This study sought to compare self-reported rates of, and reasons for, caesarean section between England and Queensland (Australia) to identify potential explanations for discrepant caesarean section rates. Overall, women in Queensland were at higher odds of having a caesarean section than women in England; they had approximately 1.5 times higher odds of a pre-labour caesarean, and those with a history of caesarean had more than double the odds of having had a caesarean. While the overall rate of intrapartum caesareans did not differ between the countries, rates were higher in Queensland than in England among women without a previous caesarean. The absence of any difference between countries in the rate of assisted vaginal deliveries suggests that England’s lower caesarean rate is not explained by a tendency to choose assisted vaginal delivery over emergency intrapartum caesarean.
In both countries, the most commonly reported reasons for pre-labour caesarean were breech presentation for women without a previous caesarean and previous caesarean for women with a history of caesarean. Among women without a previous caesarean, fetal distress was the most common reason for intrapartum caesareans in England, while failure to progress in labour was the most common reason for intrapartum caesareans in Queensland. For women with a previous caesarean, failure to progress was the most common reason for intrapartum caesarean in both countries.
While some similarities were identified between countries as to the most common reasons for caesarean, a number of key differences were also observed. Perhaps most notable were the apparent differences in the salience of previous caesarean section as the single main reason for caesarean. Compared to women in England, women in Queensland were more likely to have a pre-labour caesarean and to have an intrapartum caesarean due only to having previously had a caesarean. Although the proportion of women with a previous caesarean was higher in the Queensland sample, when those women were isolated, the proportion having a repeat caesarean due only to having previously had a caesarean remained higher in the Queensland sample.
While such vast differences could be the result of incongruent clinical standards, guidelines in Queensland and the UK [21, 26] are aligned in their recommendations for women with a previous caesarean section. Both currently recommend discussion of the risks and benefits of different modes of birth, consideration of the capabilities of the facility, and responsivity to maternal preferences for mode of birth. However, there is a longer history of guidelines legitimising the role of maternal preferences in the UK than in Australia . Such guidelines were first released in Queensland at approximately the same time as women in our sample gave birth, and represented a significant departure from existing documents that recommended such mode of birth decisions be guided by clinical expertise . The recency of the shift towards responding to women’s preferences in Queensland may explain discrepancies in the practice patterns reported by women with a previous caesarean.
Discrepancies were also identified between countries in the reasons for caesarean among women without a previous caesarean. Intrapartum caesareans due to a failure to progress in labour were more likely among women in Queensland than women in England. While the term ‘failure to progress’ can incorporate a wide range of circumstances, these are typically inter-related and often characterised by a prolonged labour. Criteria for defining prolonged labour are similar in Queensland and the UK [40, 41], however the only reference to management of delayed progress in labour in any current Queensland guideline is available in the Normal Birth guidelines and advice is restricted to ‘consulting an obstetrician’ . Previous studies have demonstrated that international variation in caesarean rates is largely attributable to rates of caesarean among nulliparous women with singleton, term, and cephalic pregnancies . The authors propose that differences in obstetric practice for the management of labour (e.g., use of oxytocin to correct dystocia) may be responsible for such variation . The absence of clear guidance for intrapartum management of potential risk factors in Queensland is likely to result in variable practice that may not be based on current evidence. Women in Queensland were also at greater odds than women in England of having had a pre-labour caesarean due to suspected disproportion. Guidelines from the UK suggest that suspected disproportion alone should not be an indication for caesarean due to the limited reliability of methods to estimate infant size while in utero [21, 38]. In Queensland and Australia there is not a single guideline for caesarean section, but rather, a collection of guidelines for specific populations or indications [26–29]. Currently there is not a clinical guideline for suspected cephalopelvic disproportion or suspected macrosomia, so it is unclear how decisions about mode of birth are made when such concerns are raised.
Although differing among primary pre-labour caesareans, maternal preference caesarean was reported at a low frequency by women in both countries (less than 1% of births). In both the UK and Australia, care providers are supported by professional bodies to perform a caesarean for maternal request in the absence of a medical indication if they consider the woman’s preference to be fully informed, and are comfortable performing the procedure [21, 28]. Despite professional endorsement under given circumstances, maternal request as a sole reason does not account for the observed differences across countries in rates of caesarean section. Our findings are consistent with previous studies [19, 43, 44] that maternal request contributes to only a small proportion of the overall rates of caesarean section.
