“…It is well accepted in health economics that it is the estimated population mean cost that is the statistic of interest to policy makers, 19,20 and this approach of modeling mean cost has been widely published in the medical literature. [21][22][23] Although several methods have been developed and used in studying health care costs, the generalized linear model (GLM), especially the gamma regression model, 24 has been commonly used in the literature, 22,25,26 as it behaves well in the estimation of population means of health care costs, and does not require the normality and homoscedasticity assumptions of the data and avoids bias associated with retransforming to the raw scale. 27 In this study, a multivariable GLM employing a gamma distribution with a log-link function 24 was used to examine the association between the main outcome of hospital costs and SMM, while adjusting for several potential confounding variables: maternal age, race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, other, and unknown), singleton/multiple gestations, primary payer (private insurance, Medicaid, self-pay, other), delivery method (vaginal, cesarean), hospital bed size (small, medium, large), hospital location/teaching status (rural, urban nonteaching, urban teaching), and hospital region (Northeast, Midwest, South, West).…”