 289 (11.2) | 68 (11.0) | 1.05 (0.96, 1.14) | 1.07 (0.98, 1.16) |
| Red blood cell transfusion | 12 (1.9) | 28 (1.3) | 51 (2.0) | 12 (1.9) | 1.08 (0.89, 1.31) | 1.06 (0.87, 1.29) |
| Uterine balloon for tamponade | 6 (0.94) | 10 (0.45) | 11 (0.43) | 4 (0.65) | 0.93 (0.63, 1.38) | ** |
| LOS post-delivery > 2 days | 80 (12.5) | 325 (14.6) | 569 (22.1) | 202 (32.6) | 1.35 (1.27, 1.44) | 1.35 (1.27, 1.45) |
Data reported as mean±standard deviation or n (%), unless otherwise specified.
CI, confidence interval; LOS, length of stay.
Continuous outcomes based on a generalized linear model with log link and gamma distribution. Categorical outcomes based on a logistic regression model.
Model adjusted for maternal age, primary trial randomization arm, modified Bishop score at randomization (< 5 or ≥5), and hospital as a random effect.
Sparse data.
Bold indicates p<0.01
In the antepartum period after randomization, BMI was not associated with an increase in scheduled and unscheduled visits, emergency room/triage visits, hospital admissions prior to delivery, sonograms, or antenatal surveillance tests. However, BMI was associated with increased laboratory blood and urine testing (such as complete blood counts or metabolic assessments, and urinalysis) and treatment with analgesics. For instance, 6.1% and 10.8% of people with BMI 30–39 kg/m2 and ≥40 kg/m2 had a blood test compared with 4.8% of participants with BMI < 25 kg/m2, and every 5-unit increase in BMI was associated with an average 43.5% increase in analgesic treatment (adjusted odds per 5-unit increase in BMI 1.44, 99% Confidence Interval 1.17, 1.76). (Tables 3 and 4)
Table 3:
Proportion (%) of women with at least one post-randomization antepartum encounter with the health care system, by body mass index
| BMI <25 kg/m2 (n=640) |
BMI 25–29 kg/m2 (n=2,222) |
BMI 30–39 kg/m2 (n=2,577) |
BMI ≥40 kg/m2 (n=619) |
Unadjusted estimate per 5-unit increase in BMI (99% CI) | Adjusted estimate per 5-unit increase in BMI (99% CI) | |
|---|---|---|---|---|---|---|
| Ambulatory | ||||||
| Office visit, routine scheduled prenatal care | 304 (47.5) | 1115 (50.2) | 1301 (50.5) | 324 (52.3) | 1.04 (0.98, 1.10) | 1.03 (0.97, 1.10) |
| Office visit, unscheduled | 8 (1.3) | 34 (1.5) | 42 (1.6) | 10 (1.6) | 1.07 (0.87, 1.31) | ** |
| Emergency Room/Urgent care, obstetric triage visit | 198 (30.9) | 651 (29.3) | 775 (30.1) | 204 (33.0) | 1.01 (0.96, 1.07) | 1.01 (0.95, 1.08) |
| Inpatient | ||||||
| Hospital admission before delivery | 8 (1.3) | 40 (1.8) | 34 (1.3) | 9 (1.5) | 0.99 (0.79, 1.24) | ** |
Data reported as n (%), unless otherwise specified.
CI, confidence interval.
Models adjusted for maternal age, primary trial randomization arm, and hospital as a random effect.
Sparse data.
