Uterine rupture and labour after a previous low transverse caesarean section
Dr M Turner, Department of Obstetrics and Gynaecology, Coombe Women’s Hospital, Dublin 8, Ireland. Email email@example.com
This study reviewed all cases of complete uterine rupture (UR) in pregnancy during the decade 1993–2002. In 69 412 deliveries, 5320 women had a single prior caesarean section. Of these, 4021 had a trial of labour and 3129 (77.8%) delivered vaginally. In nine (0.22%) cases, the previous transverse scar ruptured during labour. None of these nine ruptures resulted in maternal or fetal death, peripartum hysterectomy or fetal morbidity. In our practice, a trial of labour in women with a previous low transverse caesarean is associated with a high rate of vaginal delivery and a low rate of UR.
Complete uterine rupture (UR) can be an obstetric catastrophe, which causes maternal and fetal mortality and morbidity worldwide.1 Nowadays, in the developed world, it is strongly associated with a history of low transverse caesarean delivery.1
In the USA, the caesarean birth rate increased from 5.5% in 1970 to 15.2% in 1978. This dramatic change in practice prompted the National Institutes of Health (NIH) to hold a consensus conference on caesarean childbirth in September 1980. Prior caesarean delivery was one of the two major reasons cited for caesarean birth, and it accounted for 25–30% of the increase in caesarean rate between 1970 and 1978. The conference concluded that vaginal birth after caesarean (VBAC) delivery was an appropriate way to reverse the trend in the increasing caesarean rates. The NIH recommendations were implemented, and although the overall caesarean rate failed to fall, the overall VBAC rate did increase from 3.0% in 1980 to 28.3% in 1996.2
In a population-based study in Nova Scotia of 6138 women with a previous caesarean reported in 1996, a trial of labour was compared with an elective second caesarean.3 There were ten cases of UR, all within the trial-of-labour group. Including the ruptures, the major maternal complications were twice as likely among those deliveries managed with a trial of labour as among those that involved an elective second caesarean section.
In another population-based study of 20 095 women in Washington State, the authors concluded that for women with one prior caesarean delivery, the risk of UR was higher among those whose labour was induced than among those with repeated caesarean delivery without labour.4
A large prospective American study of women with a singleton gestation and a prior caesarean delivery was conducted at 19 academic medical centres.5 The authors concluded that a trial of labour after prior caesarean was associated with a greater perinatal risk than elective repeat caesarean, although absolute risks were low. There is now emerging evidence that these publications in the influential New England Journal of Medicine are contributing to a reversal of the trend in VBAC rates.2 Caesarean deliveries are now at the highest level ever reported in US history.2
Previously, we reported that a trial of labour in women with a previous low transverse caesarean section was associated with a low incidence of UR and that it was also associated with a high incidence of VBAC in labour. In the light of the recent publications and trends,2–5 we decided to audit our results to determine whether our obstetric practice is safe for both mother and baby.
The maternity services in the Republic of Ireland are highly centralised. About one in seven babies nationally are born in our hospital, which is a university teaching hospital that also acts as a tertiary referral centre. The obstetric services are specialist led, but normal care is provided primarily by midwives. The patients cross all social classes, with approximately one-third attending their obstetrician as a private patient.
At all times, there is a trainee obstetrician, a neonatologist and an anaesthetist in our hospital, and a senior obstetrician, neonatologist and anaesthetist is available either in the Hospital or from home.
Women with one previous caesarean section, in general, are encouraged to aim for a vaginal birth. Antenatally, an ultrasound examination is recommended to date the pregnancy and to rule out abnormal placentation. Induction of labour is advocated only for obstetric indications in selected women. Women with a previous caesarean section have a colour-coded partogram to differentiate them from those who labour without a uterine scar. The decision to administer oxytocin to augment labour is reserved for senior obstetricians who have personally reviewed the woman clinically in labour. Routinely, the fetal heart rate is monitored electronically and, if indicated, fetal blood sampling for pH status is used. A previous caesarean section is not considered a contraindication to epidural analgesia. Once a decision has been made to perform an emergency caesarean section, the aim is a short decision to delivery interval.
The study was confined to all women who delivered a baby weighing 500 g or more in the decade 1993–2002 inclusive. All deliveries were computerised in the labour ward by the midwifery staff. We also examined the obstetric and neonatal case records of all women who had a complete UR.
