attempt and success rates for vaginal birth after caesarean section in relation to complications of...

Attempt and success rates for vaginal birth afterCaesarean section in relation to complications ofthe previous pregnancy

Victoria L. Holt*{, and Beth A. Mueller*{Departments of *Epidemiology and {Health Services, School of PublicHealth and Community Medicine, University of Washington and{Division of Public Health Sciences, Fred Hutchinson Cancer ResearchCenter, Seattle, Washington, USA

Summary. The relationships between prior obstetrical complicationsand subsequent trial of labour and vaginal birth after Caesarean (VBAC)success likelihood were examined among a cohort of Washington Statewomen with a first livebirth via Caesarean delivery and a secondlivebirth between 1987 and 1993 (n = 10 110). Overall, 64% of the cohortundertook a labour trial, and 62% of those who attempted VBACdelivery were successful, for an overall VBAC rate of 40%. Women withfetal macrosomia, cephalopelvic disproportion, prolonged labour, dia-betes, or placental problems in the first pregnancy were less likely toundergo a labour trial in the second pregnancy than those without thesefactors, while women with prior induced labour, genital herpes, fetaldistress, or breech presentation were more likely to attempt vaginal birth.Approximately half of women with prior macrosomia, labour problems,or chronic medical conditions who attempted VBAC succeeded, as didthree-quarters of women with prior breech presentation or placentalconditions. Overall VBAC rates were around 33% for women withprevious fetal macrosomia, labour problems, or chronic medicalconditions, and 45±55% among those with herpes, fetal distress orbreech presentation at the first birth. Trial of labour should especially beencouraged among women without prior labour problems.

Introduction

The US Caesarean delivery rate has been over 22% since 1985, and in 1993 nearly

one million American women had this surgical procedure.1 Over one-third of all

Paediatric and Perinatal Epidemiology 1997, 11, Suppl. 1, 63±72

Address for correspondence: Dr V. L. Holt, University of Washington, School of Public Healthand Community Medicine, MCH Program, Box 357230, Seattle, WA 98195-7230, USA.

63# 1997 Blackwell Science Ltd.

Sup 1-8 Disc

Caesareans are repeats, and a prior Caesarean is the most common indication for

Caesarean delivery. This circumstance, and an expanding awareness of the safety

of vaginal birth among women with prior Caesarean delivery, have led North

American clinical and public health organisations to promote increased use of

vaginal birth after Caesarean (VBAC) procedures.2±4 VBAC rates have quadrupled

in the past decade in the US, rising from 6.6% of all women with prior Caesareans

in 1985 to 25.4% in 1993, but the current rate is still below the year 2000 goal of

35%.1,5 Reasons for the continued reliance on repeat Caesarean delivery are

complex, and some investigators have hypothesised that physicians and patients

may be unwilling to recommend or undertake a VBAC trial because the presumed

likelihood of success is unacceptably low.6±9 Knowledge of VBAC success rates

associated with various risk factors may encourage appropriate VBAC attempts,

and thereby decrease the occurrence of repeat Caesarean delivery.

Reproductive history, particularly the indication for the initial Caesarean

delivery, has been found in numerous studies to have an impact on the likelihood

of VBAC success, but these studies have been for the most part small-scale

prospective trials conducted within single institutions.10±17 The objective of this

analysis was to examine the relationships between prior obstetrical complications

and subsequent trial of labour and VBAC success likelihood in an unselected

population of women delivering in an entire state over several years. Our goals

were to determine if associations discovered in clinical trials were replicable in a

large population-based study and to investigate several previously unstudied

variables related to prior obstetrical complications.

