RESUMEN
Objective:To explore influencing factors of hospitalization cost for breast cancer inpatients, develop an applicable standard using bundled payment approach, so as to provide reference for health insurance organizations to formulate reasonable imbursement method.Methods:31 792 inpatients′ medical records from 16 tertiary hospitals in Beijing, 2015 were included in the present study. Non-parametric tests and logistic regression model for ordinal categorical variable were applied to explore potential influencing factors. Decision tree model was performed to yield case-mix related groups.Results:The breast cancer inpatients were classified into 17 groups based on the following factors: surgical treatment/non-surgical treatment, surgical modality, radiology treatment, lymphadenectomy, comorbidity. RIV(reduction in variance)was 0.26. CV(coefficient of variation)was 0.24~0.97.Conclusions:It was feasible to apply the case-mix bundled payment method as a reimbursement method for breast cancer inpatients. The present study served as reference for formulating related policies.