OBSTETRICS Contemporary cesarean delivery practice in the United States Jun Zhang, PhD, MD; James Troendle, PhD; Uma M. Reddy, MD, MPH; S. Katherine Laughon, MD, MS; D. Ware Branch, MD; Ronald Burkman, MD; Helain J. Landy, MD; Judith U. Hibbard, MD; Shoshana Haberman, MD, PhD; Mildred M. Ramirez, MD; Jennifer L. Bailit, MD, MPH; Matthew K. Hoffman, MD, MPH; Kimberly D. Gregory, MD, MPH; Victor H. Gonzalez-Quintero, MD, MPH; Michelle Kominiarek, MD; Lee A. Learman, MD, PhD; Christos G. Hatjis, MD; Paul van Veldhuisen, PhD; for the Consortium on Safe Labor OBJECTIVE: To describe contemporary cesarean delivery practice in the United States. STUDY DESIGN: Consortium on Safe Labor collected detailed labor and delivery information from 228,668 electronic medical records from 19 hospitals across the United States, 2002-2008. RESULTS: The overall cesarean delivery rate was 30.5%. The 31.2% of nulliparous women were delivered by cesarean section. Prelabor repeat cesarean delivery due to a previous uterine scar contributed 30.9% of all cesarean sections. The 28.8% of women with a uterine scar had a trial of labor and the success rate was 57.1%. The 43.8% women at- tempting vaginal delivery had induction. Half of cesarean for dystocia in induced labor were performed before 6 cm of cervical dilation. CONCLUSION: To decrease cesarean delivery rate in the United States, reducing primary cesarean delivery is the key. Increasing vaginal birth after previous cesarean rate is urgently needed. Cesarean section for dystocia should be avoided before the active phase is established, par- ticularly in nulliparous women and in induced labor. Key words: cesarean delivery, induction, labor, vaginal birth after previous cesarean Cite this article as: Zhang J, Troendle J, Reddy UM, et al, for the Consortium on Safe Labor. Contemporary cesarean delivery practice in the United States. Am J Obstet Gynecol 2010;203:326.e1-10. C esarean delivery has become the most common major surgical pro- cedure in many parts of the world. 1,2 The national rate of cesarean delivery in the United States has increased more than 50% since 1996 to 31.8% in 2007. 3 This upward trajectory appears likely to con- tinue in the near future. Reasons for the increase are multifac- eted. Delayed childbearing, increasing ma- ternal body mass, more multifetal gesta- tions, and low use of vaginal birth after previous cesarean (VBAC) are commonly cited causes. 4,5 Cesarean delivery on ma- ternal request 6 and physicians’ fear of liti- gation due to a poor obstetric outcome 7 may also be contributing to the escalating rate of cesarean delivery. Despite anecdotal evidence and the common belief that cer- tain cesarean deliveries may be unneces- sary, the magnitude of unnecessary cesar- ean section at the national level is unknown. Reliable, detailed information on labor and delivery is often unavailable in a large population. Consequently, our understanding of the underlying causes of high cesarean rate is incomplete. To meet this challenge, the Eunice Shriver Kennedy National Institute of Child Health and Human Development, National Institutes of Health, in collabora- tion with 12 institutions across the coun- try, conducted a retrospective observa- tional study entitled “the Consortium on Safe Labor.” The goal of the study was to collect comprehensive information on contemporary labor and delivery practice in multiple institutions. This paper de- scribes the contemporary cesarean deliv- ery practice in the US population. MATERIALS AND METHODS The Consortium on Safe Labor in- cluded 12 clinical centers (with 19 hos- From the Eunice Kennedy Shriver National Institute of Child Health and Human Development (Drs Zhang, Troendle, Reddy, and Laughon), National Institutes of Health, Bethesda, MD; Intermountain HealthCare and University of Utah (Dr Branch), Salt Lake City, UT; Baystate Medical Center (Dr Burkman), Springfield, MA; Georgetown University Hospital (Dr Landy), MedStar Health, Washington, DC; University of Illinois at Chicago (Dr Hibbard), Chicago, IL; Maimonides Medical Center (Dr Haberman), Brooklyn, NY; University of Texas Health Science Center at Houston (Dr Ramirez), Houston, TX; MetroHealth Medical Center (Dr Bailit), Cleveland, OH; Christiana Care Health System (Dr Hoffman), Wilmington, DE; Cedars-Sinai Medical Center (Dr Gregory), Los Angeles, CA; University of Miami (Dr Gonzalez-Quintero), Miami, FL; Indiana University–Clarian Health (Drs Kominiarek and Learman), Indianapolis, IN; Summa Health System, Akron City Hospital (Dr Hatjis), Akron, OH; and The EMMES Corporation (Dr van Veldhuisen), Rockville, MD. Received April 6, 2010; revised May 31, 2010; accepted June 21, 2010. Reprints not available from the authors. The Consortium on Safe Labor was supported by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, through Contract no. HHSN267200603425C. 0002-9378/free • Published by Mosby, Inc. • doi: 10.1016/j.ajog.2010.06.058 For Editors’ Commentary, see Table of Contents Research www. AJOG.org 326.e1 American Journal of Obstetrics & Gynecology OCTOBER 2010
Transcript
O
CJDMVC
Ot
Sdh
Rncat
CO
CcnU5ut
F(IMMMCCMMSC
R
R
TKH
0
Research www.AJOG.org
3
BSTETRICS
ontemporary cesarean delivery practice in the United Statesun Zhang, PhD, MD; James Troendle, PhD; Uma M. Reddy, MD, MPH; S. Katherine Laughon, MD, MS;. Ware Branch, MD; Ronald Burkman, MD; Helain J. Landy, MD; Judith U. Hibbard, MD; Shoshana Haberman, MD, PhD;ildred M. Ramirez, MD; Jennifer L. Bailit, MD, MPH; Matthew K. Hoffman, MD, MPH; Kimberly D. Gregory, MD, MPH;ictor H. Gonzalez-Quintero, MD, MPH; Michelle Kominiarek, MD; Lee A. Learman, MD, PhD;hristos G. Hatjis, MD; Paul van Veldhuisen, PhD; for the Consortium on Safe Labor
ti
Cradt
K
BJECTIVE: To describe contemporary cesarean delivery practice inhe United States.
