Results
During the study period 7894 out of 24129 women that delivered at EH had a CS, giving an overall CS rate of 32.7% (31.8% in BHH and 33.8% in AH).
As shown in table 1, all women were classified into one of the Robson groups as per their pregnancy characteristics. The largest population that EH serves was composed by groups 1 and 2 (nulliparous women) at 39.9%, followed by groups 3 and 4 (multiparous women) at 37.7%. The size of group 5 was 13.4% and the size of groups 8 and 10 was 0.9% and 4.4% respectively.
When looking at the ratio of nulliparous women in spontaneous labour, group 1, versus nulliparous women who had labour induced or were delivered by CS before labour, group 2, we found it to be 1.1:1. We found a similar scenario when looking at the ratio of multiparous women in spontaneous labour, group 3, versus multiparous women who had labour induced or were delivered by CS before labour, group 4, with our ratio being 1.8:1.
The CS rates varied from 3.0% in group 3 to 93.1% in group 9. The greatest contributor to the overall CS rate was group 5. Women in this group had high CS rates in both hospitals (86.1% in BHH and 86.3% in AH) and comprised 35.2% of all women delivered by CS at EH (33.8% in BHH and 36.7% in AH). The second largest group was group 2, contributing 26.6% to the overall CS rate (27.7% in BHH and 25.3% in AH). When looking at the subdivision, group 2a was responsible for the majority of this with 20.5% of the relative contribution to all CS. The third largest group was group 1 contributing 10.3% to the overall CS rate at EH (Table 1).
Over the five-year period of the study, CS rates at both hospitals increased from 30.7% in 2015 to 36.0% in 2019 (Figure 1). The overall CS rate at EH increased 1.5% per year (95% CI: 0.1-2.9, p-value: 0.04) and was mainly due to the increase in the CS rate at BHH, which had an increase of 2.4% per year (95% CI: 0.9-3.8, p-value: 0.01). AH also showed an increase in the CS rate of 0.9% per year. However, this was not statistically significant (95%CI: 0.5-2.4, p-value: 0.1).
As shown in Table 2, the CS rates were similar within most of the Robson groups when comparing BHH and AH. There was only evidence of a difference in group 2a (p-value 0.04), group 3 (p-value 0.007) and group 10 (p-value 0.008). BHH had higher CS rate among multiparous women in spontaneous labour (group 3) and AH had higher CS in women <37 weeks gestation (group 10) and induced nulliparous women with >37w (group 2A).
On assessment, we found several indicators that our data is of high quality. The number of CS and women delivered in the hospital corresponds to the same number in our data. Additionally, when crosslinking our data with the national birth registry there were no missing cases. The size of group 9 was 0.2%, which is in line with the WHO recommendations of less than 1%, however the CS rate of group 9 was 93% and not 100% as it should be by convention (9). On a closer review of patient’s records, two women were misclassified as transverse lie when they were actually compound presentations who had vaginal deliveries.