OBSTETRICS

OBSTETRICS

Comparing Rates of Trial of Labour Attempts, VBAC Success, and Fetal and Maternal Complications Among Family Physicians and Obstetricians
Balbina Russillo, MD, CCFP, FCFP,1,2,3 Maida J. Sewitch, PhD,1,4 Linda Cardinal, HRA,5 Normand Brassard, MD, FRCPC, MBA6,7
1 2 3 4 5 6 7

Department of Family Medicine, McGill University, Montreal QC Department of Obstetrics and Gynecology, St. Marys Hospital Center, Montreal QC Obstetrics Coordinator of the Family Medicine Unit, St. Marys Hospital Center, Montreal QC Department of Clinical Epidemiology and Community Studies, St. Marys Hospital Center, Montreal QC Quality Assessment Analyst, St. Marys Hospital Center, Montreal QC Department of Obstetrics and Gynecology Laval University, Quebec QC Obstetrician-gynecologist in Chief, Centre Hospitalier, Universitaire de Qubec, Qubec (Qubec)

Abstract
Objectives: To determine differences between family physicians and obstetricians in rates of trial of labour (TOL) attempt, vaginal birth after Caesarean section (VBAC) success, and maternal-fetal complications. Methods: We undertook a database evaluation study in an urban Quebec secondary care hospital centre that serves a multiethnic population. Study subjects were pregnant women with at least one previous Caesarean section (CS), who delivered singletons at St. Marys Hospital Center between January 1995 and December 2003. Outcomes were rates of TOL attempt, of VBAC success and failure, and of uterine rupture or dehiscence. Results: Of 32 500 singleton deliveries, 3694 (11.4%) women met study criteria. Of these, 3493 (94.6%) were patients of obstetricians, and 201 (5.4%) were patients of family physicians. The TOL attempt rate was 50.6% (1768) and 81.1% (163) for obstetricians and family physicians, respectively ( P < 0.001). For women having TOL, the VBAC success rate was 64.3% for obstetricians and 76.1% for family physicians (P = 0.002). Rates of uterine rupture or dehiscence in the combined failed and successful VBAC groups were 2.9% for obstetricians and 4.3% for family physicians (P = 0.33) whereas in the failed VBAC group the rates were 7.9% versus 17.9% for the family physicians (P = 0.04). Within delivery outcomes for successful and failed VBAC there were no differences in maternal characteristics and newborn outcomes by physician group.

Conclusion: More patients of family physicians than of obstetricians attempted TOL and had successful VBAC. Newborn outcomes were similar in the two groups, except that in the failed VBAC group, the family doctors had slightly higher uterine rupture or dehiscence rates; given the low power of this study, further studies are needed to confirm and explain this result. Also, given the similarity in patient profiles, the differences in delivery outcomes may be attributable to differences in physician practice styles.

Rsum
Objectifs : Identifier les diffrences entre les mdecins de famille et les obsttriciens en matire de taux dessai de travail (EDT), de taux de russite de laccouchement vaginal aprs une csarienne (AVAC) et de taux de complications materno-ftales. Mthodes : Nous avons men une tude dvaluation de base de donnes au sein dun centre hospitalier de soins secondaires urbain qubcois qui dessert une population multiethnique. Cette tude portait sur les femmes enceintes qui, ayant dj connu au moins une csarienne, avaient accouch dun enfant unique au St. Marys Hospital Center entre janvier 1995 et dcembre 2003. Parmi les critres dvaluation, on trouvait le taux dEDT, les taux de russite et dchec de lAVAC et le taux de rupture ou de dhiscence utrine. Rsultats : Dans le cadre de notre tude, 3 694 (11,4 %) des 32 500 femmes ayant connu un accouchement simple ont satisfait aux critres de slection. Parmi celles-ci, 3 493 (94,6 %) taient des patientes dobsttriciens et 201 (5,4 %) taient des patientes de mdecins de famille. Les taux dEDT taient de 50,6 % (1 768) dans le cas des obsttriciens et de 81,1 % (163) dans celui des mdecins de famille (P < 0,001). Chez les femmes tentant un EDT, le taux de russite de lAVAC tait de 64,3 % pour les obsttriciens et de 76,1 % pour les mdecins de famille (P = 0,002). Les taux de rupture ou de dhiscence utrine totaux (combinaison des rsultats des groupes chec de lAVAC et russite de lAVAC ) taient de 2,9 % pour les obsttriciens et