Table 4:
Proportion (%) of women with at least one antepartum test or intervention, stratified by body mass index
| BMI <25 kg/m2 (n=640) |
BMI 25–29 kg/m2 (n=2,222) |
BMI 30–39 kg/m2 (n=2,577) |
BMI ≥40 kg/m2 (n=619) |
Unadjusted estimate per 5-unit increase in BMI (99% CI) | Adjusted estimate per 5-unit increase in BMI (99% CI) | |
|---|---|---|---|---|---|---|
| Fetal surveillance and imaging | ||||||
| Any Antenatal surveillance test1 | 133 (20.8) | 507 (22.8) | 576 (22.4) | 178 (28.8) | 1.07 (1.01, 1.14) | 1.07 (1.00, 1.15) |
| Any antenatal sonogram2 | 63 (9.8) | 256 (11.5) | 304 (11.8) | 104 (16.8) | 1.13 (1.05, 1.22) | 1.09 (0.99, 1.19) |
| Contraction stress test3 | 2 (0.31) | 4 (0.18) | 8 (0.31) | 3 (0.48) | 1.22 (0.79, 1.89) | ** |
| Laboratory testing | ||||||
| Any blood test | 31 (4.8) | 114 (5.1) | 158 (6.1) | 67 (10.8) | 1.23 (1.12, 1.36) | 1.23 (1.10, 1.38) |
| Other† | 11 (1.7) | 64 (2.9) | 57 (2.2) | 12 (1.9) | 0.94 (0.78, 1.13) | 0.89 (0.70, 1.12) |
| Urinalysis | 24 (3.8) | 66 (3.0) | 110 (4.3) | 36 (5.8) | 1.20 (1.06, 1.36) | 1.16 (1.01, 1.32) |
| Treatment | ||||||
| Analgesic‡ | 4 (0.63) | 26 (1.2) | 39 (1.5) | 15 (2.4) | 1.31 (1.08, 1.58) | 1.44 (1.17, 1.76) |
| Intravenous hydration | 9 (1.4) | 15 (0.68) | 35 (1.4) | 10 (1.6) | 1.17 (0.93, 1.46) | 1.18 (0.93, 1.49) |
| Antibiotic | 2 (0.31) | 7 (0.32) | 15 (0.58) | 5 (0.81) | 1.19 (0.85, 1.68) | ** |
| Other medication§ | 9 (1.4) | 30 (1.4) | 47 (1.8) | 13 (2.1) | 1.17 (0.97, 1.42) | 1.17 (0.96, 1.43) |
Data reported as n (%), unless otherwise specified.
CI, confidence interval.
Models adjusted for maternal age, primary trial randomization arm, and hospital as a random effect.
A complete blood count and chemistry panel focused upon preeclampsia (e.g., creatinine, liver function tests).
Includes blood tests (e.g., such as amylase and lipase) and cultures from urine, blood, and skin.
Includes medications such as acetaminophen, butalbital, and opioids.
Includes medications such as antiemetics, antihistamines, and antivirals.
Sparse data.
Bold indicates p<0.01
Antenatal surveillance tests include non-stress test, modified biophysical profile, or Biophysical profile.
Antenatal sonogram include those performed for Doppler, fetal position, amniotic fluid index, fetal growth.
Any blood test include includes CBC, metabolic panel, preeclampsia panel.
Moreover, BMI was associated with an increase in postpartum emergency department visits and hospital readmissions, such that every 5-unit increase in BMI was associated with an average 31.0% increase in the risk of emergency department visit (adjusted odds per 5-unit increase in BMI 1.31, 99% Confidence Interval 1.19, 1.44) and 23.9% increase in postpartum hospital admission (adjusted odds per 5-unit increase in BMI 1.24, 99% Confidence Interval 1.06, 1.45). (Table 5)
Table 5:
Proportion (%) of women with at least one post-randomization postpartum encounter with the health care system, by body mass index
| BMI <25 kg/m2 (n=640) |
BMI 25–29 kg/m2 (n=2,222) |
BMI 30–39 kg/m2 (n=2,577) |
BMI ≥40 kg/m2 (n=619) |
Unadjusted estimate per 5-unit increase in BMI (99% CI) | Adjusted estimate per 5-unit increase in BMI (99% CI) | |
|---|---|---|---|---|---|---|
| Ambulatory | ||||||
| Office visit, unanticipated | 31 (4.8) | 135 (6.1) | 144 (5.6) | 53 (8.6) | 1.09 (0.98, 1.21) | 1.10 (0.99, 1.23) |
| Urgent care visit | 12 (1.9) | 24 (1.1) | 31 (1.2) | 12 (1.9) | 1.04 (0.83, 1.31) | 1.08 (0.85, 1.37) |
| Emergency department visit | 24 (3.8) | 124 (5.6) | 185 (7.2) | 78 (12.6) | 1.32 (1.21, 1.45) | 1.31 (1.19, 1.44) |
| Inpatient | ||||||
| Hospital admission | 14 (2.2) | 38 (1.7) | 64 (2.5) | 26 (4.2) | 1.23 (1.06, 1.44) | 1.24 (1.06, 1.45) |
Data reported as n (%), unless otherwise specified.