Methods

Using maternally linked birth certificates, data were obtained for all primiparous

women with a singleton livebirth via Caesarean delivery in Washington State

between 1987 and 1993 and a second singleton livebirth in Washington State

during the same period (n = 10989). Details of the methods used to create this

linkage have been reported elsewhere.18 The outcome of interest was delivery

method at second birth, categorised on the birth certificate as: (1) vaginal birth or

vaginal birth after Caesarean, (2) repeat Caesarean after trial of labour, or (3)

repeat Caesarean without a trial of labour. We excluded women with second

births prior to 1989 when the trial of labour variable was added to the birth

certificate (n = 369) and those with unknown delivery method at second birth

(n = 510), leaving 10110 women for analysis.

All demographic and obstetrical complication data were taken from the birth

certificates, with the exception of insurance coverage information, which was

obtained through linkage of birth certificates with Washington State Comprehen-

sive Hospital Abstract System (CHARS) data. Macrosomia was defined as a

birthweight 54000 g, and labour problems included the notation of cephalic

# 1997 Blackwell Science Ltd. Paediatric and Perinatal Epidemiology, 11, Suppl. 1, 63±72

64 V. L. Holt and B. A. MuellerSup 1-8 Disc

pelvic disproportion, prolonged labour, or dysfunctional labour on the birth

certificate. Dysfunctional labour was reported on the birth certificate beginning in

1989. Three rates related to delivery method at the second birth were calculated

for each of several obstetrical complications of the first birth: attempted trial of

labour rate (women who attempted VBAC/all women in cohort), successful trial

of labour rate (women with successful VBAC/women who attempted VBAC), and

VBAC rate (women with successful VBAC/all women in cohort). Differences in

VBAC rates between women with a particular first obstetrical complication and

those without that complication were determined using the chi-square test, with

P 40.05 denoting statistical significance.19

Results

This cohort of Washington State women was primarily of white race and married

at the time of their first birth (Table 1). Over 80% of study subjects did not smoke

during their first pregnancy, more than half had insurance covering delivery costs,

and most received first trimester prenatal care. With the exception of maternal age

(mean age = 25.3 years at first birth and 27.8 years at second), these characteristics

did not change appreciably from the first to the second birth.

Overall, 64% of the cohort undertook a trial of labour at their second birth, and

62% of those who attempted were successful. The resultant VBAC rate during the

study period was 40%. Several complications of the first labour and delivery

affected the likelihood of subsequent VBAC attempts. Women with a previous


VBAC success and previous obstetrical complications 65

Table 1. Demographic characteristics at time of first birth of women with a first birth byCaesarean delivery and a second birth by any delivery method in Washington State, USA,between 1987±93 (n = 10110)

n (%) n (%)

Age Smoked during pregnancy 1682 (17.6)< 20 years 1521 (15.1) Unknown 53520±24 years 3064 (30.3) Insurance25±29 years 3310 (32.8) Private 5281 (56.6)30-34 years 1768 (17.5) HMO 1338 (14.3)5 35 years 442 (4.4) Medicaid 2013 (21.6)Unknown 5 Self 501 (5.4)

Race Other 202 (2.2)White 8784 (88.5) Unknown 775Black 253 (2.5) Prenatal care onsetAsian 285 (2.9) 1st trimester 8210 (83.5)Hispanic 452 (4.6) 2nd trimester 1343 (13.7)Other 153 (1.5) 3rd trimester or no care 276 (2.8)Unknown 183 Unknown 281

Married 8032 (79.5)Unknown 10

Sup 1-8 Disc


66 V. L. Holt and B. A. Mueller

Table 2. First delivery complications of women with prior Caesarean delivery, by trial oflabour and VBAC success likelihood at second delivery, Washington State, USA, 1987±93

Women in Attempted Successful VBACFirst cohort trial of laboura trial of labourb ratec

delivery (n=10110) (n=6491) (n=4007)

n n (%) n (%) %Birthweight

<2500 g 594 388 (65.3) 261 (67.3) 43.92500±3999 g 7419 4834 (65.2) 3045 (63.0) 41.04000+ g 2082 1258 (60.4) * 693 (55.1) * 33.3 *unknown 15

Cephalopelvic disproportionno 6026 4045 (67.1) 2646 (65.4) 43.9yes 3973 2367 (59.6) * 1311 (55.4) * 33.0*unknown 111

