Heart rate variability parameters in horses distinguish atrial fibrillation from sinus rhythm before and after successful electrical cardioversion

• Published in: Equine veterinary journal 2017-11, Vol.49 (6), p.723-728
• Main Authors: Broux, B., De Clercq, D., Decloedt, A., Ven, S., Vera, L., Steenkiste, G., Mitchell, K., Schwarzwald, C., Loon, G.
• Format: Article
• Language: English
• Subjects: Animals; Arrhythmia; Atrial Fibrillation - therapy; Atrial Fibrillation - veterinary; Audio frequencies; Cardiac arrhythmia; cardiology; Cohort Studies; Electric Countershock - veterinary; Electrocardiography; equine; Female; Fibrillation; Heart Rate; heart rate monitor; horse; Horse Diseases - pathology; Horse Diseases - therapy; Horses; Male; Medical electronics; Rank tests; Sinus; Standard deviation; Statistical analysis; Variability
• ISSN: 0425-1644; 2042-3306
• DOI: 10.1111/evj.12684

Summary Background Atrial fibrillation (AF) is the most common pathological arrhythmia in horses. After successful treatment, recurrence is common. Heart rate monitors are easily applicable in horses and some devices offer basic heart rate variability (HRV) calculations. If HRV can be used to distinguish between AF and sinus rhythm (SR), this could become a monitoring tool for horses at risk for recurrence of AF. Objectives The purpose of this study was to assess whether in horses AF (before cardioversion) and SR (after cardioversion) can be differentiated based upon HRV parameters. Study design Cohort study with internal controls. Methods Six HRV parameters were determined in 20 horses, both in AF and in SR, at rest (2‐ and 5‐min and 1‐ and 4‐h recordings) and during exercise (walk and trot, 2‐min recordings). Time‐domain (standard deviation of the NN intervals, root mean squared successive differences in NN intervals and triangular index), frequency domain (low/high frequency ratio) and nonlinear parameters (standard deviation of the Poincaré plot [SD]1 and SD2) were used. Statistical analysis was done using paired Wilcoxon signed rank tests and receiver operating characteristic curves. Results HRV was higher during AF compared to SR. Results for the detection of AF were good (area under the receiver operating characteristic curve [AUC] 0.8–1) for most HRV parameters. Root mean squared successive differences in NN intervals and SD1 yielded the best results (AUC 0.9–1). Sensitivity and specificity were high for all parameters at all recordings, but highest during exercise. Although AUCs improved with longer recordings, short recordings were also good (AUC 0.8–1) for the detection of AF. In horses with frequent second degree atrioventricular block, HRV at rest is increased and recordings at walk or trot are recommended. Main limitations Animals served as their own controls and there was no long‐term follow‐up to identify AF recurrence. Conclusions AF (before cardioversion) and SR (after cardioversion) could be distinguished with HRV. This technique has promise as a monitoring tool in horses at risk for AF development.