CI, confidence interval.
Models adjusted for maternal age, primary trial randomization arm, modified Bishop score at randomization (< 5 or ≥5), and hospital as a random effect.
Bold indicates p<0.01
There were no statistically significant interactions between primary trial assignment groups (induction vs. expectant management) and BMI for any association (P >0.01). Sensitivity analyses for postdelivery resources according to mode of delivery are included in Tables 6 – 9, and did not show major differences from the overall analysis. The association between BMI and antibiotic use during the delivery admission was only significant among those who did not require cesarean delivery. (Tables 6 – 7).
Table 6:
Delivery admission use of resources, by body mass index among participants with cesarean delivery
| BMI <25 kg/m2 (n=79) |
BMI 25–29 kg/m2 (n=314) |
BMI 30–39 kg/m2 (n=618) |
BMI ≥40 kg/m2 (n=225) |
Unadjusted estimate per 5-unit increase in BMI (99% CI) | Adjusted estimate per 5-unit increase in BMI (99% CI) | |
|---|---|---|---|---|---|---|
| Number of hours from admission to delivery | 19.6±12.2 | 21.7±13.6 | 25.5±13.5 | 27.5±15.5 | 1.07 (1.03, 1.11) | 1.06 (1.03, 1.09) |
| Number of days from admission to discharge | 4.0±1.0 | 4.2±1.1 | 4.3±1.0 | 4.3±1.1 | 1.01 (1.00, 1.03) | 1.01 (1.00, 1.03) |
| Antibiotic infusion | 61 (77.2) | 260 (82.8) | 545 (88.2) | 200 (88.9) | 1.16 (0.98, 1.37) | 1.14 (0.96, 1.36) |
| Uterotonic medication | 14 (17.7) | 45 (14.3) | 99 (16.0) | 24 (10.7) | 0.88 (0.75, 1.03) | 0.90 (0.76, 1.06) |
| Red blood cell transfusion | 8 (10.1) | 15 (4.8) | 30 (4.9) | 4 (1.8) | 0.72 (0.53, 0.98) | ** |
| Uterine balloon for tamponade | 2 (2.5) | 3 (1.0) | 4 (0.7) | 0 | 0.43 (0.17, 1.12) | ** |
| LOS post-delivery > 2 days | 68 (86.1) | 262 (83.4) | 512 (82.9) | 189 (84.0) | 0.98 (0.85, 1.13) | 1.07 (0.88, 1.30) |
Data reported as mean±standard deviation or n (%), unless otherwise specified.
CI, confidence interval; LOS, length of stay.
Continuous outcomes based on a generalized linear model with log link and gamma distribution. Categorical outcomes based on a logistic regression model.
Model adjusted for maternal age, primary trial randomization arm, modified Bishop score at randomization (< 5 or ≥5), and hospital as a random effect.
Sparse data.
Bold indicates p<0.01
Table 9:
Proportion (%) of women with at least one postpartum encounter with the health care system, by body mass index among participants without cesarean delivery
| BMI <25 kg/m2 (n=561) |
BMI 25–29 kg/m2 (n=1908) |
BMI 30–39 kg/m2 (n=1959) |
BMI ≥40 kg/m2 (n=394) |
Unadjusted estimate per 5-unit increase in BMI (99% CI) | Adjusted estimate per 5-unit increase in BMI (99% CI) | |
|---|---|---|---|---|---|---|
| Ambulatory | ||||||
| Office visit, unanticipated | 30 (5.4) | 116 (6.1) | 101 (5.2) | 22 (5.6) | 0.97 (0.84, 1.12) | 0.99 (0.85, 1.14) |
| Urgent care visit | 10 (1.8) | 20 (1.1) | 28 (1.4) | 6 (1.5) | 1.02 (0.78, 1.34) | ** |
| Emergency department visit | 21 (3.7) | 101 (5.3) | 115 (5.9) | 35 (8.9) | 1.21 (1.07, 1.37) | 1.20 (1.06, 1.36) |
| Inpatient | ||||||
| Hospital admission | 10 (1.8) | 32 (1.7) | 37 (1.9) | 12 (3.1) | 1.14 (0.92, 1.41) | 1.15 (0.93, 1.42) |
Data reported as n (%), unless otherwise specified.