Prolonged labourno 9060 5866 (64.7) 3640 (62.1) 40.2yes 939 546 (58.1) * 317 (58.1) 33.8 *unknown 111

Dysfunctional labour d

no 4958 3519 (71.0) 2143 (60.9) 43.2yes 600 426 (71.0) 224 (52.6) * 37.3 *unknown 4552

Diabetesno 9361 6029 (64.4) 3739 (62.0) 39.9established 49 25 (51.0) * 9 (36.0) * 18.4 *

gestational 267 155 (58.1) * 71 (45.8) * 26.6 *unknown 433

Hypertensionno 8635 5544 (64.2) 3463 (62.5) 40.1chronic 133 90 (67.7) 44 (48.9) * 33.1pregnancy-assoc. 1044 667 (63.9) 359 (53.8) * 34.1 *unknown 427

Infant deathno 9984 6417 (64.3) 3966 (61.8) 39.7yes 126 74 (58.7) 41 (55.4) 32.5

Induced labourno 8582 5443 (63.4) 3459 (63.5) 40.3yes 1337 909 (68.0) * 463 (50.9) * 34.6 *unknown 191

Placental abruption or placenta praeviano 9872 6348 (64.3) 3903 (61.5) 39.5yes 238 143 (60.1) * 104 (72.7) * 43.7

Genital herpesno 9238 5897 (63.8) 3616 (61.3) 39.1yes 449 315 (70.2) * 207 (65.7) 46.1 *unknown 423

continued

Sup 1-8 Disc

macrosomic infant, history of cephalopelvic disproportion, or prolonged first

labour were significantly less likely to undergo a subsequent trial of labour than

those without these conditions, as were women with established or gestational

diabetes, abruptio placentae or placenta praevia in their first pregnancy (Table 2).

Conversely, women with an induced first labour, or genital herpes, fetal distress,

or breech presentation complicating the first labour and delivery were

significantly more likely than women without these factors to attempt vaginal

delivery at their second birth.

Among women who did attempt VBAC delivery, the likelihood of success also

varied by obstetrical history (Table 2). Women whose first pregnancy was

associated with fetal macrosomia, cephalopelvic disproportion, dysfunctional

labour, established or gestational diabetes, chronic or pregnancy-associated

hypertension, or induced labour were significantly less likely to succeed after a

trial of labour at the second birth than those without such histories. Women with

the first delivery complications abruptio placentae/placenta praevia or breech

presentation who attempted vaginal delivery were significantly more likely to be

successful than those without these complications. The overall VBAC rates were

significantly lower among women with previous fetal macrosomia or labour

problems, as well as those with established or gestational diabetes or pregnancy-

associated hypertension during the first pregnancy, compared with women

without these conditions. Women with genital herpes, fetal distress or breech

presentation at the first birth had significantly higher VBAC rates at the

subsequent birth than those without such histories.

There were 6658 women in the cohort without hypertension or diabetes at

either birth who had a normal weight infant (2500±3999 g) in cephalic position at

the second delivery. Two-thirds of these low-risk women had a trial of labour at



Table 2. continued

Women in Attempted Successful VBACFirst cohort trial of laboura trial of labourb ratec

delivery (n=10110) (n=6491) (n=4007)

n n (%) n (%) %Fetal distress

no 8988 5718 (63.6) 3506 (61.3) 39.0yes 1122 773 (68.9) * 501 (64.8) 44.7 *

Breechno 8220 5102 (62.1) 2981 (58.4) 36.3yes 1779 1310 (73.6) * 976 (74.5) * 54.9 *unknown 111

a=women with trial of labour/all women in cohort (64% of cohort);b=women with successful VBAC/all women with trial of labor (62% of trials);c=women with successful VBAC/all women in cohort (40% of cohort);damong women with first birth in or after 1989 only.*Significantly different rate than those without this complication (P4 0.05)

Sup 1-8 Disc

their second delivery, and two-thirds of those who attempted delivered vaginally.