Key Words: Caesarean section, trial of labour, vaginal birth after previous Caesarean section, uterine rupture, uterine dehiscence Competing Interests: None declared. Received on July 18, 2007 Accepted on September 21, 2007

FEBRUARY JOGC FVRIER 2008 l

123

OBSTETRICS

de 4,3 % pour les mdecins de famille (P = 0,33), alors que dans le groupe chec de lAVAC , les taux taient de 7,9 % pour les obsttriciens et de 17,9 % pour les mdecins de famille (P = 0,04). En ce qui concerne les issues daccouchement au sein des groupes chec de lAVAC et russite de lAVAC , aucune diffrence na t constate entre les obsttriciens et les mdecins de famille en matire de caractristiques maternelles et dissues nonatales. Conclusion : Un plus grand nombre de patientes de mdecins de famille, par comparaison avec les patientes dobsttriciens, ont tent un EDT et connu un AVAC russi. Les issues nonatales taient semblables dans les deux groupes; cependant, dans le groupe chec de lAVAC , les mdecins de famille ont connu des taux lgrement suprieurs de rupture ou de dhiscence utrine. Compte tenu de la faible envergure de cette tude, dautres tudes savrent requises pour confirmer et expliquer ces rsultats. De plus, compte tenu de la similarit des profils de patiente, les diffrences en matire dissues daccouchement pourraient tre attribuables des diffrences en ce qui concerne les styles de pratique des mdecins en question. J Obstet Gynaecol Can 2008;30(2):123128

INTRODUCTION

82% of trials of labour after previous CS result in successful vaginal birth. Evidence suggests that family physicians and obstetricians take different approaches to the management of labour and delivery of women with previous CS.10,11 For example, comparison of the intrapartum management by family physicians and obstetricians shows that family physicians intervene less often during labour without adversely affecting maternal or fetal outcome.12 When family physicians do intervene, for example with vacuum-assisted deliveries, their complication rates are similar to those of obstetricians.13 However, it is not clear whether the management of VBAC deliveries differs according to physician speciality. The purpose of this study was to determine whether there were differences in rates of VBAC success, trials of labour, and fetal and maternal complications between family practitioners and obstetricians and if there were differences, how they could be explained and how management of these patients should be changed to decrease morbidity and mortality.
METHODS

he management of women with previous CS has long been subject to debate. The increased rate of uterine rupture and the subsequent concern for maternal and perinatal morbidity have challenged the safety of vaginal births after previous CS. In general, physicians in Europe, Asia, and Africa are more inclined than those in the US and Canada to attempt a trial of vaginal delivery.1 The overall rate of CS in the US has risen from 5% in 1970 to a high of 26% in 2002, and in Canada from 6% in 1970 to 21.2% in 20002001.2,3

In 1981, concerned with the rising rate of CS, the US National Institutes of Health Consensus Development Task Force recommended that properly selected women should be encouraged to labour and deliver vaginally after a prior CS.4 By the end of the 1980s, the rate of VBAC had risen, reaching a peak of 28.3% in 1996, with a CS rate of approximately 20%.5 However, after 1997, the VBAC rate steadily decreased to 10.6% in 2003, and the CS rate rose to approximately 27%, partly because of rising medical-legal claims from adverse outcomes.6 Although the most frequent indications for CS are previous CS, dystocia, malpresentation, and non-reassuring fetal status, the practice of repeat CS nonetheless exerts a major influence on the overall increase in CS rate7,8; repeat CS accounted for 39% of all CS in 2001.9 An estimated 60% to

