Automatic Holter electrocardiogram analysis in ischaemic stroke patients to detect paroxysmal atrial fibrillation: ready to replace physicians?

Abstract

Background The detection of paroxysmal atrial fibrillation (pAF) in patients presenting with ischemic stroke shifts secondary stroke prevention to oral anticoagulation. Aims In order to deal with the time‐ and resource‐consuming manual analysis of prolonged ECG‐monitoring data we investigated the effectiveness of pAF detection with an automated algorithm (AA) in comparison to a manual analysis with software support within the IDEAS study (study analysis (SA)). Methods We used the data set of the prospective IDEAS cohort of patients with acute ischemic stroke/TIA presenting in sinus rhythm undergoing prolonged 72h‐Holter‐ECG with central adjudication of AF. This adjudicated diagnosis of AF was compared to a commercially available AA. Discordant results with respect to the diagnosis of pAF were resolved by an additional cardiological reference confirmation. Results pAF was finally diagnosed in 62 patients (5.9%) of the cohort (n=1043). AA more often diagnosed pAF (n=60, 5.8%) as compared to SA (n=47, 4.5%). Due to a high sensitivity (96.8%) and negative predictive value ((NPV) 99.8%), AA is able to identify patients without pAF, while abnormal findings in AA require manual review (specificity 96%; positive predictive value (PPV) 60.6%). SA exhibited a lower sensitivity (75.8%) and NPV (98.5%) whilst showing a specificity and PPV of 100%. Agreement between the two methods classified by Kappa coefficient was moderate (0.591). Conclusion Automated determination of ‘absence of pAF’ could be used to reduce the manual review work load associated with review of prolonged Holter ECG recordings.