The overall VBAC rate for this group was 44%. Low-risk women with breech

presentation, abruptio placentae/placenta praevia, or fetal distress complicating

their first delivery were significantly more likely to have a trial of labour at the

second birth than those with prior cephalopelvic disproportion, prolonged or

dysfunctional labour and also more likely to be successful in their VBAC attempts

(Table 3).

Discussion

In our study approximately two-thirds of women attempted VBAC delivery, and

almost two-thirds of these attempts were successful. Our finding that attempts

among women with prior breech presentation were more successful (75%

succeeded) than those among women with prior labour problems such as

cephalopelvic disproportion (55% succeeded) confirms the results of earlier

smaller studies10±17 and extends them to a general population of primarily white

women delivering in 75 hospitals across an entire state. Through use of maternally-

linked vital data, we found previously unreported associations between successful

trial of labour and other prior obstetrical complications as well: women with

chronic or pregnancy-associated hypertension or established or gestational

diabetes in the first pregnancy who undertook a trial of labour were less likely

to have a VBAC than women without these conditions. These associations may be

as a result of recurrence of the complications in the second pregnancy, and may

reflect a medical judgment to abbreviate labour trials in women with serious



Table 3. Trial of labour and VBAC success likelihood among low-risk women with priorCaesarean delivery, by complication of first delivery, Washington State, USA, 1987±93a

Women in Attempted Successful VBACFirst cohort trial of labourb trial of labourc rated

delivery (n=6658) (n=4396) (n=2906)

n n (%) n (%) %Breech 1207 934 (77.4) 751 (80.4) 62.2AP or PP e 155 101 (65.2) 78 (77.2) 50.3Fetal distress 741 516 (69.6) 343 (66.5) 46.3Labour problems f 3116 1895 (60.8) 1117 (58.9) 35.9

aLow risk defined as no history of or current hypertension or diabetes, and a cephalicposition normal weight infant (2500±3999 g) at second delivery.bwomen with trial of labour/all women in cohort (66% of cohort);cwomen with successful VBAC/all women with trial of labour (66% of trials);dwomen with successful VBAC/all women in cohort (44% of cohort);eAP=abruptio placentae, PP=placenta praevia;fcephalic pelvic disproportion, prolonged labour, or dysfunctional labour.

Sup 1-8 Disc

medical conditions. We also found that prior induced labour was associated with

decreased trial of labour success. This result was noted in one prior study, which

reported that women with a first, failed induction of labour had lower VBAC

success in the subsequent birth even than women with prior cephalopelvic

disproportion.13

Our finding that two-thirds of women with a prior Caesarean who undertook

a trial of labour in their second birth successfully delivered vaginally is somewhat

lower than the 70±80% success rates reported in earlier, smaller studies. There may

be several reasons for this discrepancy. Most of the earlier research was carried out

in single hospital settings, often in the context of a prospective trial designed to

improve trial of labour and VBAC success rates. Our study represents a

population-based experience, with a variety of practitioner styles and hospital

policies and capabilities. In this respect, our study most closely corresponds to that

of Paterson et al. who reported a 71% VBAC success rate at the second deliveries of

women in 17 hospitals in one hospital region in the UK in 1988.6 The Paterson et al.study and the current study also found a similar proportion of women with

previous Caesarean delivery who attempted a trial of labour (60±65%). These

results may most closely approximate the likely experience of obstetrical patients

in routine practice settings.

Our lower VBAC success rates may also be a function of the reproductive

experience of our study cohort. We examined only the second births of women

with a Caesarean first birth, while most earlier research included women of all

parity levels with a Caesarean at any prior delivery. Since a successful vaginal

birth after a Caesarean delivery is associated with increased VBAC success in

subsequent births, studies including such women can be expected to have higher

success rates. We believe that it is useful to practitioners and patients to report

parity-specific VBAC success rates, and women with no prior vaginal births are

those for whom predictive information is most important.