The data source for this cross-sectional study was the labour and delivery database of the St. Marys Hospital Center, a secondary care, urban hospital in Montreal serving a large multiethnic population, with approximately 3500 deliveries per year. The labour and delivery database is constructed from information recorded on standardized delivery forms that are part of the patients charts, and contains information on all deliveries performed at the hospital from 1993 to the present. The study included 13 family medicine physicians with obstetrical privileges (excluding performance of CS) and 30 obstetricians. The obstetricians take 24-hour in-house call and are available for emergency calls. The family physicians also have an on-call system but are not in-house 24 hours a day. Access to emergency CS and support for patients in labour was the same for both physician groups. The study included all pregnant women who had at least one previous CS and who had a singleton pregnancy (birth weight at least 500 g) at St. Marys Hospital Center between January 1995 and December 2003. We extracted the following maternal and neonatal data from the database: gestational age by ultrasound, date of admission to caseroom, patient date of birth, obstetric history, diabetes in pregnancy, hypertension, admitted with spontaneous or induced labour, indication for induction, rupture of membranes, labour duration, augmentation, fetal distress, fever in labour, antibiotics in labour, complications, number and rank of fetus, number of previous CS, presentation, CS indication (primary or failed VBAC), reason for repeat CS, maternal hemorrhage, live or stillbirth, birth weight,

ABBREVIATIONS
CS TOL VBAC Caesarean section trial of labour vaginal birth after Caesarean section

124

l FEBRUARY JOGC FVRIER 2008

Comparing Rates of Trial of Labour Attempts, VBAC Success, and Fetal and Maternal Complications Among Family Physicians and Obstetricians

Table 1. Summary of main outcomes

Deliveries after previous CS 19952003 (N = 3694) Obstetricians Rates TOL VBAC success Repeat CS, no TOL Total rupture or dehiscence Rupture or dehiscence in failed and successful VBAC groups n = 3493 1768 1136 1725 92 51 (94.6%) (50.6) (64.3) (49.4) (2.6) (2.9) Family physicians n = 201 163 124 38 8 7 (5.4%) (81.1) (76.1) (18.9) (4.0) (4.3) P < 0.001 0.002 < 0.001 0.25 0.33

suspected neonatal anomalies, cord pH value, respiratory depression, Apgar score, and time of delivery.
Statistical Analysis

Descriptive statistics were used to characterize study subjects. Subjects were compared using Student t tests, chi-square tests, and Fisher exact tests, as appropriate. The four study outcomes were defined as follows: (1) The TOL rate was equal to the number of women attempting vaginal delivery among women with prior CS divided by all women with prior CS; (2) The VBAC success rate was equal to the number of VBACs divided by the number of women with TOL; (3) The VBAC failure rate was equal to the number of CS divided by the number of women undergoing TOL; (4) The VBAC uterine rupture or dehiscence rate was equal to the number of uterine ruptures and dehiscences divided by the number of women with prior CS.
RESULTS

(13.2% vs. 4.2%, P = 0.02), and hypertension (7.9% vs. 2.4%, P = 0.07) and a lower proportion of patients between 37 and 41 weeks gestation (75% vs. 90%) than the obstetricians. Family physicians had a higher percentage of medical indications as the primary reason for repeat CS than the obstetrician group (57.9% vs. 40.9%). Maternal and neonatal complications were similar in the two physician groups. Table 2 shows the maternal characteristics according to VBAC outcome. For the successful VBAC group, family physicians had a higher percentage of patients with diabetes than obstetricians (11.3% vs. 4.7%, P = 0.002). Induction of labour rates were similar for the family physician and the obstetrician groups (Table 3). There were no significant differences in augmentation of labour and complications (e.g., fever or chorioamnionitis) in the two groups. Mean birth weight, Apgar scores, and cord pH in babies whose mothers had a successful VBAC were comparable between physician groups. However, the proportion of cord pH < 7.2 was higher for the family physician group than for the obstetrician group (17.4% vs. 7.8%, P = 0.05) (Table 5). Maternal characteristics and intrapartum characteristics in the failed VBAC group were comparable between physician groups. However, complications following a failed VBAC were different: the uterine rupture or dehiscence rate was 7.9% for the obstetricians and 17.9% for the family physicians (P = 0.04). Rates of fever or chorioamnionitis as well as newborn characteristics were comparable (Tables 4 and 5). Of the 15 stillbirths recorded, 11 (73%) were intrauterine fetal deaths occurring prior to admission, and four (27%) occurred after admission to hospital (2 fetal deaths occurred during labour).
FEBRUARY JOGC FVRIER 2008 l 125

