OBSTETRICS

Induction of labor versus expectantmanagement for women with a priorcesarean deliveryAnna Palatnik, MD; William A. Grobman, MD, MBA

boeanu

stioks

oalCesarean Registry thatgestational age of39an delivery. Outcomes

6safvsgpearwids

1.03e1.67; and odd1.22e6.12, respectivel

CONCLUSION: Induction of labor at 39 weeks, when compared to

nt

ooC

isacone

anagement among womenrial of labor after cesarean.nvestigated whether laboris associated with an in-e risk of uterine rupture oretric morbidities.

Research ajog.orgMATERIALS AND METHODSThis was a secondary analysis of data

Corresponding author: Anna Palatnik, MD. anna.palatnik@northwestern.edu

0002-9378/free 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2015.01.026chance ofexpectant mplanning tWe also iinductioncrease in thother obst

From the Department of Obstetrics and Gynecology, Feinberg School of Medicine, NorthwesternUniversity, Chicago, IL.

Received Oct. 18, 2014; revised Dec. 17, 2014; accepted Jan. 19, 2015.

The authors report no conict of interest.

Presented in oral format at the 35th annual meeting of the Society for Maternal-Fetal Medicine,San Diego, CA, Feb. 2-7, 2015. The racing ag logo above indicates that this article was rushed topress for the benet of the scientic community.associated with an approximately 2-foldincreased risk of uterine rupture.8-11

women without a prior cesarean de-livery, when labor induction has been

390/7 weeks of gestation would notbe associated with an increased

cesarean when compared todeliveries done for the indication of prior cesarean delivery wereexcluded from the analysis.

Cite this article as: Palatnik A, Grobman WA. Induction of labor versus expecta2015;212:358.e1-6.

I t is commonly believed that womenwith a prior cesarean delivery whoundergo induction of labor are less likelyto have vaginal birth after cesarean(VBAC).1 Indeed, observational studieshave consistently shown that womenwho are induced after a prior cesareanhave a 15-20% higher chance of cesareandelivery.2-7 In addition, several studieshave shown that induction of labor is

However, these ccomparisons with wspontaneous labor.others13,14 have demcomparison groupvant, because theinduction is not spexpectant managem358.e1 American Journal of Obstetrics& Gynecology MARCH 2015Key words: induction of labor, uterine rupture, vaginal birth aftercesarean delivery

management for women with a prior cesarean delivery. Am J Obstet Gynecol

nclusions are frommen who were inaughey et al12 andonstrated how thisnot clinically rele-tual alternative totaneous labor butnt. In fact, among

compared to expectant managementinstead of spontaneous labor, meta-analysis of observational studies hasrevealed a lower chance of cesarean de-livery among those who were induced.15

The consequences of labor induc-tion compared to expectant manage-ment among women with a priorcesarean remain uncertain. We hy-pothesized that induction of laborof induction at 39, 40, and 41 weeks were compared to expectantmanagement beyond each gestational age period using univariableand multivariable analyses. Women with scheduled repeat cesarean

expectant management, was associated with a higher chance of VBACbut also of uterine rupture.opment Maternal-Fetal Medicine Units Networkincluded women with singleton gestations at aweeks and a history of 1 low transverse cesareOBJECTIVE: Previous studies of induction of laof labor after cesarean have compared womlabor after cesarean to those undergoing spontthe clinically relevant comparison is to thosemanagement. The objective of this study waoutcomes between women undergoing inducundergoing expectant management 39 weeSTUDY DESIGN: This was a secondary analysisKennedy Shriver National Institute of Child Her in the setting of trialn undergoing trial ofeous labor. However,ndergoing expectantto compare obstetricn of labor and thoseof gestation.

f data from the Euniceth and Human Devel-

RESULTS: In all, 12,67vaginal birth after cedergoing induction omanagement (73.8%also was higher amonweeks compared to exrespectively). In multivremained associateduterine rupture (odwomen were eligible for analysis. The rate ofrean (VBAC) was higher among women un-labor at 39 weeks compared to expectant61.3%, P< .001). The risk of uterine rupturewomen undergoing induction of labor at 39ctant management (1.4% vs 0.5%, P .006,iable analysis, induction of labor at 39 weeksth a significantly higher chance of VBAC and

ratio, 1.31; 95% confidence interval,s ratio, 2.73; 95% confidence interval,y).from the Cesarean Registry of the Eunice

n e

/7

63

3/7

.4 .7

.6 .0

b (

b (

(

(

(

b (

b (

b (

ajog.org Obstetrics ResearchTABLE 1Characteristics of women undergoimanagement

Characteristic

IOL

390/7e393

wk (n[

Age, y 30.2 5Prepregnancy BMI 27.0 6Race

African American 157 (24.6)