Finally, our VBAC success rates may differ from those of hospital-based

studies because of our reliance on birth certificate data, with possible inaccuracy of

coding of delivery method at the second birth. Unpublished data from a

validation study that examined the accuracy of Washington State birth certificate

data indicate that trial of labour among women with prior Caesarean deliveries

may have been underreported in 1989 by up to 25%.20 In 1992, the format and

coding of the trial of labour variable on Washington State birth certificates

changed, and we observed a substantial increase in reported trial of labour among

births occurring in 1992 and 1993. When our analysis was stratified by whether

the second delivery occurred before 1992, the results concerning VBAC rates (the

coding of which had not changed) were essentially the same. Although trial of

labour attempts increased and trial of labour success decreased somewhat in the

latter time period, associations with prior reproductive history generally retained

the patterns seen in the overall analysis. One exception was noted: women with a



Sup 1-8 Disc

history of genital herpes were more likely than those without herpes to have a trial

of labour only in the second time period.

Non-clinical factors may influence VBAC rates differently in different settings,

which might affect the generalisability of our findings. We found that among the

non-clinical variables we examined, only insurance payer was associated with

VBAC likelihood. In our study, VBAC rates adjusted for insurance payer were not

significantly different from the unadjusted rates we presented in Table 2, and all

significant differences between women with and without a specific prior

obstetrical condition remained significant when using adjusted rates, reassuring

us that non-clinical factors were not important determinants of VBAC rates in our

population.

Results of our examination of the associations between prior obstetrical

complications and a woman's likelihood of undertaking a trial of labour may be of

note to clinicians and others interested in increasing appropriate VBAC attempts.

Women with cephalopelvic disproportion, prolonged or dysfunctional labour at

the first birth were less likely to attempt a subsequent vaginal delivery, perhaps

owing to the memory of a painful, long, unsuccessful initial labour. In contrast,

women with a history of genital herpes or breech presentation, who perhaps had

no labour before their initial Caesarean deliveries, were more likely to undertake a

labour trial at the second birth. This choice may be as a result of both the

physician's assessment of current low medical risk and of the patient's lack of

negative prior experience. Most researchers have found that the decision to

undertake a labour trial is a combination of the physician's and the patient's

choice.21-23 The patient's desire for convenience and to avoid pain play a part, as

does her assessment of the likelihood of successful completion of labour, which

often is unduly pessimistic.21

Two-thirds of women with prior Caesarean deliveries in our study attempted a

subsequent vaginal birth, and almost two-thirds of these women were successful.

Even women with prior labour problems who had a trial of labour were successful

over 50% of the time, and labour trials of low-risk women were successful up to

80% of the time. Physicians and their patients should be aware that women with

normal weight infants, without medical complications, and with non-recurring

causes of the first Caesarean delivery can achieve vaginal birth rates in their

second delivery approximating those of primiparous women. Encouragement of

trial of labour among these women especially is warranted.

Acknowledgements

This research was supported in part by grant MCJ-4093 from the Maternal and

Child Health Bureau (Title V, Social Security Act), Health Resources and Services

Administration, US Department of Health and Human Services.



