Comparing different venous thromboembolism risk assessment machine learning models in Chinese patients

Abstract

Objective: Venous thromboembolism (VTE) is a fatal complication and the most common preventable cause of death in hospitals. The risk-to-benefit ratio of thromboprophylaxis depends on the performance of the risk assessment model. A linear model, the Padua model, is recommended for medical inpatients in the United States but is not suitable for Chinese inpatients due to differences in race and disease spectrum. Currently, machine learning (ML) methods show advantages in modeling complex data patterns and have been applied to clinical data analysis. This study aimed to build VTE risk assessment ML models among Chinese inpatients and compare the predictive validity of the ML models with that of the Padua model. Methods: We used 376 patients, including 188 patients with VTE, to build a model and then evaluate the predictive validity of the model in a consecutive clinical dataset from Peking Union Medical College Hospital. Nine widely used ML methods were trained on the model derivation set and then compared with the Padua model. Results: Among the nine ML methods, random forest (RF), boosting-based methods, and logistic regression achieved a higher specificity, Youden index, positive predictive value, and area under the receiver operating characteristic curve than the Padua model on both the test and clinical validation sets. However, their sensitivities were inferior to that of the Padua model. Combined with the receiver operating characteristic curve, RF, as the best performing model, maintained high specificity with relatively better sensitivity and captured VTE patients' patterns more precisely. Conclusions: Advances in ML technology provide powerful tools for medical data analysis, and choosing models conforming to the disease pattern would achieve good performance. Popular ML models do not surpass the Padua model on all indicators of validity, and the drawback of low sensitivity should be improved upon in the future.

Machine learning for cardiology

“…Some feature types are more informative than others: quantitative features contain more details than the ordinal ones, and the same relationship holds between ordinal and categorical features. The empirical impact of this statement is present in many of the papers included in this review: risk scores obtained by reducing the number of used features often end up using more quantitative and ordinal features than categorical ones [14][15][16][17][18][19][20][21][22][23][24] .…”

“…Our literature review found that RF appeared second highest in the papers we reviewed, ranging from classification tasks 26,33,46 to regression tasks. Many authors use RFs to predict all-cause mortality, and others apply RF to particular cardiovascular diseases 37,46 or for risk assessment of heart failure 33 and venous thromboembolism 22 . In some studies, authors claim that the inferred RF models are helpful for clinical decisions, allowing to estimate whether a patient is suffering from heart failure with preserved ejection fraction or not 33 , and that their risk score assessment performs better than the state-of-the-art ones.…”

“…In some studies, authors claim that the inferred RF models are helpful for clinical decisions, allowing to estimate whether a patient is suffering from heart failure with preserved ejection fraction or not 33 , and that their risk score assessment performs better than the state-of-the-art ones. 22,26,46…”

Machine Learning to Predict Outcomes in Patients with Acute Pulmonary Embolism Who Prematurely Discontinued Anticoagulant Therapy

“…Reports on artificial intelligence in other medical scenarios are exponentially increasing in the literature. In patients with VTE, ML methods have been used for DVT diagnosis, 14 PE diagnosis, 16 risk stratification of acutely ill medical patients, 18,15 and in the prediction of VTE after surgery 17 or post-hospitalization 13 . Overall, supervised ML modestly improved the performance of LR, with some exceptions.…”

“…Recently, Wang et al reported on the performance of nine ML methods to predict VTE in hospitalized patients in a single center study. 15 In the validation subgroup, the difference in the c-statistics of the best performing method (Random…”