Complete rupture was defined as involving the full thickness of the uterine wall, including the visceral peritoneum.1,2 Dehiscence of a previous uterine scar occurs if the peritoneum remains intact.1,2 Women with an asymptomatic scar dehiscence found at elective caesarean section were excluded because labour is not an issue and a dehiscence has little or no clinical significance.
During this study, 68 443 mothers delivered 69 412 babies weighing 500 g or more. The incidence of caesarean section increased from 12.9% in 1993 to 18.9% in 2002, and the overall incidence was 15.5%. The number of women with a history of a previous single caesarean section was 5320, and of these 1299 (24.4%) had a repeat elective caesarean.
There were a total of 14 cases of complete UR. Two cases occurred in women without a known uterine scar, and one occurred preterm and prelabour in a woman with a bicornuate uterus and a previous classical caesarean section. In women with a history of a previous transverse caesarean, there were 11 URs.
Of the 11 cases, there were 2 where the rupture occurred at a site other than the caesarean scar, which remained intact. In the first case, the rupture occurred prelabour at the fundus when the woman was 25 weeks1 pregnant and the baby was stillborn. In the second case, the rupture occurred posteriorly and was diagnosed at emergency caesarean section in labour for fetal distress. The baby has been diagnosed with hypoxic ischaemic encephalopathy. In both cases, the woman had a previous uterine curettage, raising the possibility that the rupture occurred at the site of a previous unsuspected uterine perforation.
Therefore, there were nine cases in 4021 women where labour resulted in a rupture of a previous transverse caesarean scar. None of the nine cases resulted in a maternal death, peripartum hysterectomy or serious maternal morbidity. Three women received a blood transfusion (range 2–8 units).
None of the nine cases resulted in a perinatal death or hypoxic ischaemic encephalopathy. None had an Apgar score of less than 7 at 5 minutes. Seven babies had a cord pH performed, and in three of these, the pH was less than 7.20.
Of the 4021 women with a previous single transverse caesarean allowed to labour, 3129 (77.8%) delivered vaginally. Of the 1776 women with a previous caesarean and a previous vaginal delivery allowed to labour, 1587 (89.4%) delivered vaginally and the caesarean scar ruptured in 3 (0.17%) cases. In the 2245 women with a previous caesarean and no previous vaginal delivery allowed to labour, 1542 (68.7%) delivered vaginally and the caesarean scar ruptured in 6 cases (0.27%).
Of the 4021 women allowed to labour, 69 (1.7%) received vaginal prostaglandins to induce labour, 352 (8.8%) received oxytocin to induce labour and 1024 (25.5%) received oxytocin to augment labour.
Of the nine cases of caesarean scar rupture, four had oxytocin to induce labour (one woman also received prostaglandins) and two had oxytocin to augment labour. Eight women used epidural analgesia, and in six of the eight, the rupture was associated with vaginal bleeding.
We also reviewed all perinatal deaths during the study period. A policy of repeat elective caesarean potentially would have prevented three stillbirths of normally formed infants, which occurred antepartum after 39 weeks in the 4021 women allowed a trial of labour.
This audit confirmed that women with one previous low transverse caesarean section can achieve a high rate of VBAC overall and a high rate of VBAC in labour. These high rates were associated with a low risk of rupture of the caesarean scar in labour. Even when the scar ruptured, major complications for mother or baby were few.
If, during the study, we had adopted a policy of repeat elective caesarean section at term in women with one previous caesarean section, we would not have prevented any perinatal death due to labour. We might have prevented a single case of hypoxic ischaemic encephalopathy, but this complication did not result from rupture of the caesarean scar. Moreover, encephalopathy can occur after any labour irrespective of the obstetric history. We may have prevented three antepartum stillbirths, but logically, elective caesarean at term should be considered in all pregnancies if we wish to prevent all stillbirths after 39 weeks of gestation.
If we had implemented a policy of repeat elective caesarean section, we would have performed an additional 3129 repeat caesarean sections. We would also have committed these women to third, fourth and fifth caesarean sections, etc. if they were to have more children in the future. Any increase in the incidence of repeat elective caesarean section at term has also a potential downside for the baby. The incidence of neonatal respiratory problems, for example, increases after elective caesarean section compared with that after a vaginal delivery, irrespective of gestation.