TUDY DESIGN: Consortium on Safe Labor collected detailed labor andelivery information from 228,668 electronic medical records from 19ospitals across the United States, 2002-2008.
ESULTS: The overall cesarean delivery rate was 30.5%. The 31.2% ofulliparous women were delivered by cesarean section. Prelabor repeatesarean delivery due to a previous uterine scar contributed 30.9% ofll cesarean sections. The 28.8% of women with a uterine scar had a
bstet Gynecol 2010;203:326.e1-10.
ettpc
26.e1 American Journal of Obstetrics & Gynecology OCTOBER 2010
empting vaginal delivery had induction. Half of cesarean for dystocia innduced labor were performed before 6 cm of cervical dilation.
ONCLUSION: To decrease cesarean delivery rate in the United States,educing primary cesarean delivery is the key. Increasing vaginal birthfter previous cesarean rate is urgently needed. Cesarean section forystocia should be avoided before the active phase is established, par-icularly in nulliparous women and in induced labor.
ey words: cesarean delivery, induction, labor, vaginal birth after
rial of labor and the success rate was 57.1%. The 43.8% women at- previous cesarean
ite this article as: Zhang J, Troendle J, Reddy UM, et al, for the Consortium on Safe Labor. Contemporary cesarean delivery practice in the United States. Am J
tgmretseuoiuh
SCNtttSccise
MT
esarean delivery has become themost common major surgical pro-
edure in many parts of the world.1,2 Theational rate of cesarean delivery in thenited States has increased more than
0% since 1996 to 31.8% in 2007.3 Thispward trajectory appears likely to con-
inue in the near future.
rom the Eunice Kennedy Shriver National InstitDrs Zhang, Troendle, Reddy, and Laughon), Nntermountain HealthCare and University of U
edical Center (Dr Burkman), Springfield, MAedStar Health, Washington, DC; University oaimonides Medical Center (Dr Haberman), B
enter at Houston (Dr Ramirez), Houston, TX;leveland, OH; Christiana Care Health System (edical Center (Dr Gregory), Los Angeles, CA;iami, FL; Indiana University–Clarian Health (
umma Health System, Akron City Hospital (Dorporation (Dr van Veldhuisen), Rockville, M
eceived April 6, 2010; revised May 31, 2010; a
eprints not available from the authors.
he Consortium on Safe Labor was supported bennedy Shriver National Institute of Child Healtealth, through Contract no. HHSN267200603
002-9378/free • Published by Mosby, Inc. • d
For Editors’ Commentary, see Table
Reasons for the increase are multifac-ted. Delayed childbearing, increasing ma-ernal body mass, more multifetal gesta-ions, and low use of vaginal birth afterrevious cesarean (VBAC) are commonlyited causes.4,5 Cesarean delivery on ma-
of Child Health and Human Developmentnal Institutes of Health, Bethesda, MD;(Dr Branch), Salt Lake City, UT; Baystateorgetown University Hospital (Dr Landy),nois at Chicago (Dr Hibbard), Chicago, IL;klyn, NY; University of Texas Health SciencetroHealth Medical Center (Dr Bailit),Hoffman), Wilmington, DE; Cedars-Sinaiiversity of Miami (Dr Gonzalez-Quintero),Kominiarek and Learman), Indianapolis, IN;
atjis), Akron, OH; and The EMMES
pted June 21, 2010.
e Intramural Research Program of the Euniced Human Development, National Institutes ofC.
