(CCSP047) IMPLEMENTATION OF ELECTROCARDIOGRAM ATRIAL FIBRILLATION REPORTING PROMPT INCREASED ATRIAL FIBRILLATION DETECTION AND ANTICOAGULATION RATES

Background: The recognition of atrial fibrillation (AF) allows for thromboembolic risk assessment and, in the majority of patients, reduction in morbidity and mortality through the initiation of oral anticoagulant use. Automated reporting of AF in patients who demonstrate ventricular pacing coincident with electrocardiogram (ECG) acquisition is poor. We sought to evaluate whether the reporting of AF increases AF clinical diagnosis and anticoagulation initiation, and whether an atrial rhythm reporting prompt increases the reporting rates.

METHODS AND RESULTS: We tested the impact of an automated atrial rhythm prompt during ventricular paced rhythms on ECG reader completion of the underlying rhythm, and subsequent clinical diagnosis of atrial fibrillation. All ECGs from Vancouver General Hospital between June 17, 2014 to January 31, 2018 with the diagnostic code “Ventricular-paced rhythm” were included in the retrospective cohort. This diagnostic code was changed within the MUSE (GE Healthcare) ECG reporting software on April 30, 2021 to “Ventricular-paced rhythm; Atrial rhythm:”. This prompt allowed the reading cardiologist to insert the underlying atrial rhythm or delete the statement as appropriate. The subsequent one year of ECGs served as the prospective cohort.

We sought to determine whether 1) changing the diagnostic code on the ECG reporting software improves the detection of AF. 2) whether anticoagulation is appropriately implemented or assessed at the time of AF documentation on ECG. Prior AF diagnosis, AF recognition, anticoagulation status pre- and post-hospitalization, and contraindications where present were assessed through hospital chart review. Prior to the atrial rhythm prompt, 1202 of 1978 (61%) ECGs reported the atrial rhythm. This increased substantially in the cohort after the atrial rhythm prompt (402/470 (86%)).

Prior to the atrial rhythm prompt, 466 of the 776 (60%) where the atrial rhythm was not reported showed AF. Of these, 181 charts with AF on ECG were reviewed (106 where the rhythm was reported and 75 where the rhythm was not reported). Previously diagnosed AF was similar in the reported vs. not reported groups (90% vs. 89%). AF being reported on ECG increased the likelihood of diagnosis at the time of discharge (64% vs. 13%, p=0.026). In each case where AF was identified the anti-coagulation strategy was assessed and where possible implemented (5 started on OAC, 2 contraindication, 1 died prior to discharge).

Conclusion: Altering ECG reporting software to prompt atrial rhythm reporting substantially increases the rate of atrial rhythm reporting. Reporting the atrial rhythm increases both clinical AF detection and appropriate anticoagulation use.