Caucasian 391 (61.3)

Hispanic 55 (8.6)b

Other 35 (5.5)b

Cigarette use during 75 (11.7)

Pregnancy

Prior vaginal delivery 362 (56.8)

Prior VBAC 254 (39.8)

Recurrent indication for prior 173 (27.1)Kennedy Shriver National Institute ofChild Health and Human DevelopmentMaternal-Fetal Medicine Units Network.That registry was the result of a 4eyearmulticenter observational study,designed to address clinical issues relatedto cesarean childbirth.8 In the presentanalysis, we included all women from theregistry who had a history of 1 cesareandelivery via a low transverse or unknownuterine incision and were at a gestationalage of at least 390/7 weeks. Women withscheduled repeat cesarean deliveriesdone for the indication of prior cesareandelivery were excluded from the analysis.Women who underwent labor induc-

tion were divided into 3 comparisongroups based on the timing of their in-duction of labor: 390/7-393/7, 400/7-403/7,and 410/7-413/7 weeks. Gestational agewas based on the best obstetric estimate(last menstrual period compared withultrasonography), determined by healthcare providers and used for clinical

CD

Chronic medical illnessa 9 (1.4)

All data presented as mean SD or N (%).BMI, body mass index; CD, cesarean delivery; EM, expectant m

a Includes chronic hypertension, pregestational diabetes, asthma

Palatnik. Induction of labor after prior cesarean delivery. Amg induction of labor with 1 prior cesar

EM IOL EM

8)>393/7 wk(n[ 7565)

400/7e403/7 wk(n[ 522)

>40(n[

b 28.1 5.7 29.7 5.4b 2726.7 6.3 27.3 6.7 27

2556 (33.8) 159 (30.4)b 1066

2116 (27.9) 263 (50.4)b 733

2473 (32.7) 69 (13.2)b 973

421 (5.6) 30 (5.7)b 160

971 (12.8) 69 (13.2) 395

3228 (42.9) 238 (46.1) 1270

2010 (26.5) 169 (32.4)b 777

2475 (39.8) 169 (32.4) 937decision-making.8 Women who under-went induction during each gestationalage windowwere compared with womenwho were managed expectantly after thesame gestational age window. Thisdesign was used to mimic the prospec-tive choice of undertaking a labor in-duction during a given period of time atthe start of a given week of gestation orundergoing expectant managementfrom that time forward. In an effort toevaluate womenwho were not in need ofimmediate delivery due to the onset ofan acute obstetric complication, womenwere excluded from the induction groupwhen they had an acute obstetric medi-cal indication for induction (ie, pre-eclampsia, gestational hypertension,nonreassuring antenatal surveillance,oligohydramnios, fetal growth restric-tion, and antenatal intrauterine fetaldemise). However, if women developedthese conditions while they were beingexpectantly managed, they were not

63 (0.8) 6 (1.1) 17 (

anagement; IOL, induction of labor; VBAC, vaginal birth after cesarea

, seizure disorder, thyroid disease, or renal insufficiency; b P< .05 for

J Obstet Gynecol 2015.

MARCH 2015 Amerian delivery compared to expectant

IOL EM

wk2933)

410/7e413/7 wk(n[ 471)

>413/7 wk(n[ 547)

5.7 28.7 5.5b 27.4 5.6 6.4 27.6 7.3 27.3 6.1

36.3) 157 (33.3)b 220 (40.2)

25.0) 188 (39.9)b 104 (19.0)

33.2) 94 (20.0)b 191 (34.9)

5.4) 32 (6.8)b 32 (5.8)

13.5) 67 (14.2) 96 (17.6)

43.6) 188 (39.9)b 255 (47.0)

26.5) 117 (24.8) 153 (28.0)

38.5) 145 (30.8) 193 (40.2)excluded from the expectant manage-ment group, as 1 consequence ofexpectant management is that theseconditions may develop and requiredelivery.A recurrent indication for cesarean

delivery was dened as a cesarean due toany type of arrest disorder. Uterinerupture was dened as a disruption ortear of the uterine muscle and visceralperitoneum or a separation of the uter-ine muscle with extension to the bladderor broad ligament.To ensure that our results were not

solely dependent upon our primary an-alytic approach and group composition,we performed additional analyses inwhich the inclusion criteria for thegroup of women expectantly managedwas altered slightly. In 1 analysis, weincluded women who underwent laborfrom the rst day at which women alsomay have undergone labor induction (ie,laboring women were included in the