A total of 3694 pregnancies met study criteria. Of these, 3493 (94.6%) were delivered by obstetricians, and 201 (5.4%) were delivered by family physicians (Table 1). Of the obstetrician group, 50.6% (1768) had a trial of labour compared with 81.1% (163) of the family physician group (P < 0.001). The VBAC success rate was 64.3% (1136) in the obstetrician group compared with 76.1% (124) in the family physician group (P = 0.002). Conversely, the VBAC failure rate was 35.7% (632) for obstetricians versus 23.9% (39) for family physicians (P = 0.002). The uterine rupture or dehiscence rate in the failed and successful VBAC group was 2.9% (51) for obstetricians versus 4.3% (7) for family physicians (P = 0.33). For patients who underwent a repeat CS without TOL, family physicians had a greater proportion of patients with a gestational age < 37 weeks (16.7% vs. 6.2%), diabetes

OBSTETRICS

Table 2. Maternal characteristics

Successful VBAC Characteristics Maternal age (mean) Maternal comorbidity Diabetes Hypertension Gestational age < 37 weeks 3741 weeks > 41 weeks No. of previous CS 1 2 3 or more 96.0% 3.6% 0.4% 99.2% 0.8% 0.0% 4.7% 83.0% 12.3% 5.8% 82.5% 11.7% 0.21 94.3% 5.4% 0.3% 94.9% 5.1% 0.0% 4.7% 2.9% 11.3% 4.0% 0.002 0.41 0.84 5.3% 72.5% 22.2% 5.3% 65.8% 28.9% 0.99 8.45% 3.6% 5.1% 2.6% 0.76 0.99 0.62 Obstetricians 31.4 Family physicians 30.1 P 0.002 Obstetricians 32.2 Failed VBAC Family physicians 31.2 P 0.14

Table 3. Intrapartum characteristics

Successful VBAC Obstetricians Characteristics Induction Oxytocin Prostaglandin Amniotomy TOTAL Augmentation of labour 1st stage 2nd stage TOTAL 29.4 4.8 34.2 30.8 6.7 37.5 0.47 18.8 1.2 1.4 21.4 20.2 0.0 2.4 22.6 0.76 0.60 28.0 1.8 29.8 40.5 2.7 43.2 0.08 % Family physicians % P 0.52 32.6 2.4 0.5 35.4 33.3 2.6 0.0 35.9 0.95 0.23 Obstetricians % Failed VBAC Family physicians % P 0.99

DISCUSSION

In this study, family physicians had a higher TOL attempt rate than obstetricians; this was not explained by patient profiles, which were similar in terms of the maternal age, gestational age, number of previous CS, and maternal comorbidities (diabetes and hypertension). The one exception was that family physicians had a higher percentage of patients with gestational diabetes in the successful VBAC group. Although comparable percentages of patients in each physician group were eligible for a TOL, VBAC refusal rates were 35.9% for obstetricians and 23.7% for family physicians. The higher rate of patient refusal in the obstetrician group may have resulted from the way in which physicians counselled patients on TOL. Our database did
126 l FEBRUARY JOGC FVRIER 2008

not provide detailed descriptions of counselling by physicians. To avoid physician bias, a well-defined protocol for TOL, as used in a study by Gonen et al.,14 can be used to select all eligible patients for specific counselling, quoting success and failure rates. This would be feasible in a future study. Family physicians had a higher VBAC success rate than obstetricians. Current studies show that approximately 60% to 82% of TOL after previous CS result in successful vaginal delivery.1 The observed VBAC success rate among family physicians is consistent with the pooled vaginal delivery rate obtained in prospective studies (76%).6 Given the comparable patient profiles, it is possible that the higher VBAC success rate among family physicians was due to

