Introduction
A recent study of 146 NHS trusts in England (Bragg et al., 2010) found that one in four women had a CS. The rise in primary CS rates has led to an increased proportion of women of reproductive age with a scarred uterus. Thus, the issue of the most appropriate mode of delivery following a CS continues to be the subject of research and debate.
This section presents the best available evidence to facilitate antenatal counselling and decision making when planning the mode of birth following one or more previous CSs.
Overview of evidence
Four studies were included in this review (Guise et al., 2010; Cahill et al., 2010; Tahseen & Griffiths, 2010; Law et al., 2010).
Of the four studies, one is a systematic review (Guise et al., 2010), one was conducted in the USA (Cahill et al., 2010), one in the UK (Tahseen & Griffiths, 2010) and one in Hong Kong (Law et al., 2010). One study is a large, rigorous systematic review of observational studies (Guise et al., 2010). One study (Tahseen & Griffiths, 2010) performed a systematic review of observational studies of success rates and adverse maternal and neonatal outcomes of vaginal birth after one or two CSs and repeat CSs.
One study (Cahill et al., 2010) reported maternal morbidity in women with three or more prior caesarean births who attempt a vaginal birth (VBAC).
One study (Law et al., 2010) reported maternal psychological status among women with one previous caesarean birth who were randomised to planned vaginal birth or planned CS.
Evidence profile
One evidence profile summarises maternal outcomes from one systematic review plus one randomised trial of the risks and benefits of “elective” repeat CSs [ERCS] compared with trial of labour [TOL] (Guise et al., 2010, Law et al., 2010). One evidence profile summarises neonatal outcomes from one systematic review of the risks and benefits of ERCS compared with TOL (Guise et al., 2010). Three evidence profiles report maternal complications associated with repeat CS as reported by the same systematic review (Guise et al., 2010). One evidence profile reports maternal outcomes of vaginal birth or planned CS after two previous CSs compared with vaginal birth or planned CS after one previous CS (Tahseen & Griffiths, 2010). Maternal morbidity in women who plan vaginal birth after three or more prior CSs is detailed in one evidence profile reporting findings from one observational study (Cahill et al., 2010). All included studies were observational studies. Therefore, using the GRADE system, the quality of the evidence was moderate, low or very low for all studies.
Maternal outcomes
All of the data included in have been taken from one systematic review and one randomised trial and details outcomes for women who have had one previous CS. The number in the first column indicates the number of studies within the review that contribute data to that outcome.
GRADE summary of findings comparing planned CS with planned vaginal birth in women with a previous CS (maternal outcomes).
Repeat CS
Narrative discussions of maternal complications associated with multiple CSs were reported in one systematic review (Guise et al., 2010). All participants gave birth by CS. The number of studies contributing to the outcome in question is reported in the first column of the table. For Tables 11.2 to 11.4 just one study from within the systematic review (not the same study) contributed data to each outcome. All included studies involved women giving birth at any gestation.
GRADE summary of findings for repeat CS (one prior CS compared with two prior CSs).
GRADE summary of findings for repeat CS (one prior CS compared with two or more prior CSs).
GRADE summary of findings for repeat CS (one prior CS compared with three prior CSs).
Vaginal birth attempt following two or more CS
All of the data included in this section have been taken from two studies (Cahill et al., 2010; Tahseen & Griffiths, 2010) that reported maternal morbidity in women who attempted VBAC. The range of successful VBACs was 74% to 80% in one observational study (Cahill et al., 2010) and 72% to 76% in the other study (Tahseen & Griffiths, 2010). For the systematic review (Tahseen & Griffiths, 2010) the number of studies reported in the first columns of and corresponds to the number of included studies contributing findings to each reported outcome.
GRADE summary of findings for planned VBAC after two prior CSs compared with planned repeat CS after two prior CSs.
GRADE summary of findings for planned VBAC after three or more prior CSs compared with planned repeat CS after three or more prior CSs.
Neonatal outcomes
All of the data included in have been taken from one systematic review (Guise et al., 2010). The number in the first column indicates the number of studies from that review which report on those outcomes.
GRADE summary of findings comparing planned CS with planned vaginal birth in women with a previous CS (neonatal outcomes).
No pooled data was reported in the systematic review (Guise et al., 2010) for neonatal intensive care unit [NICU] admission, hypoxic-ischemic encephalopathy (HIE) and neonates' Apgar score. There were narrative discussions for these outcomes which are summarised here.
There was evidence that the rate of NICU admission was higher in neonates born by planned repeat CS compared with neonates born following planned vaginal birth (eight studies, low quality, pooled data not reported).
There was very low quality evidence from three studies that reported lower rates of HIE among neonates born by planned repeat CS compared with neonates born following planned vaginal birth (pooled data not reported) .
There was low quality evidence from four studies that found no difference in the proportion of babies with an Apgar score of 7 or below at 5 minutes in neonates born by planned repeat CS compared with neonates born following a planned vaginal birth (pooled data not reported).
Health economics
A model was developed to compare the cost effectiveness of VBAC versus a planned CS in women with one previous CS and with no plans for further children. A summary of this analysis is provided below (see Chapter 13 for further details).