0.6) 4 (0.8) 5 (0.9)

n.

comparison of labor induction vs EM at given gestational age.

can Journal of Obstetrics& Gynecology 358.e2

expectantly managed group from 390/7

weeks, instead of only from 394/7 weeksas in the primary analysis). In anotheranalysis, we included women in theexpectant management group if theyremained pregnant at least 1 week longerthan the rst day women underwent la-bor induction in the comparison groupbut then underwent a scheduled repeatcesarean with prior cesarean as the pri-mary indication, given it is possible thatthese women initially may have chosenexpectant management (instead ofplanned cesarean) but then decided toforego trial of labor after cesarean whenthey had not labored by a given gesta-tional age.All analyses were performed with

software (Stata, version 12.0; StataCorp,College Station, TX). Univariable com-parisons of maternal and neonatalcharacteristics and pregnancy outcomeswere performed using Pearson c2 testand Fisher exact test for categorical data

into the regression equation if theydiffered between groups in univariableanalysis at a level of P < .05. Odds ratioswith 95% condence intervals wereestimated from the logistic regression.This study was considered exempt by theNorthwestern University InstitutionalReview Board because only de-identieddata were used.

RESULTSIn all, 12,676 women were eligible foranalysis. Maternal characteristics of thestudy population according to thegestational age at which they underwentlabor induction or expectant manage-ment are depicted in Table 1. Womenundergoing induction of labor differedin several ways from those who wereexpectantly managed, including in theirage, race, and obstetric history.Maternal outcomes are depicted in

Table 2. Women induced at 390/7-393/7

weeks compared to those who were

not reach statistical signicance. Womeninduced at 390/7-393/7 weeks also had ahigher chance of uterine rupturecompared to women managed expec-tantly beyond that gestational age(Table 2). There were no other differ-ences in obstetric morbidity betweenwomenwho were induced and who wereexpectantly managed.Neonatal characteristics and out-

comes are depicted in Table 3. Neonatesof women who were induced at 390/7-393/7 and 400/7-403/7 weeks had lowerbirthweight compared to neonateswhose mothers were managed expec-tantly (P < .001 and P .03, respec-tively). Overall, neonatal outcomes ateach gestational age did not differsignicantly among the comparisongroups. The point estimates of the fre-quencies of neonatal intensive care unitadmission, hypoxic-ischemic encepha-lopathy, and perinatal death were higheramong neonates whose mothers were in

d e

/7

2

(6

(3

(3

(0

(1

(0

(0

(0

(0

uc

ar

Research Obstetrics ajog.organd the Student t test for continuousmeasures. Additionally, multivariablelogistic regression was performed for theoutcomes that were signicantlydifferent in univariable analysis. Poten-tial confounding variables were entered

TABLE 2Outcomes of women undergoing inmanagement

Maternal outcome

IOL

390/7e393/7

wk (n[ 638)

VBAC 471 (73.8)b

Third/fourth degree 28 (4.4)

Endometritis 13 (2.0)

Wound complication 4 (0.6)

Blood transfusion 6 (0.9)

Operative complicationsa 1 (0.1)

Uterine rupture 9 (1.4)b

Hysterectomy 0

ICU admission 1 (0.1)

Maternal death 0

All data presented as mean SD or N (%).EM, expectant management; ICU, intensive care unit; IOL, ind

a Cystotomy, ureteral injury, bowel injury; b P < .05 for compPalatnik. Induction of labor after prior cesarean delivery. Am

358.e3 American Journal of Obstetrics& Gynecolmanaged expectantly had higher rates ofVBAC. A higher chance of VBAC simi-larly was noted among womenwho wereinduced at 400/7-403/7 weeks, comparedto those expectantly managed beyondthat gestational age, but the results did

uction of labor after 1 prior cesarean d

EM IOL EM

>393/7 wk(n[ 7565 )

400/7e403/7 wk(n[ 522)

>403(n[

4640 (61.3) 340 (65.1) 1817

311 (4.1) 30 (5.7) 108

255 (3.4) 12 (2.3) 115

29 (0.4) 4 (0.7) 12

122 (1.6) 14 (2.7) 45

22 (0.3) 4 (0.7) 10

40 (0.5) 7 (1.3) 17

11 (0.1) 3 (0.6) 5

21 (0.3) 3 (0.6) 7

1 (0.01) 1 (0.2) 0

tion of labor; VBAC, vaginal birth after cesarean.

ison of labor induction vs EM at given gestational age.J Obstet Gynecol 2015.