Comparing Rates of Trial of Labour Attempts, VBAC Success, and Fetal and Maternal Complications Among Family Physicians and Obstetricians

Table 4. Complications rates

Successful VBAC Family physicians % 0.0 4.0 P 0.99 0.88 Failed VBAC Family physicians % 17.9 12.8 P 0.04 0.10 Repeat CS, no TOL Family physicians % 2.6 2.6 P 0.60 0.49

Obstetricians Complication Uterine rupture or dehiscence Fever/ chorioamnionitis % 0.1 4.3

Obstetricians % 7.9 6.0

Obstetricians % 2.4 1.7

Table 5. Fetal characteristics

Successful VBAC Obstetricians Characteristics Birth weight < 2500 g 25004000 g > 4000 g Apgar score 3 at 1 min 6 at 5 min Cord pH < 7.2 7.8 17.4 0.05 8.6 3.3 0.50 3.0 2.6 2.4 2.4 0.99 0.99 3.0 2.2 0.0 0.0 0.62 0.99 3.3 87.5 9.2 2.4 87.1 10.5 % Family physicians % P 0.81 2.2 81.6 16.1 2.6 87.2 10.3 Obstetricians % Failed VBAC Family physicians % P 0.62

different physician management approaches or styles during labour; intraspecialty differences in the way providers treat similar conditions are well documented.15,16 Although family physicians may be less interventionist than obstetricians, they also do not have the option to perform CS. Of patients who underwent a repeat CS without TOL, a higher percentage of those in the family physician group than in the obstetrician group had gestational diabetes. Family physicians had a higher percentage of patients who underwent a repeat CS for medical reasons (e.g., history of myomectomy, diabetes, history of cephalic-pelvic disproportion, macrosomia, placenta previa, etc.) versus performing a repeat CS because of patient wishes. Again, given the similar patient profiles, physician style of practice may have influenced the patients decision. The uterine rupture or dehiscence rate for patients in the obstetrician group was slightly lower than the rate for those in the family physician group. In the failed VBAC group, the uterine rupture or dehiscence rate was also lower for the

obstetrician group and statistically significant (7.9% vs. 17.9% P = 0.04). Because family physicians do not perform CS, this could have affected their management and may have prolonged the time before the decision to perform a CS; this in turn may explain the increased morbidity in terms of uterine rupture or dehiscence in the family physician group. Unclear definitions for rupture and dehiscence in our database made it impossible for us to study the two rates separately, and the morbidity of uterine rupture could not be isolated from that of uterine dehiscence. Nevertheless, this finding is of concern and requires further study. In the successful VBAC group, the percentage of babies with a cord pH < 7.2 was statistically significantly higher in the family physician group than in the obstetrician group (P = 0.05). The clinical significance of this observation is difficult to interpret because our data did not allow us to further subdivide the cords with pH < 7.0, which would have provided a better predictor of fetal compromise. Our data should be interpreted in light of these study limitations. The labour and delivery database was validated
FEBRUARY JOGC FVRIER 2008 l 127

OBSTETRICS

against the hospital database and some medical charts for variables including delivery type, previous CS, single birth, stillbirth, uterine rupture, and birth weight. However, a validation of the entire database was not attempted. Certain risk factors that are known today but not at the time of our study, such as a one-layer interlocking closure,17 interdelivery interval (within 1824 months),18 and thickness of the lower uterine segment,19 might have influenced or explained certain outcomes. Although these more recently identified risk factors will help physicians to better select appropriate candidates for TOL in future, we had no knowledge of this information and no such data were recorded in the hospital database. Also, the reasons given by patients and physicians for repeat CS and for refusal of TOL may not reflect the true reasons; details of counselling discussions were not available on our database, and there was no standard counselling guide or questionnaire that all staff used.
CONCLUSION

