Outcomes of Term Induction in Trial of Labor After.19
Outcomes of Term Induction in Trial of Labor After.19
Outcomes of Term Induction in Trial of Labor After.19
OBJECTIVE: To evaluate outcomes of induction of labor, Secondary outcomes included composite maternal mor-
compared with expectant management, in women at- bidity (hysterectomy, transfusion, intensive care unit
tempting trial of labor after cesarean delivery (TOLAC) in (ICU) transfer, venous thromboembolism, death), com-
a large obstetric cohort. posite neonatal morbidity (5-minute Apgar score less
METHODS: We performed a secondary analysis of data than 5, cord pH less than 7.0, asphyxia, hypoxic ischemic
from the Consortium on Safe Labor that included encephalopathy, neonatal death), and neonatal ICU
women with term (37 weeks of gestation or greater) admission. Multivariate logistic regression was per-
singleton gestations and a history of one prior cesarean formed with adjustment for confounding factors.
delivery who attempted TOLAC. Induction of labor was RESULTS: We identified 6,033 women attempting TOLAC
compared with expectant management by week of of whom 1,626 (27.0%) underwent induction of labor and
gestation from 37 to 40 weeks in both high- and low- 4,407 (73.0%) did not. Compared with expectant manage-
risk cohorts. The primary outcome was failed TOLAC. ment, induction was associated with an increased risk of
failed TOLAC at 37–39 weeks of gestation but not at 40
From the Division of Maternal Fetal Medicine, MetroHealth Medical Center, weeks of gestation (37 weeks of gestation, 48.5% com-
Case Western Reserve University School of Medicine, and the Division of pared with 34.3%, adjusted odds ratio [OR] 1.53, 95%
Maternal Fetal Medicine, Department of Obstetrics and Gynecology, University confidence interval [CI] 1.02–2.28]; 38 weeks of gestation,
Hospitals Case Medical Center, Cleveland, Ohio.
47.0% compared with 33.0%, adjusted OR 1.74, 95% CI
Presented at the Society for Maternal-Fetal Medicine’s 35th Annual Meeting, 1.29–2.34; 39 weeks of gestation, 45.6% compared with
February 2–7, 2015, San Diego, California.
29.8%, adjusted OR 2.16, 95% CI 1.76–2.67; 40 weeks of
The data included in this paper were obtained from the Consortium on Safe Labor,
gestation, 37.9% compared with 29.4%, adjusted OR 1.21,
which was supported by the Intramural Research Program of the Eunice Kennedy
Shriver National Institute of Child Health and Human Development, National 95% CI 0.90–1.66). Induction was associated with an
Institutes of Health, through Contract No. HHSN267200603425C. Institutions increased risk of composite maternal morbidity at 39
involved in the Consortium include, in alphabetical order: Baystate Medical Center, weeks of gestation (adjusted OR 1.87, 95% CI 1.22–2.87)
Springfield, MA; Cedars-Sinai Medical Center Burnes Allen Research Center, Los
Angeles, CA; Christiana Care Health System, Newark, DE; Georgetown Univer- and neonatal ICU admission at 37 weeks of gestation
sity Hospital, MedStar Health, Washington, DC; Indiana University Clarian (adjusted OR 2.51, 95% CI 1.62–3.90). Induction was not
Health, Indianapolis, IN; Intermountain Healthcare and the University of Utah, associated with an increased risk of neonatal morbidity.
Salt Lake City, UT; Maimonides Medical Center, Brooklyn, NY; MetroHealth
Medical Center, Cleveland, OH; Summa Health System, Akron City Hospital, CONCLUSION: Induction of labor in women with one
Akron, OH; The EMMES Corporation, Rockville, MD (Data Coordinating Cen- prior cesarean delivery, compared with expectant man-
ter); University of Illinois at Chicago, Chicago, IL; University of Miami, Miami, agement, is associated with an increased risk of failed
FL; and the University of Texas Health Science Center at Houston, Houston, TX.
TOLAC. Apart from small increases in maternal morbidity
The named authors alone are responsible for the views expressed in this manuscript,
at 39 weeks and neonatal ICU admission at 37 weeks of
which does not necessarily represent the decisions or the stated policy of the Eunice
Kennedy Shriver National Institute of Child Health and Human Development. gestation, induction is not associated with an increased
Corresponding author: Justin R. Lappen, MD, Fellow, Division of Maternal
risk of severe maternal or neonatal morbidity.
