Factors associated with coronary artery disease treatment costs in India using linear, gamma, and quantile regression models
摘要
Coronary artery disease (CAD) remains a leading cause of mortality in India and globally, contributing substantially to the healthcare and economic burden. This study aimed to estimate the annual cost of managing CAD in India and evaluate the association between healthcare expenditure and independent variables. A cross-sectional survey was conducted among CAD patients attending a cardiac-specific tertiary care hospital in New Delhi between May 2023 and October 2023. Direct and indirect costs associated with CAD management were estimated using mean and median values, and factors associated with costs were assessed using linear, gamma, and quantile regression models. The predictive performance of these models was compared using within-sample prediction measures. All costs were calculated in INR (Indian rupee) and US$(2026). 560 CAD patients (62 ± 10.5 years) were included. Among them, 182 patients received medical therapy alone, while 378 underwent invasive procedures in addition to medical therapy. The highest annual median cost was observed among patients undergoing coronary artery bypass grafting [INR 276225 (US$ 2896) per patient], whereas the lowest cost was observed among those receiving medical therapy alone [INR 18299 (US$ 192) per patient). Linear regression identified sex, distance from the hospital, and type of intervention as significant cost predictors. In gamma regression, only the type of intervention remained statistically significant, while quantile regression showed that male sex, greater distance from the hospital, and higher socioeconomic status were associated with increased costs. Appropriate statistical modelling is essential for accurately estimating CAD-related costs, as model performance varies depending on the characteristics of the data and selected evaluation metrics.