Caesarean delivery and neonatal mortality rates in 46 low- and middle-income countries: a propensity-score matching and meta-analysis of Demographic and Health Survey data.

BACKGROUND Previous research on the association between caesarean delivery (CD) and neonatal mortality has had methodological limitations and given conflicting results. We conducted a study to: (i) estimate the association between CD at the individual level and neonatal mortality rates (NMR) in 46 countries; and (ii) examine whether this association varies among countries according to country-level rates of CD or gross domestic product (GDP). METHODS We obtained data from nationally representative Demographic and Health Surveys of women aged 15-49 years and their children aged 0-59 months (N = 392 883). Propensity-score matching, meta-analysis, and meta-regression were used to address the study objectives. RESULTS The pooled odds ratio (OR) for the association between individual level CD and NMR in 46 countries was 1.67 (95% confidence interval (CI) 1.48-1.89), with moderate heterogeneity (I(2) = 39%). A meta-analysis of subgroups indicated that CD at the individual level was positively associated with NMR in countries with low (OR = 1.99, 95% CI 1.71-2.33, I(2) = 8.5%) and medium (OR = 1.53, 95% CI 1.29-1.82, I(2) = 24%) rates of CD. There was substantial heterogeneity of the effects of CD among countries with high rates of CD (I(2) = 63%). Results of meta-regression showed that the association of individual-level CD with NMR depended upon country-level rates of CD. Compared with countries with high rates of CD, the OR of the NMR associated with individual-level CD in countries with low rates of CD was estimated to increased by a factor of 1.48 (95% CI 1.09-1.97). CONCLUSIONS Studies are needed to better understand the risks posed by CD in countries with low and medium rates of CD and to identify possible reasons for the heterogeneity in effects of CD among countries with high rates of CD.