Association between maternal mortality and caesarean section in Ethiopia: a national cross-sectional study.

Background: Several studies concluded that there is a reduction of maternal deaths with improved access to caesarean section, while other studies showed the existence of a direct association between the two variables. In Ethiopia, literature about the association between maternal mortality and caesarean section is scarce. This study was aimed to assess the association between maternal mortality ratios and caesarean section rates in hospitals in Ethiopia.
Methods: Analysis was done of a national maternal health dataset of 293 hospitals that accessed from the Ethiopian Public Health Institute. Hospital specific characteristics, maternal mortality ratios and caesarean section rates were described. Pearson's correlation coefficient was used to determine the direction of association between maternal mortality ratios and caesarean section rate, taking regions into consideration. Presence of a linear association between these variables was declared statistically significant at p-value < 0.05.
Results: The overall maternal mortality ratio in Ethiopian hospitals was 149 (95% CI: 136-162) per 100,000 livebirths. There was significant regional variation in maternal mortality ratios, ranging from 74 (95% CI: 51-104) per 100,000 livebirths in Tigray region to 548 (95% CI: 251-1,037) in Afar region. The average annual caesarean section rate in hospitals was 20.3% (95% CI: 20.2-20.5). The highest caesarean section rate of 38.5% (95% CI: 38.1-38.9) was observed in Addis Ababa, while the lowest rate of 5.7% (95% CI: 5.2-6.2) occurred in Somali region. At national level, a statistically non-significant inverse association was observed between maternal mortality ratios and caesarean section rates. Similarly, unlike in other regions, there were inverse associations between maternal mortality ratios and caesarean section rates in Addis Ababa, Afar Oromia and Somali, although associations were not statistically significant.
Conclusions: At national level, a statistically non-significant inverse association was observed between maternal mortality ratios and caesarean section rates in hospitals, although there were regional variations. Additional studies with a stronger design should be conducted to assess the association between population-based maternal mortality ratios and caesarean section rates.