Revisiting the Nova Scotia study,3 the women attended 27 different hospitals, including community hospitals. Some data were missing and 319 women were excluded. The cases were coded by records personnel, not by midwives, and there was no chart review of the ten ruptures. Thus, no details was available of the previous uterine scar, the use of oxytocics and the women’s previous obstetric records. The overall VBAC rate was low at 32%, and the VBAC rate in labour was 60%.
Revisiting the Washington State study,4 there was no chart review of the 91 URs. The authors could not document the accuracy of the coding, and cases of asymptomatic dehiscence or intraoperative extensions of caesarean wounds may have been included. There were no details available of the use of oxytocin to augment labour, independent of induction. Importantly, all women with a prior vaginal delivery were excluded. A prior vaginal delivery improves the outcome of both spontaneous and induced labour after prior caesarean section.1
The American multicentre observational study5 compared the outcome in 15 801 women who underwent elective repeat caesarean section without labour with the outcome in 17 898 women who attempted vaginal delivery after one prior caesarean section. The trial-of-labour group, however, included only women presenting in labour with cervical dilatations of at least 4 cm, as well as those receiving oxytocin. In the trial-of-labour group, there was an increased risk of hypoxic ischaemic encephalopathy, endometritis and blood transfusion. There was no difference in the frequency of maternal death and hysterectomy between the two groups.
This study, however, included women who had a previous vertical incision in the uterus, and the perinatal and maternal outcomes were not analysed according to the type of previous uterine scar. This is surprising because the maternal and perinatal risks are higher with a vertical scar, and the consensus is that a trial of labour is contraindicated in such cases.1 There were also significant differences between the two groups. The trial-of-labour group had a higher incidence of younger women, single women, smokers, growth-restricted babies, preterm deliveries and women without private insurance. This may explain, for example, why women scheduled for a trial of labour had a higher incidence of antepartum stillbirth than those scheduled for elective repeat caesarean section. In the 114 term pregnancies with UR, there were no intrapartum stillbirths. There were seven cases of hypoxic ischaemic encephalopathy. Two of the seven died neonatally. No details was provided on the circumstances of these cases or on the long-term outcome of the survivors.
Our findings in Dublin are more in keeping with European experiences. In a large Scottish study of attempted VBAC section,6 the proportion of vaginal birth was 74.2% and the incidence of UR was 0.35%. The risk of perinatal death associated with UR decreased in women who went into spontaneous labour, who had a previous vaginal delivery or who were delivered in hospitals with >3000 births per annum.
The North American studies3–5 have highlighted correctly the risks of UR in women who labour with a previous caesarean section. We are concerned, however, that obstetricians individually and collectively may have overreacted to their publication and that some have been too quick to revert back to a policy of ‘once a caesarean, always a caesarean’. This simplistic mantra may have been appropriate at the start of the 20th century, but our experience suggests that such a ‘one policy fits all’ approach may not be in the interests of both mother and baby at the start of the 21st century.
The subject of the delivery of a woman after a previous low transverse caesarean section remains controversial.2 The complexity of confounding variables and the differing clinical practices make it difficult to apply general obstetric knowledge to the care of individual patients. Ideally, best practice should be resolved using randomised controlled trials. In reality, however, such trials are difficult to conduct. UR is a rare event with unpredictable clinical outcomes, and statistically, a large number of women would be required for an adequately powered study. This would require a multicentre approach, with all the disadvantages of different clinical practices. There are also other potential problems such as informed consent, observer bias and differing patient populations. Inadequate evidence from individual studies has prompted the use of meta-analysis.2 We suggest, however, that there is still a role for audit in determining clinical practice. Indeed, it is arguable that continuing local audit in larger centres may be more relevant for developing clinical practice locally than research findings in other settings.
We consider that women with a previous low transverse caesarean section can be delivered safely, provided the obstetricians select carefully on clinical grounds those women suitable for a trial of labour. We advocate avoidance of induction of labour unless clinically necessary, close supervision of labour, oxytocin usage only by experienced clinicians and expert surgical skill being immediately available to expedite delivery should problems arise. If these provisions are met, then the safety of mother and baby may be optimised and repeat elective caesarean section minimised.
We hope that our findings will help reassure doctors and women alike that the best option for women with a single previous low transverse caesarean section considering their mode of delivery is for obstetricians to continue to practice the art of obstetrics based on good clinical practice, taking into account individual patient circumstances and the quality of local services.
We thank our midwifery colleagues for their supervision of women who laboured after a previous caesarean section and Ms Emma McNamee for her assistance in collating the computerised information.