Estimating the effect of cesarean section rate on health outcome - Evidence from Swedish hospital data

This paper tests the null hypothesis of a zero effect of cesarean section r ate on health outcome against the alternative of a positive effect. Using d ata from 59 hospitals in Sweden from 1988-92, we specify two separate linea r regression models for health outcome, one with perinatal mortality, and t he other with rate of asphyxia, as dependent variable, We estimate the mode ls by single-year cross-section regressions and as pooled data systems. The null hypothesis cannot be rejected, i.e., we do not find any significant p ositive effect of cesarean section rate on health outcome. Thus, we conclud e that an increase in cesarean section rate does not imply lower perinatal mortality or lower rate of asphyxia. This in turn indicates that the minimu m cesarean section rate is optimal.