10.1016/j.ajog.2010.06.058
ontents
c
ernal request6 and physicians’ fear of liti-ation due to a poor obstetric outcome7
ay also be contributing to the escalatingate of cesarean delivery. Despite anecdotalvidence and the common belief that cer-ain cesarean deliveries may be unneces-ary, the magnitude of unnecessary cesar-an section at the national level isnknown. Reliable, detailed informationn labor and delivery is often unavailable
n a large population. Consequently, ournderstanding of the underlying causes ofigh cesarean rate is incomplete.To meet this challenge, the Eunice
hriver Kennedy National Institute ofhild Health and Human Development,ational Institutes of Health, in collabora-
ion with 12 institutions across the coun-ry, conducted a retrospective observa-ional study entitled “the Consortium onafe Labor.” The goal of the study was toollect comprehensive information onontemporary labor and delivery practicen multiple institutions. This paper de-cribes the contemporary cesarean deliv-ry practice in the US population.
ATERIALS AND METHODShe Consortium on Safe Labor in-
uteatiotah; Gef IllirooMeDrUnDrsr HD.
cce
y thh an425
oi:
of C
luded 12 clinical centers (with 19 hos-
po(u92TaegUi2sboctdlf2pimgttptitpeotlntcTti
tmccwinsNtdi
wcldrmsrcttbttgTat
spffropocucvtdseptt
ttntsnnerppfNtat
sdsota
ee2pFfipcmecn(vaerdgeorptcicaieiktcw
aotapwdasGwc
www.AJOG.org Obstetrics Research
itals) across 9 American Collegef Obstetricians and GynecologistsACOG) US districts. There were 8niversity affiliated teaching hospitals,teaching community hospitals, andnonteaching community hospitals.
hey were chosen because of the avail-bility of electronic medical records atach institution and because their geo-raphic distribution covers all ACOGS districts. A total of 228,668 deliver-
es with 233,844 newborns between002 and 2008 were included in thetudy. Between 2005 and 2007, 87%irths occurred. All births at 23 weeksr later in these institutions were in-luded. A total of 9.5% of women con-ributed more than 1 delivery to theatabase. To avoid intraperson corre-
ation, we selected the first deliveryrom each subject in the study, leaving06,969 deliveries for analysis. Partici-ating institutions extracted detailed
nformation from their electronicedical records on maternal demo-
raphic characteristics, medical his-ory, reproductive, and prenatal his-ory, labor, and delivery summary,ostpartum and newborn informa-ion. Information from the neonatalntensive care unit (NICU) was linkedo the newborn records. Data on laborrogression were extracted from thelectronic labor database. Informationn hospital and physician characteris-ics was collected from surveys of theocal investigators, and maternal andewborn discharge summaries (in In-
ernational Classification of Diseases-9odes) were linked to each delivery.his project was approved by the insti-
utional review boards of all participat-ng institutions.
Data transferred from the clinical cen-ers were mapped to predefined com-
on codes for each variable at the dataoordinating center. Data inquiries,leaning, recoding, and logic checkingere performed. We also conducted val-
dation studies for 4 key outcome diag-oses, including cesarean for nonreas-uring fetal heart rate tracing, asphyxia,ICU admission for respiratory condi-
ions, and shoulder dystocia. To validateata, eligible charts were selected, and
nvestigators were asked to recollect data c
ith chart abstraction done by hand. Weompared the information hand col-ected from the medical charts with thatownloaded from the electronic medicalecords. Appendix Table 1 indicates thatost variables that were reviewed in this
tudy are highly accurate. Although ourecords were not sampled randomly, theonsistency among different records onhe same variable (eg, singleton, gesta-ional age, attempting vaginal birth, liveirth, vertex presentation) indicates thathe information provided in the valida-ion studies is reliable and likely to beeneralizable to the entire database.hus, the electronic medical records arereasonably accurate representation of
he medical charts.Approximately 5.9% of women in our
tudy had missing information on fetalresentation. Given the importance of
etal presentation in our analysis, we per-ormed multiple imputation.8 A logisticegression model imputed the likelihoodf vertex/nonvertex presentation in aarticular subject multiple times basedn other obstetric characteristics, in-luding maternal race, parity, previousterine scar, number of fetus, externalephalic version, smoking, placenta pre-ia, cephalopelvic disproportion, gesta-ional age, reason for admission to labor/elivery, trial of labor, induction, fetalcalp electrode, operative vaginal deliv-ry, and mode of delivery. When the im-uted data were analyzed, the uncer-ainty that was related to imputation wasaken into account.
To make our study population reflecthe overall US obstetric population ando minimize the impact of the variousumber of births from different institu-
ions, we assigned a weight to eachubject based on ACOG district, mater-al race/ethnicity (non-Hispanic white,on-Hispanic black, Hispanic, and oth-rs), parity (nulliparous vs multipa-ous), and plurality (singleton vs multi-le gestation). We first calculated therobability of each delivery with these 4
actors according to the 2004 Nationalatality data9; publicly available Na-
ional Natality data can no longer be sep-rated by state after 2004. Then, based onhe number of subjects each hospital
ontributed to the database, we as- fi
OCTOBER 2010 Americ
igned a weight to each subject. Appen-ix Table 2 indicates that the weightedtudy population is close to the entire USbstetric population. Therefore, we usedhe weighted sample throughout ournalyses.