ogy MARCH 2015the expectant management groups,however these differences did not reachstatistical signicance (Table 3).Multivariable analyses for the likeli-

hood of VBAC and uterine rupture arepresented in Table 4. Induction of labor

livery compared to expectant

IOL EM

wk933)

410/7e413/7 wk(n[ 471)

>413/7 wk(n[ 547)

1.9) 277 (58.8) 330 (60.3)

.7) 27 (5.7) 14 (2.5)

.9) 20 (4.2) 16 (2.9)

.4) 1 (0.2) 2 (0.3)

.5) 9 (1.9) 11 (2.0)

.3) 0 1 (0.2)

.6) 6 (1.3) 2 (0.4)

.2) 0 3 (0.5)

.2) 0 1 (0.2)

0 0

TABLE 3Neonatal outcomes of women undergoing induction of labor after 1 prior ceexpectant management

Variable

IOL EM IOL EM

390/7e393/7

wk (n[ 638)>393/7 wk(n[ 7565 )

400/7e403/7 wk(n[ 522)

>403(n[

Male 309 (48.4) 3805 (50.3) 271 (51.9) 1517

Birthweight, g 3416 448b 3527 461 3543 443b 35915-min Apgar score 5 2 (0.3) 60 (0.8) 4 (0.8) 24Umbilical artery pH 7.0 5/183 (2.7) 58/3445 (1.7) 5/169 (2.9) 30/1Admission to NICU 46 (7.2) 703 (9.3) 48 (9.2) 309

Antepartum or intrapartum death 0 9 (0.1) 0 4

Hypoxic-ischemic Encephalopathy 0 5 (0.06) 1 (0.2) 2

Neonatal death 0 3 (0.04) 0 1

Perinatal deatha 0 12 (0.1) 0 5

All data presented as mean SD or N (%).EM, expectant management; IOL, induction of labor; NICU, neonatal intensive care unit; VBAC, vaginal birth after cesarean.

a b co

Am

ajog.org Obstetrics Researchat 390/7-393/7 weeks continued to beassociated with a signicantly higherchance of VBAC compared to expectantmanagement. Although induction oflabor at later gestational ages also

Antepartum, intrapartum, or neonatal death; P < .05 for

Palatnik. Induction of labor after prior cesarean delivery.was associated with higher odds ofVBAC, this difference did not reachstatistical signicance. Similarly, in the

TABLE 4Association of labor induction with Vto expectant managementGestational age at IOL, wk OR

VBAC

390/7e393/7 1.78

400/7e403/7 1.15

410/7e413/7 0.94

Risk of uterine rupture

390/7e393/7 2.69

400/7e403/7 2.33

410/7e413/7 3.52

aOR, adjusted odds ratio; CI, confidence interval; IOL, induction

a Adjusted for maternal age, race, recurrent indication for priopresence of prior VBAC.

Palatnik. Induction of labor after prior cesarean delivery. Ammultivariable analysis, induction of la-bor at 390/7-393/7 weeks compared toexpectant management was associatedwith a higher risk of uterine rupture; thisassociation did not reach statistical sig-

mparison of labor induction vs EM at given gestational age.

J Obstet Gynecol 2015.nicance when labor inductions werecompared to expectant management atlater gestational ages. Results did not

BAC and uterine rupture compared

95% CI aORa 95% CI

1.48e2.13 1.31 1.03e1.67

0.94e1.39 1.21 0.93e1.56

0.73e1.21 1.04 0.76e1.43

1.30e5.57 2.73 1.22e6.12

0.96e5.65 2.31 0.84e6.33

0.71e17.50 3.13 0.58e16.88

of labor; OR, odds ratio; VBAC, vaginal birth after cesarean.

r cesarean delivery, presence of prior vaginal delivery, and

J Obstet Gynecol 2015.

MARCH 2015 Ameridiffer for the analyses that used differentinclusion criteria to construct theexpectant management group.

