Performance of multilabel machine learning models and risk stratification schemas for predicting stroke and bleeding risk in patients with non-valvular atrial fibrillation

Lü, Juan; Hutchens, R.; Hung, Joseph; Bennamoun, Mohammed; McQuillan, Brendan; Briffa, Tom; Sohel, Ferdous; Murray, Kevin; Stewart, Judith; Chow, Benjamin J.W.; Sanfilippo, Frank M; Dwivedi, Girish

doi:10.1016/j.compbiomed.2022.106126

Cited by 15 publications

(6 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, in the HAS-BLED score, the labile international normalized ratio (INR) could not be assessed and was thus omitted. This method has been used in other retrospective studies among patients with AF [10][11][12]. Therefore, a maximum of 8 points was used for "modified HAS-BLED" analysis in this study.…”

Section: Introductionmentioning

confidence: 99%

Development and Validation of a Nomogram for Predicting Mortality in Patients with Atrial Fibrillation and Acute Coronary Syndrome Who Underwent Percutaneous Coronary Intervention in a Chinese Multicenter Cohort

Hua

Tian

Wang

et al. 2022

Applied Bionics and Biomechanics

View full text Add to dashboard Cite

Background. This study is aimed at to establish an effective prognostic nomogram for patients with atrial fibrillation (AF) and acute coronary syndrome (ACS) underwent percutaneous coronary intervention (PCI). Methods. The nomogram was based on a retrospective study of 977 patients with AF and ACS who underwent PCI who were admitted to any of the 11 tertiary hospitals in the Beijing area between 2009 and 2015. The predictive accuracy and discriminative ability of the nomogram were determined by a concordance index ( C -index) and calibration curve and were compared using current risk scores such as GRACE, CRUSADE, CHA2DS2-VASc, and HAS-BLED. The results were validated using bootstrap resampling and a retrospective cohort study of 409 patients enrolled in Fuwai Hospital at the same institution. Results. Independent factors derived from multivariable analysis of the primary cohort to predict all-cause mortality were age, pattern of ACS, red blood cell distribution width, N-terminal proBNP, and serum creatinine, all of which were assembled into the nomogram. The calibration curve for the probability of recurrence showed that the nomogram-based predictions were in good agreement with actual observations. The C -index of the nomogram for predicting mortality was 0.764 (95% CI, 0.718-0.810), which was statistically higher than the C -index values for the current risk scores (from 0.573 to 0.681). In the validation cohort, the C -index of the nomogram for predicting all-cause death was 0.706 (95% CI 0.601-0.811), with no significant differences compared with GRACE and CRUSADE, but better than that of CHA2DS2-VASc and HAS-BLED. Conclusions. The nomogram has good prognostic prediction for patients with AF and ACS who underwent PCI.

show abstract

Section: Introductionmentioning

confidence: 99%

Development and Validation of a Nomogram for Predicting Mortality in Patients with Atrial Fibrillation and Acute Coronary Syndrome Who Underwent Percutaneous Coronary Intervention in a Chinese Multicenter Cohort

Hua

Tian

Wang

et al. 2022

Applied Bionics and Biomechanics

View full text Add to dashboard Cite

show abstract

“…A gradient boosting decision tree 19 -based classifier chain, 20 namely, ML-GBDT, was the best-performing algorithm (compared with support vector machine and multi-layer neural networks) in previous studies for predicting risks in AF patients. 18 The order of the classifier chain was established as major bleeding, ischaemic stroke, and all-cause death based on our previous experimental results 18 and time-to-event for different outcomes. The predicted probability of major bleeding was used to predict ischaemic stroke, and both predictions were used to predict all-cause death.…”

Section: Methodsmentioning

confidence: 99%

“…Compared with our previous study, predictors were modified based on the available variables. 18 We added medical treatments for AF as predictors but were unable to incorporate predictors such as haemoglobin and indigenous ethnicity, which were not recorded in the GLORIA-AF registry. We utilized different data sources to define outcomes.…”

Section: Methodsmentioning

confidence: 99%

“…In our previous study in an Australian AF registry, we constructed and assessed different multi-label ML algorithms, among which the multi-label gradient boosting decision tree (ML-GBDT) model demonstrated superior performance for predicting stroke and bleeding risk in patients with non-valvular AF compared with clinical risk scores (CHA 2 DS 2 -VASc and HAS-BLED). 18 One major limitation of this previous study was the retrospective nature of the dataset which was sourced exclusively from one geographical region with the predominant ethnic group being Caucasians. Moreover, non-vitamin K antagonist oral anticoagulants (NOACs) had not yet reached widespread use in our previous study.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Predicting multifaceted risks using machine learning in atrial fibrillation: insights from GLORIA-AF study

