Development, validation, and clinical utility of risk prediction models for cancer-associated venous thromboembolism: A retrospective and prospective cohort study

Shuai Jin; Dan Qin; Chong Wang; Baosheng Liang; Lichuan Zhang; Weiyin Gao; Xiao Wang; Bo Jiang; Benqiang Rao; Hanping Shi; Lihui Liu; Qian Lu

doi:10.1016/j.apjon.2025.100691

Development, validation, and clinical utility of risk prediction models for cancer-associated venous thromboembolism: A retrospective and prospective cohort study

Asia Pac J Oncol Nurs. 2025 Mar 22:12:100691. doi: 10.1016/j.apjon.2025.100691. eCollection 2025 Dec.

Authors

Shuai Jin¹, Dan Qin², Chong Wang³, Baosheng Liang⁴, Lichuan Zhang², Weiyin Gao⁵, Xiao Wang², Bo Jiang⁶, Benqiang Rao⁷, Hanping Shi⁷, Lihui Liu⁸, Qian Lu²

Affiliations

¹ Department of Adult Care, School of Nursing, Capital Medical University, Beijing, China.
² Division of Medical & Surgical Nursing, School of Nursing, Peking University, Beijing, China.
³ Department of Gastrointestinal Oncology Surgery, Beijing Shijitan Hospital, Capital Medical University, Beijing, China.
⁴ Department of Biostatistics, School of Public Health, Peking University, Beijing, China.
⁵ Operating Room, Second Hospital of Shanxi Medical University, Taiyuan, Shanxi, China.
⁶ Department of Medical Oncology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China.
⁷ Department of Gastrointestinal Surgery, Beijing Shijitan Hospital, Capital Medical University, Beijing, China.
⁸ Department of Nursing, Beijing Shijitan Hospital, Capital Medical University, Beijing, China.

Abstract

Objectives: This study aims to develop cancer-associated venous thromboembolism (CA-VTE) risk prediction models using survival machine learning (ML) algorithms.

Methods: This study employed a double-cohort study design (retrospective and prospective). The retrospective cohort (n = 1036) was used as training set (70.0%, n = 725) and internal validation set (30.0%, n = 311); while the prospective cohort (n = 321) was used as external validation set. Seven survival ML algorithms, including COX regression, classification, regression and survival tree, random survival forest, gradient boosting survival machine tree, extreme gradient boosting survival tree, survival support vector analysis, and survival artificial neural network, were applied to train CA-VTE models.

Results: Univariate analysis and LASSO-COX regression both selected five predictors: age, previous VTE history, ICU/CCU, CCI, and D-dimer. The seven survival ML models (C-index: 0.709-0.760; Brier Score: 0.212-0.243) all outperformed Khorana Score (C-index: 0.632; Brier Score: 0.260) in external validation set. Among all models, the COX_DD model (COX regression + D-dimer) performed best. However, ML models and Khorana Score predicted CA-VTE risk on $\geq$ 7 days of hospitalization with an increase in Brier Score $\geq$ 0.25, showing poor calibration.

Conclusions: In this study, the CA-VTE risk prediction models developed in seven survival ML algorithms outperformed Khorana Score. Combining with D-dimer can improve model performance. Applying the nomogram based on the optimal COX_DD model allows oncology nurse to reassess CA-VTE risk once a week. The prediction models developed using survival ML algorithms in this study may contribute to the dynamic and accurate risk assessment of CA-VTE for cancer survivors.

Keywords: Decision making; Neoplasms; Risk stratification; Survival machine learning algorithm; Venous thromboembolism.