Evaluation of in‐hospital electrocardiography versus 24‐hour Holter for rate control in dogs with atrial fibrillation

Abstract

To determine if the in-clinic ECG-derived heart rate could predict the at-home Holter-derived 24-hour average heart rate (Holter24h ), and whether it is useful to identify slow versus fast atrial fibrillation in dogs. 82 pairs of 1-minute ECGs and 24-hour Holter recordings were acquired in 34 dogs with atrial fibrillation. The initial 24-hour Holter was used to test if the ECG heart rate can identify dogs with "slow" versus "fast" atrial fibrillation based on a Holter24h threshold value of 140 bpm. ECG heart rate overestimated Holter24h by 26 bpm (95% CI: 3 bpm, 48 bpm; P < 0 · 015) with a 95% limit of agreement of -21 to 83 bpm. The in-clinic ECG-derived heart rate Ä155 bpm had a sensitivity of 73% and a specificity of 100% for identifying a Holter24h HR Ä140 bpm; an in-clinic ECG-derived HR <160 bpm had a sensitivity and specificity of 91% each. In-clinic ECG assessment of heart rate in dogs with atrial fibrillation does not reliably predict the heart rate in their home environment. However, an in-clinic heart rate greater than 155 bpm is useful in identifying "fast" atrial fibrillation, allowing clinicians to stratify which case may benefit from antiarrhythmic therapy.