To analyze the composition and influencing factors of hospitalization costs for stroke patients, providing data support for rational control of medical expenses and alleviating the economic burden on patients.

The data of 66 524 stroke patients discharged from January 1, 2020 to December 31, 2022 were collected from the medical expense accounting management reporting system of a province in central China. The grey relational analysis and degree of structural variation analysis were performed to analyze the hospitalization cost composition for stroke and its subtypes. A Generalized Linear Model (GLM) and an eXtreme Gradient Boosting (XGBoost) model were constructed to explore the key influencing factors.

Among 66 524 stroke patients, 85.80%, 10.84%, and 3.36% had ischemic stroke (IS), intracerebral hemorrhage (ICH), and subarachnoid hemorrhage (SAH), respectively. The M (P₂₅, P₇₅) of the hospitalization cost for stroke patients was 6 992.32 (3 997.76, 13 475.91) CNY. The hospitalization cost for SAH patients was the highest, reaching 88 389.28 (12 494.50, 151 711.34) CNY. The top three categories of hospitalization expenses for stroke patients were drug costs, other costs, and treatment costs, with the average proportions of 29.29%, 24.61%, and 19.43%, respectively. The hospitalization costs of IS and ICH patients were mainly attributed to drug costs. For SAH patients, other costs accounted for the largest proportion of hospitalization expenses. The grey relational analysis results showed that the top three factors correlated with total hospitalization costs were Western medicine costs, other costs, and treatment costs, with correlation coefficients of 0.951 8, 0.936 2, and 0.900 4, respectively. The degree of structural variation analysis showed that the contribution rates of structure variation (CSV) for medication costs, examination costs, and other costs were 24.34%, 19.50%, and 27.74%. The GLM results indicated that longer hospital stays (≥ 7 days), emergency admission, surgery, intensive care, ICH, SAH, and for-profit medical institutions were associated with higher hospitalization costs. The XGBoost model analysis revealed that the top five factors influencing hospitalization costs for stroke patients were the type of medical institution, length of stay, whether surgery was performed, the level of medical institution, and the subtype of stroke.

IS is the most common stroke subtype, and hospitalization costs for stroke patients are primarily driven by drug expenses. SAH patients have higher hospitalization costs, mainly due to other expenses. By implementing appropriate admission assessments, improving medical efficiency, shortening hospital stays, and optimizing treatment plans, the economic burden on stroke patients can be reduced.