Background: Only 1 conventional score is available for assessing bleeding risk in patients with cancer-associated thrombosis (CAT): the CAT-BLEED score.
Objectives: Our aim was to develop a machine learning-based risk assessment model for predicting bleeding in CAT and to evaluate its predictive performance in comparison to that of the CAT-BLEED score.
Methods: We collected 488 attributes (clinical data, biochemistry, and International Classification of Diseases, 10th Revision, diagnosis) in 1080 unique patients with CAT. We compared CAT-BLEED score, Ridge and Lasso logistic regression, random forest, and Extreme Gradient Boosting (XGBoost) algorithms for predicting major bleeding or clinically relevant nonmajor bleeding occurring 1 to 90 days, 1 to 365 days, and 90 to 455 days after venous thromboembolism (VTE).
Results: The predictive performances of Lasso logistic regression, random forest, and XGBoost were higher than that of the CAT-BLEED score in the prediction of bleeding occurring 1 to 90 days and 1 to 365 days after VTE. For predicting major bleeding or clinically relevant nonmajor bleeding 1 to 90 days after VTE, the CAT-BLEED score achieved a mean area under the receiver operating characteristic curve (AUROC) of 0.48 ± 0.13, while Lasso logistic regression and XGBoost both achieved AUROCs of 0.64 ± 0.12. For predicting bleeding 1 to 365 days after VTE, the CAT-BLEED score achieved a mean AUROC of 0.47 ± 0.08, while Lasso logistic regression and XGBoost achieved AUROCs of 0.64 ± 0.08 and 0.59 ± 0.08, respectively.
Conclusion: This is the first machine learning-based risk model for bleeding prediction in patients with CAT receiving anticoagulation therapy. Its predictive performance was higher than that of the conventional CAT-BLEED score. With further development, this novel algorithm might enable clinicians to perform personalized anticoagulation strategies with improved clinical outcomes.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1016/j.jtha.2023.12.034 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Prediction model for major bleeding in anticoagulated patients with cancer-associated venous thromboembolism using machine learning and natural language processing.
Clin Transl Oncol
September 2024
Pfizer S.L.U. Medical Department, Madrid, Spain.
Article Synopsis
- A predictive model was created to estimate the risk of major bleeding in cancer patients undergoing anticoagulant treatment for venous thromboembolism (VTE) within six months following their diagnosis.
- The study analyzed data from electronic health records across nine hospitals in Spain, using natural language processing and machine learning to identify key predictors of bleeding and develop various predictive algorithms.
- Findings indicated that about 10.9% of the patients experienced major bleeding events after VTE diagnosis, with significant predictors being factors like hemoglobin levels and age, and the new models outperformed the existing CAT-BLEED score.
Developing a machine learning model for bleeding prediction in patients with cancer-associated thrombosis receiving anticoagulation therapy.
J Thromb Haemost
April 2024
Department of Research, Østfold Hospital, Sarpsborg, Norway; Institute of Clinical Medicine, University of Oslo, Oslo, Norway; Department of Hematology, Oslo University Hospital, Oslo, Norway.
Background: Only 1 conventional score is available for assessing bleeding risk in patients with cancer-associated thrombosis (CAT): the CAT-BLEED score.
Objectives: Our aim was to develop a machine learning-based risk assessment model for predicting bleeding in CAT and to evaluate its predictive performance in comparison to that of the CAT-BLEED score.
Methods: We collected 488 attributes (clinical data, biochemistry, and International Classification of Diseases, 10th Revision, diagnosis) in 1080 unique patients with CAT.
In Search of the Appropriate Anticoagulant-Associated Bleeding Risk Assessment Model for Cancer-Associated Thrombosis Patients.
Cancers (Basel)
April 2022
Médecine Interne, Hôpital Louis Mourier, Assistance Publique Hôpitaux de Paris, 92700 Colombes, France.
Patients with venous thromboembolism events (VTE) in the context of cancer should receive anticoagulants as long as the cancer is active. Therefore, a tailor-made anticoagulation strategy should rely on an individualized risk assessment model (RAM) of recurrent VTE and anticoagulant-associated bleeding. The aim of this review is to investigate the applicability of the currently available RAMs for anticoagulant-associated bleeding after VTE in the CAT population and to provide new insights on how we can succeed in developing a new anticoagulant-associated bleeding RAM for the current medical care of CAT patients.
View Article and Find Full Text PDFEstimating Bleeding Risk in Patients with Cancer-Associated Thrombosis: Evaluation of Existing Risk Scores and Development of a New Risk Score.
Thromb Haemost
May 2022
Department of Acute Internal Medicine, University Medical Center Utrecht, Utrecht, The Netherlands.
Background: Bleeding risk is highly relevant for treatment decisions in cancer-associated thrombosis (CAT). Several risk scores exist, but have never been validated in patients with CAT and are not recommended for practice.
Objectives: To compare methods of estimating clinically relevant (major and clinically relevant nonmajor) bleeding risk in patients with CAT: (1) existing risk scores for bleeding in venous thromboembolism, (2) pragmatic classification based on cancer type, and (3) new prediction model.
Want AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!