CI, confidence interval.
Models adjusted for maternal age, primary trial randomization arm, modified Bishop score at randomization (< 5 or ≥5), and hospital as a random effect.
Sparse data
Bold indicates p<0.01
Table 7:
Delivery admission use of resources, by body mass index among participants without cesarean delivery
| BMI <25 kg/m2 (n=561) |
BMI 25–29 kg/m2 (n=1,908) |
BMI 30–39 kg/m2 (n=1,959) |
BMI ≥40 kg/m2 (n=394) |
Unadjusted estimate per 5-unit increase in BMI (99% CI) | Adjusted estimate per 5-unit increase in BMI (99% CI) | |
|---|---|---|---|---|---|---|
| Number of hours from admission to delivery | 14.5±8.5 | 16.0± 9.7 | 18.6±10.5 | 21.3±11.9 | 1.10 (1.08, 1.12) | 1.09 (1.07, 1.10) |
| Number of days from admission to discharge | 2.6±0.5 | 2.6±0.7 | 2.8±0.6 | 2.9±0.7 | 1.03 (1.02, 1.04) | 1.03 (1.02, 1.03) |
| Antibiotic infusion | 158 (28.2) | 625 (32.8) | 674 (34.4) | 171 (43.4) | 1.14 (1.07, 1.22) | 1.13 (1.06, 1.21) |
| Uterotonic medication | 40 (7.1) | 172 (9.0) | 190 (9.7) | 44 (11.2) | 1.08 (0.97, 1.20) | 1.10 (0.99, 1.23) |
| Red blood cell transfusion | 4 (0.7) | 13 (0.7) | 21 (1.1) | 8 (2.0) | 1.29 (0.98, 1.69) | ** |
| Uterine balloon for tamponade | 4 (0.7) | 7 (0.4) | 7 (0.4) | 4 (1.0) | 1.14 (0.75, 1.76) | ** |
| LOS post-delivery > 2 days | 12 (2.1) | 63 (3.3) | 57 (2.9) | 13 (3.3) | 1.03 (0.86, 1.24) | 0.98 (0.81, 1.19) |
Data reported as mean±standard deviation or n (%), unless otherwise specified.
CI, confidence interval; LOS, length of stay.
Continuous outcomes based on a generalized linear model with log link and gamma distribution. Categorical outcomes based on a logistic regression model.
Model adjusted for maternal age, primary trial randomization arm, modified Bishop score at randomization (< 5 or ≥5), and hospital as a random effect.
Sparse data.
Bold indicates p<0.01
Pair-wise differences (per 1,000 participants) for resources that were significantly associated with BMI are presented in Table 10 for the obese and morbidly obese BMI groups, compared with those in the normal BMI group. In the antepartum period after randomization, people in the obese BMI group had approximately 13 more blood tests and 4 more urine tests per 1,000 people than those in the normal BMI group. During the delivery admission, people in the obese BMI group had approximately 5,110 more hours on labor and delivery and 366 more days in the hospital than people in the normal BMI group. They also received 116 more ripening agents, 135 more oxytocin infusions, 165 more intrauterine pressure catheters, and 79 more fetal scalp electrodes per 1,000 people than those in the normal BMI group. In the postpartum period, people in the obese BMI group had approximately 42 more emergency department visits. For each of the aforementioned resources, people in the morbidly obese BMI group had even greater differences compared to those in the normal BMI group (Table 10).