Fetal Medicine, Department of Obstetrics and Gynecology, MetroHealth Medical (Obstet Gynecol 2015;126:115–23)
Center, Case Western Reserve University School of Medicine, 2500 MetroHealth DOI: 10.1097/AOG.0000000000000922
Drive, Suite G240, Cleveland, OH; e-mail: jlappen@metrohealth.org.
LEVEL OF EVIDENCE: II
Financial Disclosure
T
The authors did not report any potential conflicts of interest.
© 2015 by The American College of Obstetricians and Gynecologists. Published
he rate of vaginal birth after cesarean delivery in
by Wolters Kluwer Health, Inc. All rights reserved. the United States, after peaking in 1996, has rap-
ISSN: 0029-7844/15 idly declined and was 9% in 2011. In contrast, the rate
Fig. 1. Flow diagram of cohort selection and comparison groups for elective induction compared with expectant man-
agement by week of pregnancy. TOLAC, trial of labor after cesarean delivery.
Lappen. Outcomes of Induction of Labor in TOLAC. Obstet Gynecol 2015.
VOL. 126, NO. 1, JULY 2015 Lappen et al Outcomes of Induction of Labor in TOLAC 117
Table 1. Demographic and Obstetric expectant management at each week of gestation
Characteristics of Women Attempting between 37 and 40 completed weeks. The induction
Trial of Labor After Cesarean Delivery groups were defined using the aforementioned criteria
for the primary and secondary “low-risk” cohorts,
Total Cohort respectively. For both cohorts, the expectant manage-
Characteristic (n56,033)
ment group was defined to include all women deliv-
Age (y) 29.665.6 ering at gestational ages at or greater than that of
BMI (kg/m2) 31.766.3 women being induced.23
Birth weight (g) 3,3966452 The primary outcome measure was failed TOLAC.
Race
White 2,983 (49.4)
Failed TOLAC was not a predefined variable in the data
African American 1,324 (22.0) set and was defined as any women attempting TOLAC
Latina 1,086 (18.0) who delivered by repeat cesarean delivery. Secondary
Other 640 (10.6) outcome measures included composite maternal mor-
Insurance status bidity, composite neonatal morbidity, and neonatal
Private 3,751 (62.2)
Public 1,862 (30.8)
intensive care unit (NICU) admission. Composite
Other 420 (7.0) maternal morbidity was defined as: hysterectomy, trans-
Hospital type fusion, intensive care unit admission, venous thrombo-
University teaching 2,264 (37.5) embolism (deep venous thrombosis and pulmonary
Community teaching 3,304 (54.8) embolism), or death. Composite neonatal morbidity
Community nonteaching 465 (7.7)
History of vaginal delivery 3,061 (49.3)
was defined as: 5-minute Apgar score less than 5,
Chronic hypertension 108 (1.8) arterial cord pH less than 7.0, neonatal death, or the
Pregnancy-related hypertensive 218 (3.6) clinical diagnoses of asphyxia or hypoxic–ischemic
disorders* encephalopathy. Given the association of uterine rup-
Diabetes† 352 (5.8) ture with severe maternal and neonatal morbidity, uter-
Cardiovascular disease 26 (0.4)
HIV 14 (0.2)
ine rupture was also assessed as a separate secondary
Patients excluded in secondary 500 (8.2) outcome. Uterine rupture was defined as a disruption of
cohort‡ the uterine muscle and visceral peritoneum and was
Nonmedically indicated induction of 1,007 (62.0) a predefined variable distinct from asymptomatic uter-
labor ine scar dehiscence, which was coded separately.
BMI, body mass index; HIV, human immunodeficiency virus. Multivariate logistic regression was performed
Data are mean6standard deviation or n (%).
* Includes gestational hypertension, preeclampsia, and eclampsia.
with adjustment for potential confounding factors.
†
Includes gestational and pregestational diabetes. Maternal outcomes were adjusted for age, body mass
‡
Includes chronic hypertension, diabetes, cardiovascular disease, index (calculated as weight (kg)/[height (m)]2), and
and HIV.
history of prior vaginal birth, whereas neonatal out-
comes were adjusted for history of prior vaginal deliv-
ery and neonatal sex. Given that Bishop score was not
a risk of expectant management, these women were known contemporaneously for expectantly managed
retained in the cohort and analyzed as part of the women, it could not be included in the multivariate
expectant management group. To generate the non- model. Lastly, maternal race, a factor that has been
medically indicated induction group, we identified pa- demonstrated to affect the success of TOLAC,24 was
tients using previously defined methodology from dropped from the final regression models because it
publications from the Consortium on Safe Labor.19,22 did not significantly affect the adjusted odds ratios.