Sup 1-8 Disc

References

1 Centers for Disease Control and Prevention. Rates of cesarean delivery - United States,1993. Morbidity and Mortality Weekly Report 1995; 44:303±307.2 Norman P. Vaginal birth after caesarean section (commentary). Lancet 1995; 345:142.3 Committee on Obstetrics: Maternal-fetal Medicine. Statement on Vaginal Birth after Cesarean(VBAC). Washington, DC: American College of Obstetricians and Gynecologists, 1988.4 Pitkin RM. Once a cesarean? (editorial) Obstetrics and Gynecology 1991; 77:939.5 Public Health Service. Healthy People 2000: National Health Promotion and DiseasePrevention Objectives. Washington, DC: US Department of Health and Human Services,Public Health Service 1991; DHHS publication no. (PHS)91±50212.6 Paterson CM, Saunders NJ. Mode of delivery after one caesarean section: audit ofcurrent practice in a health region. British Medical Journal 1991; 303:818±821.7 Jakobi P, Weissman A, Peretz BA, Hocherman I. Evaluation of prognostic factors forvaginal delivery after cesarean section. Journal of Reproductive Medicine 1993; 38:729±733.8 Kirk EP, Doyle KA, Leigh J, Garrard ML. Vaginal birth after cesarean or repeat cesareansection: Medical risks or social realities? American Journal of Obstetrics and Gynecology 1990;162:1398±1405.9 Targett C. Caesarean section and trial of scar. Australian and New Zealand Journal ofObstetrics and Gynaecology 1988; 28:249±262.10 Flamm BL, Newman LA, Thomas SJ, Fallon D, Yoshida MM. Vaginal birth aftercesarean delivery: Results of a 5-year multicenter collaborative study. Obstetrics andGynecology 1990; 76:750±754.11 Jarrell MA, Ashmead GG, Mann LI. Vaginal delivery after cesarean section: A five-yearstudy. Obstetrics and Gynecology 1985; 65:628±632.12 Ollendorff DA, Goldberg JM, Minogue JP, Min D, Socol ML. Vaginal birth aftercesarean section for arrest of labor: Is success determined by maximum cervical dilatationduring the prior labor? American Journal of Obstetrics and Gynecology 1988; 159:636±639.13 Davies JA, Spencer JAD. Trial of scar. British Journal of Hospital Medicine 1988; 40:379±381.14 Hadley CB, Mennuti MT, Gabbe SG. An evaluation of the relative risks of a trial oflabor vs. elective repeat cesarean section. American Journal of Perinatology 1986; 3:107±114.15 Paul RH, Phelan JP, Yeh S. Trial of labor in the patient with a prior cesarean birth.American Journal of Obstetrics and Gynecology 1985; 151:297±304.16 Clark SL, Eglinton GS, Beall M, Phelan JP. Effect of indication of previous cesareansection on subsequent delivery outcome in patients undergoing a trial of labor. Journal ofReproductive Medicine 1984; 29:22±25.17 Meier PR, Porreco RP. Trial of labor following cesarean section: A two-year experience.American Journal of Obstetrics and Gynecology 1982; 144:671±678.18 Herman AA, McCarthy BJ, Bakewell JM, Ward RH, Mueller BA, Maconochie NE et al.Data linkage methods used in maternally-linked birth and infant death surveillance datasets from the United States (Georgia, Missouri, Utah and Washington State), Israel,Norway, Scotland and Western Australia. Paediatric and Perinatal Epidemiology 1997;11(Suppl. 1), 5±22.19 Dean AG, Dean JA, Burton AH, Dicker RC. Epi Info, Version 5: a Word Processing,Database, and Statistics Program for Epidemiology on Microcomputers. Stone Mountain,Georgia: USD, Incorporated, 1990.20 Parrish KM, Holt VL, Connell FA, Williams B, LoGerfo JP. Variations in the accuracy ofobstetric procedures and diagnoses on birth records in Washington State, 1989. AmericanJournal of Epidemiology 1993; 138:119±127.



Sup 1-8 Disc

21 Joseph GF, Stedman CM, Robichaux AG. Vaginal birth after cesarean section: Theimpact of patient resistance to a trial of labor. American Journal of Obstetrics and Gynecology1991; 164:1441±1447.22 McClain CS. The making of a medical tradition: Vaginal birth after cesarean. SocialScience and Medicine 1990; 31:203±210.23 Kline J, Arias F. Analysis of factors determining the selection of repeated cesareansection or trial of labor in patients with histories of prior cesarean delivery. Journal ofReproductive Medicine 1993; 38:289±292.



Sup 1-8 Disc

attempt and success rates for vaginal birth after caesarean section in relation to complications of...

Documents