COMMENT

sarean delivery compared to

IOL EM/7 wk2933)

410/7e413/7

wk (n[ 471)>413/7 wk(n[ 547)

(51.7) 241 (51.2) 285 (52.1)

466 3645 428 3600 493(0.8) 2 (0.4) 8 (1.4)

417 (2.1) 3/172 (1.7) 6/294 (2.0)

(10.5) 39 (8.3) 64 (11.7)

(0.1) 0 4 (0.7)

(0.07) 0 1 (0.2)

(0.03) 0 0

(0.2) 0 4 (0.7)In this secondary analysis we sought toevaluate outcomes in women who wereat least 390/7 weeks of gestation with 1prior cesarean delivery and who under-went induction of labor vs expectantmanagement. We found that the rates ofboth VBAC and uterine rupture werehigher among women undergoing in-duction of labor at 390/7-393/7 weekscompared to those undergoing expec-tant management. A similar pattern inthese outcomes was seen with inductionof labor at later gestational ages,although the differences did not reachstatistical signicance, perhaps becauseof lesser power associated with a smallersample size. Other maternal andneonatal outcomes were similar betweenthe 2 groups.The results of this study in regard to

VBAC are in contrast to the results ofthose studies that have compared laborinductionwith spontaneous labor. Thesecohort studies of nulliparous womendescribed approximately a 2-foldincreased risk of cesarean delivery with

can Journal of Obstetrics& Gynecology 358.e4

Research Obstetrics ajog.orginduction of labor.6,7,16-18 However, ashas been noted by others, to evaluateoutcomes associated with labor induc-tion in an actual clinical context, therelevant clinical comparison is betweenlabor induction and expectant manage-ment.12-14 When this comparison hasbeen made among women without aprior cesarean, induction of labor 39weeks actually has been associated with alower chance of cesarean.14,19-22 Onegroup of investigators who evaluatedwomen with a prior cesarean also foundthat labor induction was associated witha lower chance of cesarean.23 This studyadditionally suggested that there was nodifference in the risk of uterine ruptureassociated with labor induction or peri-natal mortality, although there was asignicant increase in the frequency ofneonatal intensive care unit admissionand postpartum hemorrhage amongwomen who underwent labor inductionat some gestational ages. However, thedata source used was an administrativedatabase collected over>3 decades; sucha data source may be prone to multiplebiases.24 In contrast, our study used adatabase derived from direct chartabstraction by trained research nurses.Nevertheless, our results that labor in-duction was associated with a lowerchance of repeat cesarean and a greaterchance of VBAC are similar to those ofStock et al.23

In contrast, our result with regard tothe association between labor inductionand uterine rupture is not consistentwith that of Stock et al,23 but is consis-tent with other observational studies,even if those studies used spontaneouslabor as the comparison group.3,9-10 Arecent metaanalysis comparing induc-tion and spontaneous labor after priorcesarean showed 1.6-fold increase in theincidence of uterine rupture with in-duction (95% condence interval,1.13e2.31).25 Similarly, in this study, theodds of uterine rupture in the context oflabor induction were approximatelydoubled, although the attributable riskincrease was

17. Yeast JD, Jones A, Poskin M. Induction oflabor and the relationship to cesarean delivery: areview of 7001 consecutive inductions. Am JObstet Gynecol 1999;180:628-33.18. Smith KM, Hoffman MK, Sciscione A. Elec-tive induction of labor in nulliparous women in-creases the risk of cesarean delivery. ObstetGynecol 2003;101:45S.19. Cheng YW, Kaimal AJ, Snowden JM,Nicholson JM, Caughey AB. Induction of la-bor compared to expectant management inlow-risk women and associated perinataloutcomes. Am J Obstet Gynecol 2012;207:502.e1-8.

20. Glmezoglu AM,Crowther CA,Middleton P,Heatley E. Induction of labor for improving birthoutcomes for women at or beyond term.Cochrane Database Syst Rev 2012;6:CD004945.21. Darney BG, Snowden JM, Cheng YW, et al.Elective induction of labor at term comparedwith expectant management: maternal andneonatal outcomes. Obstet Gynecol 2013;122:761-9.22. Gibson KS, Waters TP, Bailit JL. Maternaland neonatal outcomes in electively inducedlow-risk termpregnancies. AmJObstet Gynecol2014;211:249.e1-16.

23. Stock SJ, Ferguson E, Duffy A, Ford I,Chalmers J, Norman JE. Outcomes of inductionof labor in women with previous cesarean de-livery: a retrospective cohort study using apopulation database. PLoS One 2013;8:e60404.24. Grimes DA. Epidemiologic research usingadministrative databases: garbage in, garbageout. Obstet Gynecol 2010;116:1018-9.25. Rossi AC, Prefumo F. Pregnancy outcomesof induced labor in women with previous ce-sarean section: a systematic review and meta-analysis. Arch Gynecol Obstet 2015;291:273-80.

ajog.org Obstetrics ResearchMARCH 2015 American Journal of Obstetrics& Gynecology 358.e6

Induction of labor versus expectant management for women with a prior cesarean deliveryMaterials and MethodsResultsCommentAcknowledgmentReferences