Briefly, the database contained predefined variables Categorical variables were assessed using x2 and
for labor induction by indication, which could be Fisher’s exact test, whereas continuous variables were
broadly categorized as “indicated,” “elective,” and assessed using unpaired t test and Wilcoxon rank-sum
“no recorded indication.” In our study, we defined test where appropriate. Results are presented as odds
nonmedically indicated induction to include all those ratios (ORs) or adjusted ORs with 95% confidence
categorized as “elective” and “no recorded indication” interval (CI) and P values. All statistical analysis was
given that the results of prior analyses suggest that performed using STATA 13.1. Institutional review
patients coded as “no recorded indication” from this board approval by MetroHealth Medical Center and
data set likely represent elective inductions.22 University Hospitals Case Medical Center was ob-
Both study cohorts were then analyzed according tained before initiation of this study (institutional
to whether a patient underwent induction of labor or review board project # EM-14-21).
Table 2. Multivariate Analysis of Maternal and Neonatal Outcomes of Labor Induction Compared With
Expectant Management for Women Attempting Trial of Labor After Cesarean Delivery
Week Gestation and Outcome Induction Expectant Management Adjusted OR (95% CI)
VOL. 126, NO. 1, JULY 2015 Lappen et al Outcomes of Induction of Labor in TOLAC 119
labor was not associated with a statistically significant longer associated with an increased risk of NICU
increased risk of composite neonatal morbidity. admission at 37 weeks of gestation. Conversely,
The results of the secondary analysis of the nonmedically indicated induction was associated
“low-risk” cohort are presented in Table 3. After with a decrease in the risk of NICU admission at
adjustment for confounding factors, nonmedically 39 weeks of gestation (adjusted OR 0.52, 95% CI
indicated induction of labor in the “low-risk” 0.31–0.88). No association between nonmedically
cohort was associated with an increased risk of indicated induction and adverse neonatal outcome
failed TOLAC at 38 and 39 weeks of gestation was present.
but not at 37 and 40 weeks of gestation (37 weeks Overall, 19 women (0.3%) experienced a uterine
of gestation542.1% compared with 33.8%, adjusted rupture in this cohort, of whom four had an induction
OR 1.53, 95% CI 0.85–2.77; 38 weeks of ges- of labor. Induction of labor was not associated with
tation555.6% compared with 32.4%, adjusted OR a statistically significant increase in the risk of uterine
2.67, 95% CI 1.81–3.93; 39 weeks of ges- rupture when assessed either in comparison with
tation548.0% compared with 29.4%, adjusted OR expectant management by week of gestation or by
2.34, 95% CI 1.86–2.95; and 40 weeks of ges- exposure to induction for the entire cohort (Table 4).
tation537.5% compared with 29.3%, adjusted OR
1.27, 95% CI 0.91–1.76). Nonmedically indicated DISCUSSION
induction was associated with an increased risk of In the current analysis, when comparing induction of
composite maternal morbidity at 39 weeks of ges- labor with expectant management in a cohort of
tation (adjusted OR 2.14, 95% CI 1.36–3.35). women with one prior cesarean delivery, we found
Again, this association was attributable to a signifi- that induction increased the risk of failed TOLAC.