We defined “attempting vaginal deliv-ry or a trial of labor” as all vaginal deliv-ries plus cesarean deliveries with at leastvaginal examination data in the labor
rogression (or labor curve) database.or indications for cesarean delivery, werst listed all major indications and theercent of cesarean deliveries with a spe-ific indication (1 woman may haveore than 1 indication). “Elective cesar-
an delivery” was defined as cesarean forlinical indications of: (1) elective as de-oted in the electronic medical record,2) declining a trial of labor, and (3) aariety of factors that are not consideredccepted indications for cesarean deliv-ry such as elderly gravida, multiparity,emote from term, postterm/postdates,iabetes, chorioamnionitis, chronic orestational hypertension without pre-clampsia/eclampsia, premature rupturef the membranes, human papillomavi-us infection, Group B streptococcusositive, polyhydramnios, fetal demise,ubal ligation, and social/religion con-erns. We then grouped all indicationsnto 3 hierarchical, mutually exclusiveategories: “clinically indicated,” “mixed,”nd “truly elective.” The “mixed” groupncluded cesarean deliveries where notnough detailed information (eg, humanmmunodeficiency virus with an un-nown viral load or unknown presenta-ion of twins) was available to judge the ne-essity or where the clinical indicationsere not that strong (eg, preeclampsia).Duration of labor arrest was calculated
s the duration of no appreciable changef cervical dilation in the first stage andhe time interval between the first 10 cmnd delivery in the second stage. “No ap-reciable change in cervical dilation”as defined as within 1 cm of change inilation before delivery. All statisticalnalyses were performed using SAS ver-ion 9.1 (SAS Institute, Inc, Cary, NC).iven that this is a descriptive analysisith a very large sample size, no statisti-
al testing was performed; nor were con-
dence intervals provided.
an Journal of Obstetrics & Gynecology 326.e2
Research Obstetrics www.AJOG.org
3
TABLE 1Cesarean rate by maternal characteristics and type of cesarean delivery
Variable
Proportionof thepopulation, %
Overallcesareanrate,a %
Primarycesareandelivery,a %
Repeatcesareandelivery,a %
Prelaborcesareandelivery,a %
Intrapartumcesareandelivery,a %
n (unweighted) 206,969 60,866 38,336 22,530 32,380 28,486................................................................................................................................................................................................................................................................................................................................................................................
n (weighted) 3,997,436 1,220,877 727,941 492,935 696,583 524,294................................................................................................................................................................................................................................................................................................................................................................................
Other 3.1 29.3 17.4 11.8 17.2 12.1................................................................................................................................................................................................................................................................................................................................................................................
Health insurance.......................................................................................................................................................................................................................................................................................................................................................................
Number of fetuses.......................................................................................................................................................................................................................................................................................................................................................................
Labor induction in women attempting vaginal delivery.......................................................................................................................................................................................................................................................................................................................................................................
No 56.2 11.8 9.1 2.7 0 11.8.......................................................................................................................................................................................................................................................................................................................................................................
Zhang. Contemporary cesarean delivery practice in the US. Am J Obstet Gynecol 2010. (continued )
26.e3 American Journal of Obstetrics & Gynecology OCTOBER 2010
RTotts(ito23pa
snswi(p
wcngea
4nan1wvuAVwefw
. Am
www.AJOG.org Obstetrics Research
ESULTShe overall rate of cesarean delivery inur cohort was 30.5%, varying from 20%o 44% among the participating hospi-als. Table 1 presents the overall andpecific categories of cesarean deliveryprimary vs repeat, and prelabor vsntrapartum) by maternal characteris-ics. With advancing maternal age, theverall cesarean rate doubled from1.0% at age �20 years to 42% in women5 years or older, mainly due to repeat,relabor cesarean deliveries. Obesity wasssociated with substantially higher ce-
Overall...................................................................................................................a Based on classification scheme proposed by Robson.10
Zhang J. Contemporary cesarean delivery practice in the US. A
arean rates in all categories. One in 3ulliparous women was delivered by ce-arean section (31.2%). Multiparousomen had an overall cesarean rate sim-
lar to that of nulliparous women30.0%), primarily due to prelabor, re-eat cesarean delivery.A total of 65.9% multifetal gestationsere delivered by cesarean section (in
omparison to 29.9% in singleton preg-ancies), and a majority of multifetalestations did not attempt vaginal deliv-ry. The induction rate was 36.2%, usingll deliveries as the denominator, or
3.8% among women attempting vagi-al delivery. The cesarean rate was twices high in induced labor than in sponta-eous labor in all pregnancies (21.1% vs1.8%) and in singleton nulliparousomen with vertex presentation (31.4%s 14.2%). In women with a previousterine scar, 28.8% had a trial of labor.mong them, the rate of successfulBAC was 57.1%. Overall, 83.6% ofomen with a uterine scar were deliv-
red by cesarean section. A total of 92.8%etuses with nonvertex presentationere delivered by cesarean section.