Table 10:
Comparison of utilization per 1000 women in types of resources that significantly differed between group
| BMI <25 kg/m2 (n=640) |
BMI 30–39 kg/m2 (n=2,577) |
Difference (99% CI) (obese vs. normal BMI) | BMI ≥40 kg/m2 (n=619) |
Difference (99% CI) (morbidly obese vs. normal BMI) | |
|---|---|---|---|---|---|
| Antepartum | |||||
| Blood tests | 41 | 54 | 13 (0, 26) | 101 | 61 (39, 82) |
| Urinalysis | 32 | 36 | 4 (−8, 16) | 44 | 11 (−5, 27) |
| Delivery admission | |||||
| Hours from admission to delivery | 15,150 | 20,259 | 5,110 (4437, 5783) | 23,506 | 8,357 (7166, 9547) |
| Days from admission to discharge | 2,743 | 3,109 | 366 (310, 422) | 3,426 | 683 (587, 779) |
| Cervical ripening | 380 | 496 | 116 (61, 172) | 574 | 194 (123, 265) |
| Oxytocin used | 691 | 826 | 135 (84, 186) | 874 | 183 (125, 242) |
| Intrauterine pressure catheter placed | 278 | 444 | 165 (113, 218) | 630 | 352 (284, 420) |
| Fetal scalp electrode placed | 202 | 281 | 79 (33, 126) | 465 | 264 (198, 330) |
| Antibiotic administration | 342 | 473 | 131 (76, 185) | 599 | 257 (187, 327) |
| Postpartum | |||||
| Emergency department visit | 41 | 83 | 42 (18, 67) | 157 | 116 (73, 159) |
| Hospital admission | 28 | 26 | −2 (−21, 17) | 45 | 17 (−10, 44) |
Discussion
In this analysis, we provide contemporary data on the association of BMI with antepartum, intrapartum, and postpartum resource utilizations in low-risk nulliparous at term. Higher BMI was associated with longer time from admission to delivery and total hospital LOS as well as more frequent utilization of intrapartum and postpartum resources.
We demonstrate that utilization of health care resources differs significantly between pregnant people with increased BMI and those with normal BMI. Specifically, the former had significantly longer time spent in L&D during the delivery admission and longer overall hospital length of stay from admission until discharge. Moreover, they had significantly higher utilization of blood and urine tests and analgesic treatments in the antepartum period after randomization, and cervical ripening, use of oxytocin, placement of intrauterine pressure catheter and fetal scalp electrode, and administration of antibiotics in the intrapartum period. Of the 2,694 patients who received antibiotics during the delivery admission, the majority received prophylactic antibiotics (49% against group B streptococcus and 21% as surgical prophylaxis prior to cesarean section) whereas 711 (26%) patients received antibiotics as treatment for suspected or confirmed chorioamnionitis. Lastly, BMI was associated with an increase in postpartum emergency department visits and hospital readmissions.
Previous studies have shown an increase in total direct healthcare costs for every unit increase in BMI among pregnant people, with more hospitalizations, higher rates of prescription drugs, and greater outpatient visits in those with obese BMI. However, many of these studies relied on administrative data, had outcomes that were not prospectively collected, and were not contemporary. To our knowledge, none were specifically in low-risk people.5,13–17 For example, analysis of electronic data records of 13,422 pregnant people from a U.S. group-practice health maintenance organization from 2000–2004 showed that higher than normal BMI was associated with prolonged length of hospital stay for delivery (4.0±0.1 days and 4.4±0.1 days for those with obese and extremely obese BMI compared with 3.6±0.1 days for those with normal BMI), with the bulk of increase related to higher rates of cesarean delivery.5 However the difference remained significant after correction for mode of delivery. In addition, a higher-than-normal BMI was associated with significantly more obstetrical ultrasonographic examinations, medications dispensed from the outpatient pharmacy, and prenatal visits with physicians.5 However, this study included high-risk pregnant people such as those with diabetes mellitus and hypertensive disorders. Also, data from approximately 38 thousand pregnant people from Queensland in 2008 showed that mothers with obese BMI had significantly longer maternal hospital stays and increased hospitalization costs compared with those of normal weight.15
The economic burden of obesity in general is substantial.14,16–20 A cross sectional analysis of 16,262 patients from the 2000 Medical Expenditure Panel Survey in the U.S. demonstrated that excess costs among morbidly obese adults resulted from greater expenditures for office-based visits, outpatient and in-patient hospital care, and prescription drugs. Moreover, overweight and obese BMI status in pregnant people was associated with increased costs of care for the newborn and child compared with those with normal BMI,14,18 and modeling estimates of the average incremental costs associated with pregnancy and birth were more than $18,290.19
We demonstrate a robust association between maternal BMI and increased utilization of health care resources in low-risk pregnancies. Further exploration of other factors that could be driving the use of resources, especially social determinants of health, is needed. Additionally, further evaluation of the association between length of stay and health-related outcomes could help explain whether the increased use of health care resources for pregnant people with obesity, but without high-risk conditions, represents significant and valuable benefit or unnecessary variation in care. In turn, variation in care may be related to health care professionals’ implicit and explicit biased attitudes toward people with obesity,21 or greater needs related to other, unknown complications associated with higher BMI. Moreover, further research is needed to identify interventions to reduce the incidence and prevalence of morbid obesity and improve the health of adults with obesity. Last, studies should focus on prospectively collecting data to quantify the total cost of health care use among mothers with increased BMI.
