Development of a Computerised Antithrombotic Risk Assessment Tool (CARAT) to optimise therapy in older persons with atrial fibrillation

Aim: To develop and evaluate a novel Computerised Antithrombotic Risk Assessment Tool (CARAT) to aid clinicians' decision making regarding the riskbenefit of antithrombotic therapy in older patients. Methods: CARAT was developed in an iterative process involving multidisciplinary feedback and computerisation of previously trialled algorithms. Hospital-based clinicians then applied the tool to patient cases, to evaluate its usability. Results: Overall, 94% of clinicians (n= 27 yielding 216 responses) were satisfied with CARAT's format. Most (72%) clinician responses agreed with CARAT recommendations; over two-thirds agreed with estimates of stroke and bleeding risk. However, geriatricians were 3.5 times more likely to disagree with CARAT recommendations than cardiologists, particularly in cases of high fall risk. Overall, 63% responded that CARAT was at least somewhat useful for clinical practice; 22% indicating it was very useful. Conclusion: CARAT has potential as a useful decision-support tool to assist clinicians in decision making regarding appropriate antithrombotic therapy in older patients.