To further examine the relative contri-ution of obstetric factors to the cesar-an rate, we grouped the women into 7ategories according to the classificationcheme proposed by Robson.10 Table 2ndicates that term pregnancies with aertex, singleton fetus, and previousterine scar contributed the most cesar-an deliveries in the United States30.9%), followed by term gestationsith a singleton, vertex fetus whose laboras induced (19.2%). Multifetal gesta-
ions and pregnancies with nonvertexresentation accounted for only 6.7% ofll births but contributed to one-fifth ofll cesarean deliveries.
The main indication for prelabor ce-arean delivery was previous uterinecar, followed by fetal malpresentationTable 3). Among intrapartum cesareaneliveries, approximately half of themere performed for “failure to progress”
Failure to progress/cephalopelvicdisproportion..........................................................................................................
Total...................................................................................................................a Women may have more than 1 indication. The total percentag
include “elective”, declining trial of labor, elder gravid, multrioamnionitis, chronic or gestational hypertension without prhuman papillomavirus infection, Group B streptococcus posreligion concerns; c Clinically indicated includes emergency,progress, cephalopelvic disproportion, failed induction, failedabruption, placenta previa, shoulder dystocia, and history obreech/malpresentation, fetal anomalies, fetal macrosomia,preeclampsia/eclampsia, other; e Truly elective: without any
Zhang. Contemporary cesarean delivery practice in the US
r “cephalopelvic disproportion,” and i
26.e5 American Journal of Obstetrics & Gynecolo
ore than a quarter were performed forndications of nonreassuring fetal testingr fetal distress. “Truly elective” cesareanelivery accounted for 9.6% of prelabornd 2.1% of intrapartum cesarean deliv-ries As some prelabor cesarean deliver-es in the “mixed” group were eligible for
trial of labor (eg, single previous lowransverse cesarean section and vertex-ertex twin gestation), the “truly elec-ive” cases may be an underestimate inhe current study.
The cesarean rate was higher at thearlier gestational ages of delivery (Fig-re 1). For example, the cesarean de-
ivery rate was over 60% at 28 weeksestation and declined gradually withdvancing gestation. In nulliparousomen, intrapartum cesarean deliveries
onstituted the majority of cesarean sec-ions at term (�37 weeks). Among mul-iparous women, most cesarean deliver-
y exceed 100%; b Indications for elective cesarean deliveryy, remote from term, postterm/postdates, diabetes, cho-mpsia/eclampsia, premature rupture of the membranes,polyhydramnios, fetal demise, tubal ligation, and social/reassuring fetal heart rate tracing/fetal distress, failure toeps, failed vaginal birth after previous cesarean, placentaoulder dystocia; d Mixed includes: previous uterine scar,an immunodeficiency virus infection, multiple gestation,ation in the “clinically indicated” or “mixed” categories.
J Obstet Gynecol 2010.
es occurred before the onset of labor
gy OCTOBER 2010
cross all gestational ages. Unlike in mostomen, the cesarean rate in women withuterine scar did not decline with ad-
ancing gestation until 40 weeks.We further examined the timing of ce-
arean delivery relative to cervical dila-ion among those women with vertex,ingleton gestation attempting vaginalelivery (Figure 2). Induced labor wasssociated with twice as many cesareaneliveries as spontaneous labor in bothulliparous and multiparous women be-
ore and after 39 weeks, and the cesareanection was performed earlier in terms ofervical dilation in induced labor than inpontaneous labor. The cesarean rateas substantially higher in women with aterine scar (47% and 33% before andfter 39 weeks, respectively), and the ce-arean section was performed at lowerervical dilation before 39 weeks than af-er 39 weeks.
Table 4 presents total duration of la-or arrest before intrapartum cesareanelivery for “failure to progress,” “ceph-lopelvic disproportion,” or “failed in-uction” among singleton, vertex pre-entation with, and without a uterinecar. The duration of labor arrest de-lined as labor advanced. At 4 cm of cer-ical dilation, for example, the medianuration for nulliparous women was 4ours and less than 3 hours at 6 cm. Theuration was similar between spontane-us and induced labors. However, nearlyalf of cesarean sections were performedefore 6 cm in induced labor and inomen with a previous uterine scar.ne-third of cesarean deliveries at the
econd stage were performed at less thanhours in nulliparous women, whereas,quarter were performed at less than 2ours in multiparous women.
OMMENTur study is a large, contemporary as-
essment of cesarean delivery practiceor the US obstetric population. Weound that:
One in 3 nulliparous women was de-livered by cesarean section.Prelabor repeat cesarean delivery dueto a previous uterine scar was the mostcommon reason for cesarean section,
reaPde
.........
4.........
.........
2.........
.........
1.........
.........
.........
.........
.........
.........
8.........
.........