cantly increased risk of transfusion among women This association was present both for the entire study
failing their TOLAC attempt. In regard to second- cohort and for a secondary “low-risk” cohort of non-
ary neonatal outcome measures in the “low-risk” medically indicated inductions. As demonstrated by
cohort, nonmedically indicated induction was no recent observational data, the magnitude and
Table 3. Multivariate Analysis of Maternal and Neonatal Outcomes of Nonmedically Indicated Induction of
Labor Compared With Expectant Management for Women Attempting Trial of Labor After
Cesarean Delivery in the “Low-Risk” Cohort
Week Gestation and Outcome Induction Expectant Management Adjusted OR (95% CI)
direction of the association between induction of labor with expectant management over a gestational age
and cesarean delivery depend on the definition of the range that includes the early-term period.18,19 Addi-
comparison group.16,21 In women without a history of tionally, our findings are consistent with other stud-
cesarean delivery, induction of labor has been associ- ies on induction in women attempting TOLAC,
ated with a decreased risk of cesarean delivery when which have not demonstrated an increased risk of
compared with expectant management.17–19 For neonatal morbidity.9 However, despite the large
women attempting TOLAC, however, studies analyz- sample size of our cohort, the infrequency of adverse
ing outcomes of labor induction have only used spon- neonatal outcomes reflects a lack of power to discern
taneous labor as the comparison group, which may a difference in these rare events. Our study also
bias against induction. The use of an expectant man- detected an increased risk of NICU admission at
agement comparison group represents one of the pri- 37 weeks of gestation, which was no longer present
mary contributions of our study to the literature. when analyzing nonmedically indicated inductions
Despite this difference in methodology, our analysis of labor in the “low-risk” cohort. This finding sug-
found a similar association between induction and gests that the NICU admissions were potentially
failed TOLAC as demonstrated by prior investiga- related to the underlying medical indication for the
tions that used a spontaneous labor comparison induction, a lack of statistical power to detect a differ-
group.2,7,25 ence, or an increased risk of respiratory morbidity
We observed an isolated increase in the risk of for early term births at 37 weeks of gestation.26 Addi-
composite maternal morbidity with induction at 39 tionally, this increase in the risk of NICU admission
weeks of gestation, which was attributable to an may be secondary to precautionary measures and
increased risk of transfusion among women failing may not represent a true increased risk of neonatal
their TOLAC attempt. The fact that this finding was morbidity because there was no standard protocol
isolated to 39 weeks of gestation likely represents the for NICU admission.
fact that the largest number of inductions occurred at There were multiple strengths to this analysis,
this gestational age and therefore our study was including the use of a large, reliable data set generated
underpowered to detect a difference at other weeks from a multicenter U.S. cohort of laboring women
of gestation. The finding that induction for women representative of the current obstetric population, and
attempting TOLAC is associated with an increased practice patterns. Compared with studies that use
risk of transfusion is consistent with the findings of administrative data, the use of medical record data
Grobman et al. However, unlike the study by improves the accuracy of variables of particular
Grobman,9 our study did not demonstrate an importance to studies of induction compared with
increase in other markers of severe maternal mor- expectant management such as gestational age, the
bidity or uterine rupture. However, given the rarity classification of labor type or onset, indication for
of severe maternal morbidity and uterine rupture, delivery, maternal or fetal medical conditions, and
our study had limited power to detect a difference maternal and neonatal outcome data. The validation
between groups. Information on the indications for of key variables directly from medical
and volume of transfusion was not available. records minimizes, but does not eliminate, misclassi-
In regard to neonatal outcomes, we did not detect fication and ascertainment bias. The definition of
a difference in composite neonatal morbidity with “trial of labor” in the original Safe Labor cohort
induction of labor, which is consistent with other was a patient with two or more documented
publications assessing outcomes of induction compared intrapartum cervical examinations. This definition
VOL. 126, NO. 1, JULY 2015 Lappen et al Outcomes of Induction of Labor in TOLAC 121
may overestimate the determination of who was hav- expectant management, is associated with an
ing a trial of labor and thus artificially increased the increased risk of failed TOLAC. Despite the change
rate of failed TOLAC in the expectant management in methodology to use expectant management as the
group (eg, patient in latent labor with two examina- comparison group, our findings support previous
tions before a planned repeat cesarean delivery literature that used spontaneous delivery in the same
counted as failed TOLAC). This misclassification week. The results of studies comparing induction with
would bias our results toward the null. Thus, any find- expectant management among women without a his-
ing of a significant difference between failed TOLAC tory of cesarean delivery should not be generalized to
in induction and expectant management groups may women attempting TOLAC. Our study findings
be an underestimate of the magnitude of the associa- should not alter current obstetric practice, which
tion. Lastly, in addition to assessing pertinent out- supports an individualized approach to the use of
comes such as failed TOLAC and uterine rupture, induction in appropriate candidates attempting
we assessed other maternal and neonatal outcome TOLAC.
measures that reflect severe morbidity.