Carer or hospital recommendation was not provided in the list of reasons for caesarean for women in England; nor did any women spontaneously report this as the main reason for their caesarean. In Queensland, 1.5% of intrapartum caesareans and 6.5% of pre-labour onset caesareans were reported by women as being due to carer/hospital recommendation alone (i.e., without specifying concurrent clinical indications). Given the variation in measurement across the two countries, it remains difficult to determine whether the discrepant rates are indicative of true differences in practice between the countries.
Another possible explanation for the observed cross-cultural differences in practice patterns is differences in the training received by those supporting women through labour and birth. In England, more than half of women (56.3%) are cared for primarily by a midwife during labour and birth . While a similar rate of midwife care (57.5%) is evident for the approximately 70% of Queensland women who birth in the public sector, a large majority of the 30% of Queensland women who birth in the private sector (89.6%) are cared for by an obstetrician . In Queensland, women who birth in the private sector are more likely than women in the public sector to have a caesarean section, and these differences are not attributable to maternal risk or preference . Further examination of the impact of training of the primary accoucheur, on both care decisions made during pregnancy or labour and the associated outcomes, is an important avenue for study in this field.
Strengths and limitations of the current study must be acknowledged. Unlike many previous studies that have used routinely collected data (such as hospital records or birth registrations) to examine rates of caesarean section [14, 18, 23], our findings are based on data reported by women and thus hinge on the reliability of this information. Despite this difference, the reported rates of caesarean section and assisted vaginal delivery in our samples vary little from national population data in both countries [24, 25]. This is consistent with previous literature highlighting the congruence of maternal self-report and medical records on indicators such as mode of birth, reason for caesarean, reproductive and obstetric history, onset of labour, use of analgesia, perineal status after birth, and infant birthweight [30–33, 47]. As with any method, self-report is not without bias and relies on women having been adequately informed regarding their treatment and being able to reliably recall this information. While only one study appears to have examined informant concordance of reason for caesarean, mismatch was mainly demonstrated for classifications reported by the clinician as ‘failed induction’, with women being more likely to provide the reason for the attempted induction as the indication for caesarean . Thus discordance of indications for caesarean may be more likely when multiple factors are present, however reports are still well-aligned in such cases. Given that the focus of this paper was on examining differences between countries in women's self-reported indication, and any self-reported error is likely to be similar in both samples, this is unlikely to have influenced our main findings. It is possible that there are systematic differences between the countries in the classification and communication of reasons for caesarean, however the potential influence of this on how women ascribe reasons for their caesarean, along with the observed differences in maternal education, is unknown.
The response rates of the respective surveys may limit the generalisability of the findings. Other self-report population-level surveys of maternity care experience have achieved higher rates of response [48, 49], however recruitment strategies relied on hospital and care provider involvement which may interfere with the perceived independence of the survey. While the women who participated in this study were not representative of the population of birthing women in their respective countries on some demographic measures, the respondent samples were largely representative on key clinical indicators examined in this study (e.g., mode of birth and experience of labour). Overrepresented characteristics in the sample that are often associated with increased likelihood of caesarean, such as maternal age in England and multiple pregnancy or private facility in Queensland, do not appear to have increased the observed rates for mode of birth. It is unclear how this may have affected the reporting of reasons for caesarean.
The measure of single (main) reason for caesarean, derived based on hierarchical ordering of clinical indications [see Additional file 1], may not have accounted for the possible complex interactions between indications. As already discussed, previous literature relating to indications for caesarean has relied on routinely collected data wherein a single indication for caesarean is provided by the attending clinician. How decisions are made when multiple indications are present or the consistency of approach between different clinicians is unknown. Most women in our sample provided only one or two reasons for their caesarean and where multiple reasons were provided this was often the pairing of a clinical indicator with maternal preference or carer/hospital recommendation. Decisions about coding were held consistent across the two countries to avoid artificial inflation of differences in main reason for caesarean. However, it should also be noted that the checklists provided to women to assess reasons for caesarean differed slightly between countries (see Table 1). While we have been intentionally cautious around interpretation of findings relating to carer/hospital recommendation and multiple pregnancy as reasons for caesarean, it is possible that the absence of these options may have altered how women responded to the question.