10.........
e maipariteeclaitive,non-forcf shhumindic
contributing almost a third of all ce-
●
●
VTdaddtopwoq23irfTtottTaamctidi
otLr2rhfodctoflpdhfow
C(Z
CoZ
www.AJOG.org Obstetrics Research
sarean deliveries. In contrast, the rateof trial of labor is disappointingly low,and the associated success rate for vag-inal birth has declined.Approximately 44% of women at-tempting vaginal delivery had induced
FIGURE 1Cesarean delivery rate by gestation
esarean delivery rate by type of cesarean (prnulliparous women, multiparous women, and whang. Contemporary cesarean delivery practice in the US. A
FIGURE 2Cervical dilation at cesarean delive
ervical dilation at intrapartum cesarean deliverynset of labor (induced vs spontaneous onset), phang. Contemporary cesarean delivery practice in the US. A
labor, and that within this group thecesarean rate was twice as high as inwomen with spontaneous labor.A high percentage of intrapartum ce-sarean deliveries were performed be-fore 6 cm of cervical dilation, particu-
age
or vs intrapartum), gestational age and parityn with a uterine scar).bstet Gynecol 2010.
ng women attempting vaginal delivery by parity,ious uterine scar in singleton gestations.
bstet Gynecol 2010.
t
OCTOBER 2010 Americ
larly in nulliparous women, inducedlabor, and women attempting VBAC.
BAChe VBAC rate experienced precipitousecline since 1996, which coincided withn over 50% increase in cesarean rateuring the same period.3 Our study in-icates that prelabor repeat cesarean sec-ions currently have a profound impactn the overall cesarean rate. Some ex-erts estimate that two-thirds of womenith a uterine scar are eligible for a trialf labor.11 Yet, the rate of trial of labor isuite low in the United States. In 1999-002, Landon et al12 reported a rate of8.9% in a NICHD Maternal-Fetal Med-cine Unit Research Network study. Theate of trial of labor declined significantlyrom 48.3% in 1999 to 30.7% in 2002.his is consistent with our finding that
he rate was 28.8% in 2005-2007, andurs may be an overestimate for the na-ional level because academic institu-ions are overrepresented in our study.he low rate of trial of labor has beenttributed in part to a slight increase inbsolute risks of maternal and neonatalorbidity, professional liability con-
erns and physician and patient’s atti-ude toward VBAC.5 The recent changen ACOG recommendations also has re-uced the availability of VBAC services
n many hospitals.13
The success rate of attempted VBAC inur study was markedly lower (57.1%)han that in previous large studies.ieberman et al14 noted a VBAC successate of 87% in 41 birth centers in 1990-000. Macones et al15 found a successate of 75.5% in a regional study with 17ospitals in 1996-2000, similar to that
ound by Landon et al12 (73.4%). Greg-ry et al,16 using 2002 California hospitalischarge data, reported an overall suc-ess rate of 67%. One possible explana-ion of our finding is that the success ratef attempted VBAC is more recently in-uenced by other factors. Physicians andatients may be less committed evenuring a trial of labor.17 For instance,alf of the intrapartum cesarean sections
or dystocia were performed before 6 cmf cervical dilation in our study. Manyomen may not be in active phase before
al
elabomem J O
ry
amorevm J O
hat time.18 The frequency of oxytocin
an Journal of Obstetrics & Gynecology 326.e6
ulr
stuvdtcsntpaeh
IIrio
t92osiwtio
bfvwdleoltm
iratdbs
TcTsoudtdsnwhch
. Am
Research Obstetrics www.AJOG.org
3
se for labor augmentation in the trial ofabor was one-sixth of that in multipa-ous women without a uterine scar.
Another possible reason for the lowuccess rate may be related to our defini-ion of “a trial of labor.” A woman with aterine scar, minimal dilation, and 2 pel-ic examinations recorded in the laboratabase would have been considered arial of labor. Such cases may have in-reased the VBAC rate but reduced theuccess rate. However, findings of vagi-al examinations that were entered in
he labor progression (labor curve) com-uter system on the labor floor are oftenn indication for intended vaginal deliv-ry. Thus, the above scenario may notave substantially affected our results.
nduction of laborn parallel with the increasing cesareanate, the rate of labor induction also hasncreased significantly in many devel-
TABLE 4Duration of labor arrest (in hours) pto progress/cephalopelvic dispropo
Second stage 3.8 (100)...................................................................................................................
Among the second stage cesarean deliveries..........................................................................................................
�2 h 12..........................................................................................................
2-3 h 33..........................................................................................................
3-4 h 55..........................................................................................................
�4 h 100...................................................................................................................