Our study has several limitations, including the
REFERENCES
retrospective nature of the data. Additionally, we were
1. Martin JA, Hamilton BE, Ventrura SJ, Osterman MJ, Matthews TJ.
unable to control for all obstetric factors that may Births: final data for 2011. Natl Vital Stat Rep 2013;62:1–69, 72.
affect the success of induction, such as Bishop score, 2. Landon MB, Hauth JC, Leveno KJ, Spong CY, Leindecker S,
because contemporaneous data were not available for Varner MW, et al. Maternal and perinatal outcomes associated
the expectant management group. Other factors with a trial of labor after prior caesarean delivery. N Engl J Med
2004;351:2581–9.
related to the success of TOLAC such as the
3. Macones GA, Peipert J, Nelson DB, Odibo A, Stevens EJ,
indication for the previous cesarean delivery or Stamilio DM, et al. Maternal complications with vaginal birth
history of successful vaginal birth after cesarean after cesarean delivery: a multicenter study. Am J Obstet Gy-
delivery were not included in regression models necol 2005;193:1656–62.
because data on these predictors were unavailable. 4. Ravasia DJ, Wood SL, Pollard JK. Uterine rupture during
induced trial of labor among women with a previous cesarean
Factors inherent to the original Safe Labor cohort
delivery. Am J Obstet Gynecol 2000;183:1176–9.
deserve mention. The original publication from the
5. Zelop CM, Shipp TD, Repke JT, Cohen A, Caugey A,
Consortium on Safe Labor reported that intrapartum Lieberman E. Uterine rupture during induced or augmented
cesarean deliveries occurred at less advanced cervical labor in women with one prior cesarean delivery. Am J Obstet
dilation for induced compared with spontaneous Gynecol 1999;181:882–6.
labors and also reported an overall rate of TOLAC 6. Lydon-Rochelle M, Holt VL, Easterling TR, Martin DP. Risk of
uterine rupture during labor among women with a prior cesar-
success of 57%, which is slightly lower than previously ean delivery. N Engl J Med 2001;345:3–8.
published data.4,5,23 These findings may reflect
7. Vaginal birth after previous cesarean delivery. Practice Bulletin
a change in commitment to TOLAC by patients and No. 115. American College of Obstetricians and Gynecologists.
physicians, a change in obstetric practice after previ- Obstet Gynecol 2010;116:450–63.
ously published studies regarding the risks of TOLAC, 8. Rossi AC, Prefumo F. Pregnancy outcomes of induced labor in
or the current medicolegal climate surrounding women with previous cesarean section: a systematic review and
meta-analysis. Arch Gynecol Obstet 2015;291:273–80.
TOLAC. Overall, each of these factors may bias
9. Grobman WA, Gilbert S, Landon MB, Spong CY, Leveno KJ,
our results against labor induction. Rouse DJ, et al. Outcomes of induction of labor after one prior
Lastly, since the collection of this data set, the cesarean. Obstet Gynecol 2007;109:262–9.
American College of Obstetricians and Gynecologists 10. Yeast JD, Jones A, Poskin M. Induction of labor and the rela-
and the Society for Maternal-Fetal Medicine have tionship to cesarean delivery: a review of 7001 consecutive
inductions. Am J Obstet Gynecol 1999;180:628–33.
recommended the avoidance of nonmedically indicated
inductions before 39 weeks of gestation.27 As a result, 11. Smith KM, Hoffman MK, Sciscione A. Elective induction of
labor in nulliparous women increases the risk of cesarean deliv-
data regarding nonmedically indicated induction before ery. Obstet Gynecol 2003;101:45S.
39 weeks of gestation may not be considered clinically 12. Hoffman MK, Vahratian A, Sciscione AC, Troendle JF, Zhang J.
appropriate in current obstetric practice. However, by Comparison of labor progression between induced and nonin-
including both medically and nonmedically indicated duced multiparous women. Obstet Gynecol 2006;107:1029–34.
inductions in this study, the results remain generalizable 13. Vahratian A, Zhang J, Troendle JF, Sciscione AC, Hoffman MK.
Labor progression and risk of cesarean delivery in electively
and clinically pertinent because medically indicated
induced nulliparas. Obstet Gynecol 2005;105:698–704.
early-term inductions remain common.
14. Luthy DA, Malmgren JA, Zingheim RW. Cesarean delivery
In conclusion, induction of labor for women with after elective induction in nulliparous women: the physician
one prior cesarean delivery, in comparison with effect. Am J Obstet Gynecol 2004;191:1511–5.
VOL. 126, NO. 1, JULY 2015 Lappen et al Outcomes of Induction of Labor in TOLAC 123
Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.
Alternative Proxies: