<p>In large-scale disasters, coordinated interactions among hospitals are critical for effective medical response, which should be viewed as an interconnected system rather than isolated units. This paper presents a novel network modeling and analysis framework to describe the collaboration modalities of hospitals and identify key hospitals that play crucial roles in disaster rescue and treatment. The proposed model incorporates hospital interconnection patterns and geographical accessibility across different rescue phases. Five node importance metrics are employed to rank hospital significance, including degree-strength centrality, betweenness centrality, closeness centrality, PageRank, and hospital capacity. Meanwhile, four network robustness metrics are used to evaluate network robustness and node-ranking effectiveness, including network efficiency, number of components, giant component size, and giant component capacity. Applied to Beijing, China, the model successfully identifies high-level hospitals and highlights that large-scale hospitals are not always the most critical nodes. Geographic location significantly constrains the capacity of certain hospitals to participate in post-disaster treatment. The findings suggest that urban healthcare planners should consider not only hospital scale but also spatial positioning when allocating resources and planning new facilities. The flexibility and adaptability of the proposed general model ensure that it can be iteratively improved and customized, making it a valuable tool for advancing disaster preparedness and response strategies. It also offers practical applications for improving patient transport, resource allocation, and overall disaster preparedness by optimizing hospital roles within the healthcare network.</p>