Joint use of cardio-embolic and bleeding risk scores in elderly patients with atrial fibrillation

• We observed acutely ill medical elderly patients with atrial fibrillation.
• We investigated how scores for cardio-embolic and bleeding risk co-classify them.
• Overall they were both at high cardio-embolic and high bleeding risk.
• Overall predicted cardio-embolic risk tended to be higher than bleeding risk.
• Antithrombotic prescription was primarily influenced by bleeding risk scores.

BackgroundScores for cardio-embolic and bleeding risk in patients with atrial fibrillation are described in the literature. However, it is not clear how they co-classify elderly patients with multimorbidity, nor whether and how they affect the physician's decision on thromboprophylaxis.MethodsFour scores for cardio-embolic and bleeding risks were retrospectively calculated for ≥ 65 year old patients with atrial fibrillation enrolled in the REPOSI registry. The co-classification of patients according to risk categories based on different score combinations was described and the relationship between risk categories tested. The association between the antithrombotic therapy received and the scores was investigated by logistic regressions and CART analyses.ResultsAt admission, among 543 patients the median scores (range) were: CHADS2 2 (0–6), CHA2DS2–VASc 4 (1–9), HEMORR2HAGES 3 (0–7), HAS-BLED 2 (1–6). Most of the patients were at high cardio-embolic/high-intermediate bleeding risk (70.5% combining CHADS2 and HEMORR2HAGES, 98.3% combining CHA2DS2–VASc and HAS-BLED). 50–60% of patients were classified in a cardio-embolic risk category higher than the bleeding risk category. In univariate and multivariable analyses, a higher bleeding score was negatively associated with warfarin prescription, and positively associated with aspirin prescription. The cardio-embolic scores were associated with the therapeutic choice only after adjusting for bleeding score or age.ConclusionREPOSI patients represented a population at high cardio-embolic and bleeding risks, but most of them were classified by the scores as having a higher cardio-embolic than bleeding risk. Yet, prescription and type of antithrombotic therapy appeared to be primarily dictated by the bleeding risk.