—, the number of subjects is too small to have a meaningful
Zhang. Contemporary cesarean delivery practice in the US
ped countries, including a doubling of s
26.e7 American Journal of Obstetrics & Gynecolo
he US rate between 1990-2000 (from.5% in 1990 to 20.2% in 2000, and2.6% in 2006).19 In comparison, theverall induction rate was 35.2% in ourtudy (around 2006); the induction raten women attempting vaginal deliveryas 43.8%. These findings suggest that
he national figures based on birth certif-cates may have been an underestimatef the true national induction rate.20
For medically indicated induction of la-or, the benefits for the mother and/or theetus are indisputable. However, contro-ersies arise when labor is induced foromen with absent or marginal clinical in-ications.21-23 Our data show that induced
abor was twice as likely to result in cesar-an delivery than spontaneous labor. Halff cesarean sections for dystocia in induced
abor were performed before 6 cm of dila-ion, suggesting that clinical impatience
ay play a role in decision making. Some
r to intrapartum cesarean for “failuon” or “failed induction”
ncreases the risk of postpartum hemor-hage and blood transfusion.22 Therefore,lthough more research is warranted, cau-ion is needed to perform elective labor in-uction, as it may result in maternal mor-idity and repeat cesarean deliveries inubsequent pregnancies.
iming of intrapartumesarean deliveryhe high proportion of intrapartum ce-
arean deliveries performed before 6 cmf cervical dilation is concerning, partic-larly in nulliparous women and in in-uced labor. The active phase of labor isypically considered starting at 4 cm ofilation.24,25 However, a recent studyhowed that multiparous women mayot start the active phase until 6 cm,hereas, nulliparous women may notave a typical active phase of labor.18 Inontemporary nulliparous women whoad vaginal delivery and normal perina-
Women with a previousuterine scar having atrial of labor, median(cumulative percentage)
ours at 4 cm or 3 hours at 5 cm of dila-ion (results not shown). In the currenttudy, the median duration of labor ar-est in nulliparous women, including pe-iods before and after any intervention,as 4 hours at 4 cm. Thus, a cesarean
ection for dystocia before 6 cm in nul-iparous women appears too soon inome cases.
We also found that one-third of cesar-an deliveries at the seond stage wereerformed at less than 3 hours in nullip-rous women, whereas, a quarter wereerformed at less than 2 hours in multip-rous women. This is of concern becauseCOG guidelines define arrest of de-
cent as greater than 3 hours in nullipa-ous women with epidural analgesia andreater than 2 hours in multiparousomen with epidural analgesia.26 Given
hat 90% of the subjects used epiduralnalgesia and that the median durationn the second stage arrest includeddecision-to-incision” time, the trueaiting period in the second stage arrest
ppears shorter than ACOG recommen-ations in many cesarean cases.It should be noted that even though we
elected 12 institutions across 9 ACOGistricts, our study subjects were not aandom sample of all births in thenited States. Academic institutions areverrepresented in our cohort. This may
n part explain why our weighted pre-erm birth rate was higher (14.1%) thanhe national average (12.8% in 2006)19
nd why the induction rate in our studyas higher than the published national
verage. Yet, our overall cesarean rateas slightly lower than the correspond-
ng national average (30.5% vs 31.1% in006).19 Nonetheless, for the first timee were able to separate cesarean deliv-
ry by prelabor and intrapartum in aarge multicenter study. Our study pro-ides detailed information on when ce-arean section was performed, allowings to identify potential areas for optimalanagement.In summary, our study shows that 1 innulliparous women is delivered by ce-
arean section, which has a tremendousmpact on subsequent prelabor repeat
esarean delivery. Prelabor repeat cesar- t
an delivery now contributes almost ahird of all cesarean deliveries. To make aignificant impact on the high cesareanelivery rate in the United States, the fo-us should be preventing unnecessaryrimary cesarean deliveries from severalspects. First, we need to decrease theate of cesarean delivery associated withhigh rate of induction of labor. Cesar-
an section for dystocia should bevoided before active phase of labor isstablished particularly in nulliparousomen, induced labor, and VBAC at-
empts. Second, there should be a clini-ally accepted indication for performingesarean delivery. Finally, increasing ac-ess to and patient education on trial ofabor in women with a previous uterinecar and improving the success rate arergently needed. f
EFERENCES. Villar J, Valladares E, Wojdyla D, et al. Cae-arean delivery rates and pregnancy outcomes:he 2005 WHO global survey on maternal anderinatal health in Latin America. Lancet 2006;67:1819-29.. Lumbiganon P, Laopaiboon M, GülmezogluM, et al. Method of delivery and pregnancyutcomes in Asia: the WHO global survey onaternal and perinatal health 2007-08. Lancet010;375:490-9.. Hamilton BE, Martin JA, Ventura SJ. Births:reliminary data for 2007. National Vital Statis-ics Reports, Web release; vol. 57 no. 12.yattsville, MD: National Center for Health Sta-
istics. Released March 18, 2009.. Joseph KS, Young DC, Dodds L, et al.hanges in maternal characteristics and ob-tetric practice and recent increases in primaryesarean delivery. Obstet Gynecol 2003;102:91-800.. Macones GA. Clinical outcomes in VBAC at-empts: what to say to patients? Am J Obstetynecol 2008;199:1-2.. Habiba M, Kaminski M, Da Fré M, et al. Ce-arean section of request: a comparison of ob-tetricians’ attitudes in eight European coun-ries. BJOG 2006;113:647-56.. Murthy K, Grobman WA, Lee TA, Holl JL.ssociation between rising professional liability
nsurance premiums and primary cesarean de-ivery rates. Obstet Gynecol 2007;110:1264-9.. Rubin DB. Multiple imputation for nonre-ponse in surveys. New York: John Wiley; 1987.. Martin JA, Hamilton BE, Sutton PD, et al.irths: final data for 2004. National Vital Statis-
ics Reports; vol. 55 no. 1. Hyattsville, MD: Na-
ional Center for Health Statistics. 2006. n
OCTOBER 2010 Americ
0. Robson SM. Can we reduce the cesareanection rate? Best Practice Res Clin Obstetynaecol 2001;15:179-94.1. Landon MB. Vaginal birth after cesareanelivery. Clin Perinatol 2008;35:491-504.2. Landon MB, Hauth JC, Leveno KJ, et al.aternal and perinatal outcomes associatedith a trial of labor after prior cesarean delivery.Engl J Med 2004;351:2581-9.