Strengths of our study include that it was a secondary analysis of a large, contemporary, randomized trial of singleton low-risk nulliparous people, thus minimizing bias due to confounding and misclassification of variables. This approach, along with the multicenter aspect of the study and the geographically and ethnically diverse population enhances the generalizability of our results. In addition, the study outcomes were defined a priori and assessed prospectively by trained research staff. Contrary to other studies, the primary variable of interest (BMI) was collected by trained research staff. This method is important to avoid recall bias and misreporting of data on BMI based solely on patient report, as there is evidence that adults tend to overestimate their height and underestimate their weight, which, could potentially result in an underestimation of the association under study.22 We used BMI at randomization, since it occurred between 38 and 39 weeks and it is unlikely that patients had excessive weight gain or loss leading to significant change in their BMI category between randomization and delivery. Moreover, using BMI at randomization allowed us to avoid using two different exposure groups i.e. BMI at randomization for antepartum resources and BMI at delivery for delivery admission and post-delivery resources. We also accounted for potentially important confounding factors. In our analysis of intrapartum factors, we adjusted for the modified Bishop score at randomization.
Our findings should be interpreted in the context of the following limitations. We were unable to collect actual costs of resource utilization, due to the lack of standardized costs across the different hospitals. Additionally, we did not have reliable data on, or adjust for, gestational weight gain. There are data supporting gestational weight gain as a risk factor for pregnancy complications and a modifier of the association between maternal obesity and pregnancy outcomes.23,24 However, we are not able to distinguish between whether study associations are related to the presence of obesity at the start of pregnancy or excessive gestational weight gain during pregnancy. On the other hand, despite randomization in the parent trial by induction and expectant management, the four groups in this study were categorized based on BMI. This may have led to unbalance in multiple variables, especially CD rates. However, mode of delivery was considered a mediator for postdelivery outcomes and not included as a covariate in the model. It is possible that differences in obstetric care health care professionals caring for different BMI groups may have biased the results, however prior studies showed limited evidence that interhospital outcome variation, with the exception of postpartum hemorrhage, was related to health care professionals or hospital factors.25 Most patients across the different BMI groups were cared for by OB/GYN and MFM physicians, but 6.1% of patients across all groups were cared for by a certified nurse midwife (4.7% for BMI <25 and 6% for those in BMI ≥40 kg/m2, see Table 1). We do not believe that this biased the study because care provided by obstetric health care professionals is concordant across the different BMI groups at the individual trial centers. Lastly, this study was limited to nulliparous patients with singleton gestations, and cannot be extrapolated to other people.
We provide contemporary data on the association of BMI on health care resources utilization in low-risk nulliparous people at term. Higher BMI was associated with longer time from admission to delivery and total hospital LOS as well as more frequent utilization of intrapartum and postpartum resources. These associations were similar regardless of whether individuals underwent induction or expectant management at 39 weeks. Even a small increase in health care resources, given that, at present, almost 1 of 3 pregnant people in the U.S. have an obese BMI, could lead to significant additional use of L&D resources and have potentially substantial economic implications.
