Like many other high-income countries, caesarean rates have increased in Australia and there is great interest in how a reduction in rates might be achieved.16–18
In our study population, the fundamental feature has been a sharp rise in caesarean delivery at first-birth, which then feeds into increased repeat caesareans. Since 1994, intrapartum caesarean deliveries have accounted for two-thirds of all first-birth caesareans. This finding is consistent with a recent US study (2003–2007) where over 60% of primary caesarean births were for labour arrest disorders or non-reassuring fetal heart rate tracing, with relative increases of 21% and 62%, respectively, for caesarean delivery in these situations.19
Other international intrapartum rates are difficult to identify. A comparative analysis of international caesarean delivery rates from single hospitals in nine countries in 2005–2006 reported intrapartum caesarean rates for nulliparae with singleton, cephalic presentations ≥37 weeks with spontaneous labour ranging from 5.7% in Norway to 20.6% in New Zealand.20
The comparable rate of 16.7% in 2009 in our study population was at the high end of this range.
This population-based study illustrates how a rising caesarean delivery rate can develop a momentum that is difficult to reverse. Until 2003, caesarean rates for multiparous women lagged behind the rising first-birth rate. However, as the TOLAC/VBAC rate decreased and repeat caesareans increased, the caesarean rate among multipara has exceeded that of nullipara. The steady increase in first-birth caesareans in the 2000's means that overall caesarean rates may still rise for some years, even if first-birth caesarean rates plateau.21
If rates of TOLAC and VBAC could recover from the decreases since the 1990s, this has the potential to reduce overall caesarean rates by up to 5%—only 3.5 women would have to be considered for a VBAC attempt in order to avoid one repeat caesarean. However, increasing TOLAC would not address the rising first-birth caesarean rate.1
Furthermore, any substantial recovery in TOLAC and VBAC rates appears less likely given the recent publication of a prospective study reporting a lower risk of fetal and infant death or serious infant morbidity following elective repeat caesarean compared with planned TOLAC.22
Policies encouraging uptake of private health insurance could have been expected to be associated with some of the rise in primary caesareans. Deliveries in private hospitals increased during the study period, and private hospitals did have higher rates of caesarean delivery. However, even after adjusting for changes in private/public care, and for increased maternal age and other factors, our predictive model found that only 24% of the increase in the primary caesarean rate could be explained. Information was not available for the entire study period on some factors associated with caesarean delivery (eg, placenta praevia and maternal obesity) but other studies assessing changes in maternal characteristics and pregnancy complications have similarly been unable to fully explain increasing caesarean rates.4–6
The large increase in caesarean deliveries in public patients as well as private patients suggests that the trend reflects more general changes in attitudes to obstetric risk factors and delivery management. Women and obstetricians may have become more averse to the perceived risks associated with vaginal delivery, or alternatively the perceived risks of caesarean section, relative to vaginal delivery, may have decreased. One possible contributor in our study population may have been a widely publicised NSW civil suit alleging negligence in a vaginal delivery.23
This case was finally settled in the appeals court in 2001 (coinciding with the single largest annual increase in caesarean rates 2000–2001), with a final payout of $11 million dollars The case contributed to one of the larger medical liability insurers going into liquidation in 2002, again with much media coverage.
Although intuitively it seems that reducing first-birth prelabour caesareans should have great potential as a target for interventions aimed at reducing overall caesarean section rates, the scenario is more complicated for two reasons. For one thing, even when the alternative scenario (await spontaneous labour) is a safe and acceptable option, it does not necessarily result in vaginal birth. Second, the nulliparae with a singleton cephalic presenting fetus who have a prelabour caesarean at term (a likely target group) represent a minority of all first-births (6% in 2009) that would dilute any impact on the overall rates.9
Further, these women are likely to have disparate medical and pregnancy conditions that may not be amenable to a single intervention.
The issue of whether efforts can or should be made to decrease the intrapartum caesarean rate is vexed. Intrapartum caesarean delivery is indicated following either concerns over fetal welfare or a failure to progress in labour, although the threshold for intervening may have changed. Interventions demonstrated to be effective in randomised trials include continuous support for women during childbirth, early amniotomy and early administration of oxytocin in spontaneous labour, and high-dose oxytocin for augmentation of delayed labour.24–26
Level of skill and obstetric training in labour management and operative vaginal deliveries are key issues, and there is some evidence that involvement of consultant obstetricians in decision-making can reduce the likelihood of caesarean.27
Evidence-based protocols for evaluating fetal status and managing dysfunctional labour need to be developed and promoted. In one study, application of a strict protocol dramatically reduced elective no-medical-indication births before 39 weeks, although the impact on caesarean rates was not an outcome.28
Finally, making performance data public has also resulted in changes in obstetrical services.29
However, changing practice requires that interventions are adapted to local circumstances.30
Breech presentation almost uniformly resulted in caesarean section by the 2000's (97.5% in Cohort 2, ) and makes a continued and stable contribution to caesarean section rates.6
Decreasing breech presentation as a means of reducing caesarean section rates should not be overlooked, as external cephalic version is evidence-based and likely to be achievable with training and education for both women and clinicians.32
Another contributor was caesarean delivery for multifetal pregnancies, which increased by 50% across the two cohorts. There is an absence of clear evidence about management, although the outcome of a randomised trial of prelabour vaginal birth versus caesarean section for twins is imminent.33
The strength of this population-based study is the availability of reliably collected labour and delivery data, and the ability to differentiate prelabour and intrapartum caesareans. This study reports caesarean section as an outcome, but not the outcomes of caesarean section which may be affected by changes in both obstetric and neonatal care. However, from 2001 to 2009, increasing caesarean section rates have not been accompanied by any significant change in perinatal mortality, but have been accompanied by a small (3%–3.2%), but statistically significant, increase in severe neonatal morbidity.34
An important consideration for this study is that women with more than one birth are not the same as those having only one. Women who intend to have more than one child may have a greater sense of commitment to a vaginal birth. For women who continue on to have more children, the relative benefit of increasing VBAC in the second birth would play an enhanced role. However, in a population where one-child families are common, reducing primary caesareans would be of increasing relevance. Of note, the steepest relative increase in caesarean sections (by 56%) occurred among women having a primary caesarean for their second birth. The reasons for this are unclear and warrant exploration.