Lu,

Bisson,

Bennamoun

et al. 2024

European Heart Journal - Digital Health

Self Cite

View full text Add to dashboard Cite

Aims Patients with atrial fibrillation (AF) have a higher risk of ischaemic stroke and death. While anticoagulants are effective at reducing these risks, they increase the risk of bleeding. Current clinical risk scores only perform modestly in predicting adverse outcomes, especially for the outcome of death. We aimed to test the multi-label gradient boosting decision tree (ML-GBDT) model in predicting risks for adverse outcomes in a prospective global AF registry. Methods and results We studied patients from phase II/III of the GLORIA-AF registry between 2011 and 2020. The outcomes were all-cause death, ischaemic stroke and major bleeding within one year following the AF. We trained the ML-GBDT model and compared its discrimination with the clinical scores in predicting patient outcomes. A total of 25,656 patients were included (mean age 70.3 years (SD 10.3); 44.8% female). Within one-year after AF, ischaemic stroke occurred in 215 (0.8%), major bleeding in 405 (1.6%), and death in 897 (3.5%) patients. Our model achieved an optimized area under the curve (AUC) in predicting death (0.785, 95% CI: 0.757-0.813) compared to Charlson Comorbidity Index (0.747, p=0.007), ischaemic stroke (0.691, 0.626-0.756) comparing to CHA2DS2-VASc (0.613, p=0.028), and major bleeding (0.698, 0.651-0.745) as opposed to HAS-BLED (0.607, p=0.002), with improvement in net reclassification index (10.0%, 12.5% and 23.6% respectively). Conclusions The ML-GBDT model outperformed clinical risk scores in predicting the risks in patients with AF. This approach could be used as a single multifaceted holistic tool to optimize patient risk assessment and mitigate adverse outcomes when managing AF.

show abstract

“…Previous studies have explored the use of machine learning (ML) models in predicting bleeding risk in AF patients, demonstrating improved performance over conventional scoring systems 16,17 . However, these studies have limitations, such as relying on registry-based data that may not accurately represent the complexity and heterogeneity of real-world clinical practice, focusing on broader contexts like predicting bleeding risk from antithrombotic therapy in general patient populations, or concentrating only on AF subpopulations 18,19 .…”

Section: Introductionmentioning

confidence: 99%

Machine Learning - Based Bleeding Risk Predictions in Atrial Fibrillation Patients on Direct Oral Anticoagulants

Chaudhary,

Nourelahi,

Thoma

et al. 2024

Preprint

View full text Add to dashboard Cite

Importance: Accurately predicting major bleeding events in non-valvular atrial fibrillation (AF) patients on direct oral anticoagulants (DOACs) is crucial for personalized treatment and improving patient outcomes, especially with emerging alternatives like left atrial appendage closure devices. The left atrial appendage closure devices reduce stroke risk comparably but with significantly fewer non-procedural bleeding events. Objective: To evaluate the performance of machine learning (ML) risk models in predicting clinically significant bleeding events requiring hospitalization and hemorrhagic stroke in non-valvular AF patients on DOACs compared to conventional bleeding risk scores (HAS-BLED, ORBIT, and ATRIA) at the index visit to a cardiologist for AF management. Design: Prognostic modeling with retrospective cohort study design using electronic health record (EHR) data, with clinical follow-up at one-, two-, and five-years. Setting: University of Pittsburgh Medical Center (UPMC) system. Participants: 24,468 non-valvular AF patients aged ≥18 years treated with DOACs, excluding those with prior history of significant bleeding, other indications for DOACs, on warfarin or contraindicated to DOACs. Exposure(s): DOAC therapy for non-valvular AF. Main Outcome(s) and Measure(s): The primary endpoint was clinically significant bleeding requiring hospitalization within one year of index visit. The models incorporated demographic, clinical, and laboratory variables available in the EHR at the index visit. Results: Among 24,468 patients, 553 (2.3%) had bleeding events within one year, 829 (3.5%) within two years, and 1,292 (5.8%) within five years of index visit. We evaluated multivariate logistic regression and ML models including random forest, classification trees, k-nearest neighbor, naive Bayes, and extreme gradient boosting (XGBoost) which modestly outperformed HAS-BLED, ATRIA, and ORBIT scores in predicting clinically significant bleeding at 1-year follow-up. The best performing model (random forest) showed area under the curve (AUC-ROC) 0.76 (0.70-0.81), G-Mean score of 0.67, net reclassification index 0.14 compared to 0.57 (0.50-0.63), G-Mean score of 0.57 for HASBLED score, p-value for difference <0.001. The ML models had improved performance compared to conventional risk across time-points of 2-year and 5-years and within the subgroup of hemorrhagic stroke. SHAP analysis identified novel risk factors including measures from body mass index, cholesterol profile, and insurance type beyond those used in conventional risk scores. Conclusions and Relevance: Our findings demonstrate the superior performance of ML models compared to conventional bleeding risk scores and identify novel risk factors highlighting the potential for personalized bleeding risk assessment in AF patients on DOACs.

show abstract

Performance of multilabel machine learning models and risk stratification schemas for predicting stroke and bleeding risk in patients with non-valvular atrial fibrillation

Cited by 15 publications

References 26 publications

Development and Validation of a Nomogram for Predicting Mortality in Patients with Atrial Fibrillation and Acute Coronary Syndrome Who Underwent Percutaneous Coronary Intervention in a Chinese Multicenter Cohort

Development and Validation of a Nomogram for Predicting Mortality in Patients with Atrial Fibrillation and Acute Coronary Syndrome Who Underwent Percutaneous Coronary Intervention in a Chinese Multicenter Cohort

Predicting multifaceted risks using machine learning in atrial fibrillation: insights from GLORIA-AF study

Machine Learning - Based Bleeding Risk Predictions in Atrial Fibrillation Patients on Direct Oral Anticoagulants

Contact Info

Product

Resources

About