Supplementary Material
Table 8:
Proportion (%) of women with at least one postpartum encounter with the health care system, by body mass index among participants with cesarean delivery
| BMI <25 kg/m2 (n=79) |
BMI 25–29 kg/m2 (n=314) |
BMI 30–39 kg/m2 (n=618) |
BMI ≥40 kg/m2 (n=225) |
Unadjusted estimate per 5-unit increase in BMI (99% CI) | Adjusted estimate per 5-unit increase in BMI (99% CI) | |
|---|---|---|---|---|---|---|
| Ambulatory | ||||||
| Office visit, unanticipated | 1 (1.3) | 19 (6.1) | 43 (7.0) | 31 (13.8) | 1.28 (1.07, 1.54) | 1.27 (1.05, 1.55) |
| Urgent care visit | 2 (2.5) | 4 (1.3) | 5 (0.5) | 6 (2.7) | 1.13 (0.72, 1.78) | ** |
| Emergency department visit | 3 (3.8) | 23 (7.3) | 70 (11.3) | 43 (19.1) | 1.38 (1.19, 1.61) | 1.37 (1.17, 1.61) |
| Inpatient | ||||||
| Hospital admission | 4 (5.1) | 6 (1.9) | 27 (4.4) | 14 (6.2) | 1.22 (0.96, 1.56) | 1.26 (0.98, 1.61) |
Data reported as n (%), unless otherwise specified.
CI, confidence interval.
Models adjusted for maternal age, primary trial randomization arm, modified Bishop score at randomization (< 5 or ≥5), and hospital as a random effect.
Sparse data.
Bold indicates p<0.01
Acknowledgements
The authors thank Gail Mallett, R.N., M.S., C.C.R.C. and Kim Hill, R.N., B.S.N. for protocol development and coordination between clinical research centers; Lindsay Doherty, M.S. for managing the data and protocol; and Elizabeth A. Thom, Ph.D. and Madeline M. Rice, Ph.D., for protocol development and oversight.
Funding Sources:
Funding: Supported by grants (HD40512, U10 HD36801, HD27869, HD34208, HD68268, HD40485, HD40500, HD53097, HD40560, HD40545, HD27915, HD40544, HD34116, HD68282, HD87192, HD68258, HD87230) from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) and the National Center for Advancing Translational Sciences (UL1TR001873). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Financial Disclosure
Geeta K. Swamy reports money was paid to them for consultant activities from GlaxoSmithKline, Pfizer, and Moderna for vaccine-related activities and UpToDate contributions. Hyagriv N. Simhan is a co-founder of Naima Health, LLC. The authors did not report any potential conflicts of interest.
Each author has confirmed compliance with the journal’s requirements for authorship.
Presentation Information: Presented at the 40th Annual Scientific Meeting of the Society for Maternal-Fetal Medicine, February 2020, Grapevine, TX.
References:
- 1.Driscoll AK, Gregory ECW. Increases in Prepregnancy Obesity: United States, 2016–2019. NCHS Data Brief. 2020(392):1–8. [PubMed] [Google Scholar]
- 2.Hales CM, Fryar CD, Carroll MD, Freedman DS, Ogden CL. Trends in Obesity and Severe Obesity Prevalence in US Youth and Adults by Sex and Age, 2007–2008 to 2015–2016. JAMA. 2018;319(16):1723–1725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.American College of O, Gynecologists’ Committee on Practice B-O. Obesity in Pregnancy: ACOG Practice Bulletin, Number 230. Obstet Gynecol. 2021;137(6):e128–e144. [DOI] [PubMed] [Google Scholar]
- 4.Gaillard R, Durmus B, Hofman A, Mackenbach JP, Steegers EA, Jaddoe VW. Risk factors and outcomes of maternal obesity and excessive weight gain during pregnancy. Obesity (Silver Spring). 2013;21(5):1046–1055. [DOI] [PubMed] [Google Scholar]
- 5.Chu SY, Bachman DJ, Callaghan WM, et al. Association between obesity during pregnancy and increased use of health care. N Engl J Med. 2008;358(14):1444–1453. [DOI] [PubMed] [Google Scholar]
- 6.Heslehurst N, Lang R, Rankin J, Wilkinson JR, Summerbell CD. Obesity in pregnancy: a study of the impact of maternal obesity on NHS maternity services. BJOG. 2007;114(3):334–342. [DOI] [PubMed] [Google Scholar]
- 7.Galtier-Dereure F, Boegner C, Bringer J. Obesity and pregnancy: complications and cost. The American journal of clinical nutrition. 2000;71(5 Suppl):1242S–1248S. [DOI] [PubMed] [Google Scholar]
