This study conducted a descriptive analysis of the number of cases, average length of stay, per-case cost, daily cost and cost composition based on the medical record header data of 4,653 special cases from a provincial tertiary hospital in Hefei in 2024. Univariate analysis revealed that age, gender, length of stay, medical insurance type, presence of complications, DRG grouping category and department category all significantly influenced hospitalization costs for special cases. Structural equation modeling analysis indicated that special cases constituted 12.64% of total cases but accounted for 22.71% and 29.81% of total length of stay and costs, respectively, demonstrating significant resource consumption. High-cost cases primarily incurred expenses from material fees, while low-cost cases were dominated by laboratory fees. Both cost structures exhibited a pattern of “heavy material consumption and light labor costs.” The study suggests that adjusting the medical cost structure, shortening the average length of stay and refining medical insurance payment policies through the DRG payment tool can achieve the goals of “cost control, quality improvement and efficiency enhancement,” thereby promoting a win-win situation for healthcare providers, insurers and patients.