3. Roberts RG. Deutchman M, King VJ, FryerE, Miyoshi TJ. Changing policies on vaginalirth after cesarean: impact on access. Birth007;34:316-22.4. Lieberman E, Ernst EK, Rooks JP, Staple-on S, Flamm B. Results of the national study ofaginal birth after cesarean in birth centers. Ob-tet Gynecol 2008;111:285-91.5. Macones GA, Peipert J, Nelson DB, et al.aternal complications with vaginal birth after
esarean delivery: a multicenter study. Am Jbstet Gynecol 2005;193:1656-62.6. Gregory KD, Korst LM, Fridman M, et al.aginal birth after cesarean: clinical risk factorsssociated with adverse outcome. Am J Obstetynecol 2008;198:452.e1-12.7. Kalish RB, McCullough L, Gupta M, ThalerT, Chervenak FA. Intrapartum elective cesar-an delivery: a previously unrecognized clinicalntity. Obstet Gynecol 2004;103:1137-41.8. Zhang J, Troendle J, Mikolajczyk R,undaram R, Beaver J, Fraser W. The naturalistory of the normal first stage of labor. Obstetynecol 2010;115:705-10.9. Martin JA, Hamilton BE, Sutton PD, et al.irths: final data for 2006. National Vital Statis-
ics Reports; vol. 57 no. 7. Hyattsville, MD: Na-ional Center for Health Statistics; 2009.0. Roberts CL, Bell JC, Ford JB, Morris JM.onitoring the quality of maternity care: howell are labour and delivery events reported inopulation health data? Pediatr Perinatal Epide-iol 2009;23:144-52.1. Vahratian A, Zhang J, Troendle JF,ciscione AC, Hoffman MK. Labor progressionnd risk of cesarean delivery in electively in-uced nulliparas. Obstet Gynecol 2005;105:98-704.2. Grobman WA. Elective induction: when?ver? Clin Obstet Gynecol 2007;50:537-46.3. Nicholson JM, Parry S, Caughey AB, Rosen, Keen A, Macones GA. The impact of thective management of risk in pregnancy at termn birth outcomes: a randomized clinical trial.m J Obstet Gynecol 2008;198:511.e1-15.4. Albers LL, Schiff M, Gorwoda JG. The
ength of active labor in normal pregnancies.bstet Gynecol 1996;87:355-9.5. Rouse DJ, Owen J, Hauth JC. Active-phase
abor arrest: oxytocin augmentation for at leasthours. Obstet Gynecol 1999;93:323-8.6. American College of Obstetricians and Gy-ecologists. Dystocia and augmentation of la-or. ACOG practice bulletin no. 49. Obstet Gy-
ecol 2003;102:1445-54.
an Journal of Obstetrics & Gynecology 326.e8
A
Research Obstetrics www.AJOG.org
3
PPENDIX
TABLE 1Validity of data from electronic me
Sample selection criteriaInele
Nulliparous, singleton, live, vertexfetus at admission, Attemptingvaginal delivery and cesareandelivery due to non-reassuring fetalheart rate tracing (n � 5662)
Singleton, gestational age �34 wk,attempting vaginal delivery, had anApgar score at 5 min �4 orumbilical cord pH �7.0 or basedeficit �12 mmol/L (n � 503)
e fetus at admission 99.9 0.1 0.0.............................................................................................................................................................................................................................................................
26.e9 American Journal of Obstetrics & Gynecology OCTOBER 2010
Zhang. Contemporary cesarean delivery practice in the US. Am J Obstet Gynecol 2010.
www.AJOG.org Obstetrics Research
OCTOBER 2010 American
TABLE 2Description of the study population in comparisonto the 2004 US birth cohort
CharacteristicsCSL cohort(Nonweighted)
CSL cohort(Weighted)
US birthcohorta
No. of women 206,969 3,997,436 4,018,091..............................................................................................................................................................................................................................................
Non-Hispanic white 49 57 57.....................................................................................................................................................................................................................................
Non-Hispanic black 22 14 14.....................................................................................................................................................................................................................................
CSL, The Consortium on Safe Labor.a 2006 National Natality data;18 b These factors plus American College of Obstetricians and Gynecologists district were used to
