Can we improve the early detection of atrial fibrillation in a stroke unit? Detection rate of a monitor with integrated detection software.

Introduction: It is unknown whether monitors that include atrial fibrillation recognition software (AF-RS) increase the rate of early atrial fibrillation (AF) detection in acute stroke. We aimed to evaluate the AF detection rate of an AF-RS monitor and compare it with standard monitoring.

Methods: This was a retrospective, single-centre observational study conducted on consecutive patients with acute transient ischaemic attack or brain infarction attended in a stroke unit (SU) with six beds. Five beds had a standard monitor with a three-lead electrocardiogram (ECG)-tracing monitor that did not automatically detect AF, and one bed had a 12-lead ECG monitor with integrated AF-RS. All patients were monitored for at least 24 h and underwent a daily ECG during their stay in the SU. In case of unknown stroke aetiology, the patients underwent 24 h Holter monitoring.

Results: A total of 76 patients were included: 59 patients in the standard monitor group and 17 patients in the AF-RS monitor group. The mean age was 72.11 (±13.09) years, and 59.2% were men. A total of 20 new cases of AF were identified. The AF-RS monitor showed a higher rate of AF detection than the standard devices (57.1% vs 7.7%, p=0.031). The AF-RS monitor showed sensitivity, specificity, positive predictive value, and negative predictive values of 57.1%, 100%, 100% and 76.9%, respectively. For the standard monitors, these values were 7.7%, 100%, 100% and 79.3%, respectively.

Conclusion: The monitor with AF-RS demonstrated a higher detection rate for AF than standard ECG monitoring in acute stroke patients in a SU.