- 8.Galtier-Dereure F, Montpeyroux F, Boulot P, Bringer J, Jaffiol C. Weight excess before pregnancy: complications and cost. Int J Obes Relat Metab Disord. 1995;19(7):443–448. [PubMed] [Google Scholar]
- 9.Grobman WA, Rice MM, Reddy UM, et al. Labor Induction versus Expectant Management in Low-Risk Nulliparous Women. N Engl J Med. 2018;379(6):513–523. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Grobman WA, Sandoval G, Reddy UM, et al. Health resource utilization of labor induction versus expectant management. Am J Obstet Gynecol. 2020;222(4):369 e361–369 e311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Keppel K, Bilheimer L, Gurley L. Improving population health and reducing health care disparities. Health Aff (Millwood). 2007;26(5):1281–1292. [DOI] [PubMed] [Google Scholar]
- 12.Institute of Medicine CoUaE-iRaEDiHCSB, Stith AY, Nelson AR, eds. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC: Natinal Academic Press; 2003. [PubMed] [Google Scholar]
- 13.Lindholm ES, Altman D, Norman M, Blomberg M. Health Care Consumption during Pregnancy in relation to Maternal Body Mass Index: A Swedish Population Based Observational Study. J Obes. 2015;2015:215683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kuhle S, Muir A, Woolcott CG, et al. Maternal pre-pregnancy obesity and health care utilization and costs in the offspring. Int J Obes (Lond). 2019;43(4):735–743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Watson M, Howell S, Johnston T, Callaway L, Khor SL, Cornes S. Pre-pregnancy BMI: costs associated with maternal underweight and obesity in Queensland. Aust N Z J Obstet Gynaecol. 2013;53(3):243–249. [DOI] [PubMed] [Google Scholar]
- 16.Solmi F, Morris S. Overweight and obese pre-pregnancy BMI is associated with higher hospital costs of childbirth in England. BMC Pregnancy Childbirth. 2018;18(1):253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Denison FC, Norwood P, Bhattacharya S, et al. Association between maternal body mass index during pregnancy, short-term morbidity, and increased health service costs: a population-based study. BJOG. 2014;121(1):72–81; discussion 82. [DOI] [PubMed] [Google Scholar]
- 18.Moran PS, Wuytack F, Turner M, et al. Economic burden of maternal morbidity - A systematic review of cost-of-illness studies. PLoS One. 2020;15(1):e0227377. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Lenoir-Wijnkoop I, van der Beek EM, Garssen J, Nuijten MJ, Uauy RD. Health economic modeling to assess short-term costs of maternal overweight, gestational diabetes, and related macrosomia - a pilot evaluation. Front Pharmacol. 2015;6:103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Arterburn DE, Maciejewski ML, Tsevat J. Impact of morbid obesity on medical expenditures in adults. Int J Obes (Lond). 2005;29(3):334–339. [DOI] [PubMed] [Google Scholar]
- 21.Lawrence BJ, Kerr D, Pollard CM, et al. Weight bias among health care professionals: A systematic review and meta-analysis. Obesity (Silver Spring). 2021;29(11):1802–1812. [DOI] [PubMed] [Google Scholar]
- 22.Connor Gorber S, Tremblay M, Moher D, Gorber B. A comparison of direct vs. self-report measures for assessing height, weight and body mass index: a systematic review. Obes Rev. 2007;8(4):307–326. [DOI] [PubMed] [Google Scholar]
- 23.Kiel DW, Dodson EA, Artal R, Boehmer TK, Leet TL. Gestational weight gain and pregnancy outcomes in obese women: how much is enough? Obstet Gynecol. 2007;110(4):752–758. [DOI] [PubMed] [Google Scholar]
- 24.Chung JG, Taylor RS, Thompson JM, et al. Gestational weight gain and adverse pregnancy outcomes in a nulliparous cohort. Eur J Obstet Gynecol Reprod Biol. 2013;167(2):149–153. [DOI] [PubMed] [Google Scholar]
- 25.Grobman WA, Bailit JL, Rice MM, et al. Can differences in obstetric outcomes be explained by differences in the care provided? The MFMU Network APEX study. Am J Obstet Gynecol. 2014;211(2):147 e141–147 e116. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.