Objective  This study aimed to identify high-risk factors associated with high resource consumption during hospitalization among tumor patients using machine learning (ML) models, and to screen high-risk patients for pre-admission management to evaluate its impact on length of stay and hospitalization costs.

Methods  The date were retrospectively collected from inpatients treated at First Hospital of Shanxi Medical University between April 2021 and October 2022. Lasso regression was applied for feature selection, and multiple ML models were constructed to identify key factors influencing length of stay and hospitalization costs. SHAP values were used to interpret model results. Based on the modeling results, a risk factor screening and pre-admission management program was established between November and December 2022. The program was implemented from January to June 2023, and data were collected from patients who received or did not receive pre-admission management. Propensity score matching (PSM) was used to control baseline differences between groups, and causal forest (CF) analysis was performed to assess the effects of pre-admission intervention on hospitalization duration and costs.

Results  A total of 5,211 patients were included. The CatBoost model had the best results among the nine ML models constructed, and the prediction results showed that tumor type, preoperative waiting time, and age were the main high-risk factors affecting hospital resource utilization. Based on the model, 698 patients received pre-admission management. After PSM, 563 matched pairs were analyzed. CF analysis showed that the pre-admission group had a mean reduction in length of stay of 3.004 [95%CI (-3.334 to -2.675)] days compared with the control group and a mean reduction in hospitalization costs of 1,473.124  [95%CI (-2,166.09 to -780.16)] CNY, indicating that pre-admission management effectively improved hospital resource utilization efficiency.

Conclusion  Implementing pre-admission management for high-risk patients identified by ML models can significantly reduce length of stay and hospitalization costs for tumor patients, and may positively contribute to optimizing hospital workflows and resource allocation.

1.郑荣寿, 陈茹, 韩冰峰, 等. 2022年中国恶性肿瘤流行情况分析[J]. 中华肿瘤杂志, 2024, 46(3): 221-231. [Zheng RS, Chen R, Han BF, et al. Cancer incidence and mortality in China, 2022[J]. Chinese Journal of Oncology, 2024, 46(3): 221-231.] DOI: 10.3760/cma.j.cn112152-20240119-00035.

2.张欣,李维,张晨,等. 基于数据中心的预住院模式效果评价 [J].中国卫生标准管理, 2020, 11(17): 18-21. [Zhang  X, Li W, Zhang C, et al. Effect assessment of pre-registration hospitalization based on clinical data repository[J]. China Health Standard Management, 2020, 11(17): 18-21.] DOI: 10.3969/j.issn.1674-9316.2020.17.007.

3.李小芳,宋慧,蔡滨,等. 基于预住院开展多学科参与的麻醉术前访视应用思考[J]. 现代医院, 2021, 21(10): 1557-1560. [Li XF, Song H, Cai B, et al. Application of multidisciplinary participation in anesthesia preoperative visit based on pre-admission [J]. Modern Hospital, 2021, 21(10): 1557-1560. ] DOI: 10.3969/j.issn.1671-332X.2021.10.024.

4.丁四清,陆晶,秦春香,等. 数据挖掘在护理不良事件管理中的应用进展[J]. 中华护理杂志, 2019, 54(6): 873-877. [Ding SQ, Lu J, Qin CX, et al. Application progress of data mining in the management of nursing adverse events[J]. Chinese Journal of Nursing, 2019, 54(6): 873-877.] DOI: 10.3761/j.issn.0254-1769.2019.06.013.

5.Chang TH, Stuart EA. Propensity score methods for observational studies with clustered data: a review[J]. Stat Med, 2022, 41(18): 3612-3626. DOI: 10.1002/sim.9437.

6.Bonander C, Svensson M. Using causal forests to assess heterogeneity in cost-effectiveness analysis[J]. Health economics, 2021, 30(8): 1818-1832. DOI: 10.1002/hec.4263.

7.Brinkman-Stoppelenburg A, Polinder S, Olij BF, et al. The association between palliative care team consultation and hospital costs for patients with advanced cancer: an observational study in 12 Dutch hospitals[J]. Eur J Cancer Care (Engl), 2020, 29(3): e13198. DOI: 10.1111/ecc.13198.

8.Mooney K, Titchener K, Haaland B, et al. Evaluation of oncology hospital at home: unplanned health care utilization and costs in the huntsman at home real-world trial[J]. J Clin Oncol, 2021, 39(23): 2586-2593. DOI: 10.1200/JCO.20.03609.

9.Numico G, Cristofano A, Mozzicafreddo A, et al. Hospital admission of cancer patients: avoidable practice or necessary care?[J]. Plos One, 2015, 10(3): e0120827. DOI: 10.1371/journal.pone.0120827.