In addition to the costs of birth, the model also estimated ‘downstream’ costs and quality adjusted life years (QALYs) based on the risk of adverse events for each planned mode of birth. The base case analysis considered only the outcomes that were reported in the review undertaken for this guideline to determine the risks and benefits for both women and babies of planned CS compared with planned vaginal birth in women who have had a previous CS. Secondary analyses also used outcomes that were only reported in the review which compared the risk and benefits of planned CS compared with planned vaginal birth. However, it should be recognised that these risks are likely to be underestimated for this population and that the relative risk for these adverse outcomes may also be different in this population.
The results tended to show that VBAC was more likely to be cost effective, although this was a borderline finding and considerable uncertainty remains, especially with respect to all the outcomes that may differ between the different modes of planned birth in this population.
Evidence to recommendations
Relative value placed on outcomes considered
It was noted that findings from studies of babies born at term were very similar to those that included babies born at any gestational age. This was thought to reflect the relatively low numbers of preterm babies included in the studies involving all gestational ages. The extra statistical power afforded by the larger numbers where studies of all gestational ages have been included meant the GDG was happy to consider this evidence when making recommendations.
Maternal mortality is clearly a vitally important outcome at an individual level, but in terms of informing decision making for a whole population, the very low numbers of deaths reported in the studies (absolute numbers range from 1.9 to 7.5 per 100,000) mean this outcome does not necessarily drive the recommendations made. Other important outcomes, including uterine rupture and neonatal mortality, were more common, although still rare, meaning that although differences between study groups were statistically significant, the low incidence meant that they were also considered less clinically significant in terms of driving clinical practice and advice to women. In the context of low actual risk, then absolute risk will be a more important consideration than relative risk.
The reported neonatal outcomes of bag and mask ventilation and transient tachypnoea were felt to be of limited value as it was not possible to determine how these outcomes related to ongoing health problems or disability.
Trade-off between clinical benefits and harms
Mortality-related outcomes are very rare. However, the GDG noted that for women planning birth following one previous CS, maternal mortality is statistically significantly higher for women planning a repeat CS while neonatal mortality is statistically significantly higher for women planning a vaginal birth. Perinatal mortality is not statistically different between the two groups.
The GDG agreed that although it is right to give women all available information when planning mode of birth, the important thing to emphasise to women planning birth after one previous CS is that serious adverse outcomes, including maternal and neonatal mortality, uterine rupture, need for blood transfusion and hysterectomy, are rare, no matter whether women choose a planned repeat CS or VBAC.
The GDG noted that the relative risk of adverse outcomes may vary from woman to woman depending upon her obstetric history, including reasons for previous CS(s) and whether or not a woman has previously given birth vaginally. These individual considerations need to be taken into account when discussing mode of birth following one or more previous CSs.
When considering increasing numbers of previous CSs, the evidence showed no difference between planned vaginal birth and planned CS in rates of blood transfusion, fever and hysterectomy after two prior CSs. However, with an increasing number of CSs, there is an increasing risk of need for blood transfusion, wound complications and injuries to the bladder, regardless of the mode of birth.
Trade-off between net health benefits and resource use
An economic model developed for this guideline to compare the cost effectiveness of planned CS versus planned vaginal birth in women who have had a previous CS did not strongly suggest a preferred mode of birth. As a result, this model would, given the current state of evidence, support a recommendation allowing women to choose their preferred method of birth in consultation with the healthcare professionals responsible for her care. Considerations about any future pregnancies may be an important factor in the decisions made, given the increased risks in, for example, incidence of placenta praevia and morbidly adherent placenta, which are associated with repeat CSs.
Quality of evidence
The evidence for outcomes following one previous CS was drawn from one large meta-analysis of observational studies and one randomised controlled trial (RCT), and ranged from moderate to very low. The large sample sizes reported for the meta-analysis meant the GDG felt more confident in the validity of the findings regarding rare outcomes.
The evidence examining outcomes following more than one previous CS is of low and very low quality. While evidence comparing outcomes for women having a CS following one, two or more previous CSs is interesting and can be used to provide general information about increasing risks following two or more CSs, it does not help a woman decide the level of risk associated with her choice of mode of birth in the current pregnancy. However, this information is helpful for decision making about future births. The evidence comparing outcomes for women choosing a planned vaginal birth compared with those choosing a planned CS after two or more previous CSs is useful in this respect as this reflects the choice women have. Unfortunately, this evidence is of very low quality, thus lowering the validity of the reported findings.
No good quality evidence was available for women planning birth following five or more previous CSs.
Other considerations
The GDG noted that many women leave hospital following a caesarean birth without understanding the implications for planning future pregnancies and births. It was felt that it is important to provide this information to women and their partners so that they can have an accurate picture of what this means for them when planning their family, including options for future modes of birth. The GDG agreed that there is a benefit to providing this information to women and their partners prior to leaving the hospital because the medical records are easily available to refer to. As a result, the GDG recommended that this discussion take place with women after the CS. However, the GDG also recognised that some women may prefer to have this discussion at a later date, so highlighted that this discussion can be deferred. Due to the large amount of information women and their partners receive during the immediate postnatal period, this information should be provided both verbally and in written formats. It is important to emphasise to women that, regardless of future choice of mode of birth, poor outcomes are very rare.