REFERENCES
1. Biswas. A. Management of previous Cesarean section. Curr Opin Obstet Gynecol 2003;15:123201. 2. Harer WB. Vaginal birth after Cesarean deliveries. JAMA 2002;287(20):229. 3. Martel M, Mackinnon C. Guidelines for vaginal birth after previous Caesarean birth. SOGC Clinical Practice Guidelines, No. 155, Feb 2005. J Obstet Gynaecol Can 2005;27:16474. 4. Health Canada, Canadian Perinatal Health Report 2003. Ottawa (ON): Health Canada;2003:33. 5. United States Department of Health & Human Service, Public Health Services. Healthy People 2000. National Health Promotion & Disease Prevention Objectives. Washington, DC: Department of Health & Human Services. Pub N0.1990; 9150212. 6. Vaginal Birth After Cesarean (V.B.A.C). Agency of Health Care Research & Quality. Rockville MD: AMRQ Publication No. 03-E018. March 2003. 7. Macdonald S, Voaklander K, Birtwhistle R. Comparison of family physicians& obstetricians intrapartum management of low risk pregnancies. J Fam Pract 1993;37(5):45762. 8. Rosenblatt R, Dodie S, Hart G, Schneeweiss R, Gould D, Raine TR, et al. Interspeciality differences in the obstetric care of low risk women. Am J Public Health 1997;87(3):34451. 9. Chauhan S, Martin J, Henrichs C, Morrison J, Magann E. Maternal & perinatal complication with uterine rupture in 142,075 patients who attempted vaginal birth after Caesarean delivery: a review of the literature. Am J Obstet Gynecol 2003;189(2):40816. 10. Berman D, Johnson T, Apgar B, Schwenk T. Model of family medicine and obstetrics-gynecology in obstetric care at the University of Michigan. Obstet Gynecol 2000;96(2):308313. 11. Klein M. The Canadian family practice accoucheur. Can Fam Physician 1986;32:53340. 12. Meyer B. Audit of obstetrical care: comparison between family practitioners and obstetricians. Fam Pract Res J 1981;1:207. 13. Yarrow C, Benoit G, Klein M. Outcomes after vacuumassisted deliveries. Can Fam Physician 2004; 50:110916. 14. Gonen R, Nisenblat V, Barak S, Tamir A, Ohel G. Results of a well-defined protocol for a Trial of Labor After Prior Cesarean Delivery. Obstet Gynecol 2006;107:2405. 15. Eisenberg JM. Physician utilization: the state of research about physicians practice patterns. Med Care 1985;23:46183. 16. Schroeder SA. Strategies for reducing medical costs by changing physicians behavior and impact on quality of care. Int J Technol Assess Health Care 1987;3:3950. 17. Lalonde AB, Senikas V. Is VBAC a viable option for Canadian Women? J Obstet Gynaecol Can 2005;27(2):16374. 18. Bujold E, Bujold C, Hamilton EF, Harel F, Gauthier R. The impact of a single-layer or double-layer closure on uterine rupture. Am J Obstet Gynecol 2002; 186:132630. 19. Cheung VY. Sonographic measurement of the lower uterine segment thickness in women with previous Caesarean section. J Obstet Gynaecol Can 2005;27(7):67481.

In our hospital labour and delivery database study, family physicians had higher rates of TOL and of successful VBAC than obstetricians. Given the similarity in patient profiles, these findings may be the result of different practice styles between the physician specialties. Rates of maternal and neonatal complications between the physician groups were comparable for all categories except the failed VBAC group, in which family physicians had a higher uterine rupture or dehiscence rate than obstetricians. Future research is needed that calculates the rates of uterine rupture and dehiscence separately in order to increase our understanding of maternal morbidity. The clinical implications of our findings are (1) that all pregnant women who are eligible for TOL should be informed, in a standardized way, of the advantages and disadvantages of each mode of delivery and (2) that criteria for intrapartum management must be applied to improve the VBAC success rate without increasing morbidity. For all treating physicians, it may be possible to increase VBAC delivery rates without increasing maternal and fetal morbidity and mortality.
ACKNOWLEDGMENTS

This research was supported by St. Marys Hospital Center. Maida J. Sewitch, PhD is a Research Scientist of the Canadian Cancer Society through an award from the National Cancer Institute of Canada.

128

l FEBRUARY JOGC FVRIER 2008