10.May P, Garrido MM, Cassel JB, et al. Using length of stay to control for unobserved heterogeneity when estimating treatment effect on hospital costs with observational data: issues of reliability, robustness, and usefulness[J]. Health Serv Res, 2016, 51(5): 2020-2043. DOI: 10.1111/1475-6773.12460.

11.刘同柱,赵昕昱,童贵显. 我国医院预住院模式研究现状与对策[J]. 中国医院管理, 2018, 38(7): 26-28. [Liu TZ, Zhao  XY, Tong  GX. Research status and improvement strategies of pre-registration hospitalization mode in Chinese hospitals[J]. Chinese Hospital Management, 2018, 38(7): 26-28.] DOI: CNKI:SUN:YYGL.0.2018-07-014.

12.刘亚楠,张瑞,弓玉红,等. 基于PSM法的预住院模式在甲状腺外科手术中的应用效果评价[J]. 中国医院管理, 2023, 43(7): 38-41, 52. [Liu YN, Zhang R, Gong YH, et al. Evaluation on the application effect of pre-hospitalization mode in thyroid surgery based on propensity score matching method[J]. Chinese Hospital Management, 2023, 43(7): 38-41, 52.] https://d.wanfangdata.com.cn/periodical/ChVQZXJpb2RpY2FsQ0hJMjAyNTA2MjISD3pneXlnbDIwMjMwNzAxMBoIZTdpMzY5NXE%3D

13.Chen Y, Huang L, Chen X. Application of a selective hospitalization model in the clinical practice of breast surgery[J]. Medicine (Baltimore), 2023, 102(26): e34209. DOI: 10.1097/MD.0000000000034209.

14.曾洪兰,姚嘉,周钰,等. 预住院管理模式在乳腺外科手术中的应用效果[J]. 中国地方病防治杂志, 2017, 32(4): 448. [Zeng HL, Yao J, Zhou Y, et al. Application effect of pre-admission management mode in breast surgery[J]. Chinese Journal of Control of Endemic Diseases, 2017, 32(4): 448.] DOI: CNKI:SUN:DYBF.0.2017-04-053.

15.鲍萍,申建民,查浩,等. 某三甲医院预住院管理模式的实践与体会[J]. 江苏卫生事业管理, 2016, 27(5): 36-37. [Bao P, Shen JM, Zha H, et al. Practice and experience of pre-admission management mode in a tertiary hospital[J]. Jiangsu Healthcare Administration, 2016, 27(5): 36-37.] DOI: CNKI:SUN:JWSG.0.2016-05-022.

16.Lundberg S, Lee SI. A unified approach to interpreting model predictions[J]. In Proceedings of the 31st International Conference on Neural Information Processing Systems (Long Beach, California, USA) (NIPS'17). Curran Associates Inc., Red Hook, NY, USA, 4768-4777. 2017. DOI: 10.48550/arXiv.1705.07874.

17.徐良辰,郭崇慧. 基于集成学习的胃癌生存预测模型研究 [J]. 现代图书情报技术, 2021, 5(8): 86-99. [Xu LC, Guo CH. Predicting survival rates for gastric cancer based on ensemble learning[J]. Data Analysis and Knowledge Discovery, 2021, 5(8): 86-99.] DOI: 10.11925/infotech.2096-3467.2021.0045.

18.王鑫,廖彬,李敏,等. 融合LightGBM与SHAP的糖尿病预测及其特征分析方法[J].小型微型计算机系统, 2022, 43(9): 1877-1885. [Wang X, Liao B, Li M, et al. Combination of light GBM and SHAP for diabetes prediction and feature analysis[J]. Journal of Chinese Computer Systems, 2022, 43(9): 1877-1885.]DOI: 10.20009/j.cnki.21-1106/TP.2022-09-016.

19.陈小昆,左航旭,廖彬,等. 融合XGBoost与SHAP的冠心病预测及其特征分析模型[J]. 计算机应用研究, 2022, 39(6): 1796-1804. [Chen XK, Zuo HX, Liao B, et al. Coronary artery disease prediction and feature analysis model based on XGBoost and SHAP[J]. Application Research of Computers, 2022, 39(6): 1796-1804.] DOI: 10.19734/j.issn.1001-3695.2021.11.0639.

20.张国杰,张红梅,柏庆花,等. 公立医院预住院流程建设探索与思考[J/OL]. 协和医学杂志, 2025: 1-14. [2025-09-15]. [Zhang GJ, Zhang HM, Bai QH, et al. Exploration and reflection on the construction of pre-admission process in public hospitals [J/OL]. Xiehe Medical Journal, 2025: 1-14. Published online September 15, 2025.] https://link.cnki.net/urlid/11.5882.R.20250711.1016.002