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

BACKGROUND: Appropriate anticoagulant therapy for patients with atrial fibrillation (AF) requires assessment of stroke and bleeding risks. However, risk stratification schemas such as CHA2DS2-VASc and HAS-BLED have modest predictive capacity for patients with AF. Multilabel machine learning (ML) techniques may improve predictive performance and support decision-making for anticoagulant therapy. We compared the performance of multilabel ML models with the currently used risk scores for predicting outcomes in AF patients. METHODS: This was a retrospective cohort study of 9670 patients, mean age 76.9 years, 46% women, who were hospitalized with non-valvular AF, and had 1-year follow-up. The outcomes were ischemic stroke (167), major bleeding (430) admissions, all-cause death (1912) and event-free survival (7387). Discrimination and calibration of ML models were compared with clinical risk scores by area under the curve (AUC). Risk stratification was assessed using net reclassification index (NRI). RESULTS: Multilabel gradient boosting classifier chain provided the best AUCs for stroke (0.685 95% CI 0.676, 0.694), major bleeding (0.709 95% CI 0.703, 0.716) and death (0.765 95% CI 0.763, 0.768) compared to multi-layer neural networks and classifier chain using support vector machine. It provided modest performance improvement for stroke compared to AUC of CHA2DS2-VASc (0.652, NRI = 3.2%, p-value = 0.1), but significantly improved major bleeding prediction compared to AUC of HAS-BLED (0.522, NRI = 22.8%, p-value < 0.05). It also achieved greater discriminant power for death compared with AUC of CHA2DS2-VASc (0.606, p-value < 0.05). ML models identified additional risk features such as hemoglobin level, renal function. CONCLUSIONS: Multilabel ML models can outperform clinical risk stratification scores for predicting the risk of major bleeding and death in non-valvular AF patients.

Antithrombotic Therapy in Atrial Fibrillation Management in Western Australia: Temporal Trends and Evidence-Treatment Gaps.

OBJECTIVE: To describe temporal trends in appropriate antithrombotic therapy use in hospitalised atrial fibrillation (AF) patients and identify evidence-treatment gaps in clinical practice. DESIGN: Retrospective cohort study from January 2009-March 2016. SETTING: Tertiary and secondary teaching hospitals in Perth, Western Australia. PARTICIPANTS: Hospitalised adults with non-valvular AF. RESULTS: We identified 11,294 index AF admissions, with a mean age of 76.9 years, 45.8% women and 86.3% at high risk of stroke (CHA2DS2-VASc score ≥2 in men and ≥3 in women). In high risk subjects use of appropriate antithrombotic therapy improved over time with increasing oral anticoagulant (OAC) use and declining sole antiplatelet use (both trend p<0.001). However, by study end only 45.3% of high-risk patients were receiving OAC therapy. In low risk patients, receipt of OAC therapy was steady throughout the study at 40.5% (trend p=0.10). The gender gap in OAC use narrowed over time, with no significant difference between high risk men and women by study end. Use of OAC therapy in elderly patients (age ≥75 years) remained lower than younger patients (age <65 years) over the entire period, with only 31% of elderly patients receiving OAC therapy at study end. From 2012 onwards use of non-vitamin K oral anticoagulants (NOACs) doubled each year with declining warfarin use (both trend p<0.001). CONCLUSION: Despite substantial uptake of NOACs, OAC therapy in AF patients at high risk of stroke remains under-utilised in Western Australia and over-utilised in low risk patients. Further work is required to reduce treatment-risk mismatch for stroke prevention in AF patients.