Abstract 14396: Characterizing Pulse Deficits in Atrial Fibrillation Using Simple ECG Features
Introduction: Atrial fibrillation can cause pulse deficits, or absence of a pulse wave following ventricular depolarization. We sought to characterize pulse deficits based on ECG features and correlate presence of pulse deficits with clinical characteristics, such as symptom severity and findings on cardiac imaging.
Methods: Data was collected from patients having radiofrequency ablation. 60s epochs of 12-lead surface ECG and radial intra-arterial blood pressure were extracted. RR intervals (RRI) were calculated using Pan-Tompkins algorithm. Pulse deficits were identified using intra-arterial blood pressure and defined as a wave with pulse pressure < 20% mean pulse pressure. The preceding RRI was defined as S, and the pre-preceding RRI was defined as L. RRIs associated with normal pulses were also calculated. Multivariable logistic regression was used to assess detection capabilities of ECG features. Patients with and without pulse deficits were compared on symptom severity (CCS AF, Mayo Symptom Severity Index) and findings on cardiac MRI.
Results: Eighteen patients had pulse deficits while 26 patients had none. In total, 628 pulse deficits were observed with L (621±194ms) greater than S (438±54.5ms) (p <0.0001). L and S were both lower in pulse deficits than normal pulses (p < 0.001); however, the percent decrease was greater in pulse deficits (p < 0.001). Using L and S in multivariable logistic regression produced an accurate model (c-statistic 0.83). CCS AF score was higher in patients with pulse deficits (2.50 v. 2.33), but difference was not significant (p = 0.81). Significant differences were not noted in total (p = 0.83) and itemized Mayo AF Symptom Severity Inventory. Patients with pulse deficits were more likely to have mitral regurgitation (0.33 v. 0.12) with near significance (p = 0.09).
Conclusion: It is possible to characterize pulse deficits using simple ECG features. Further work is needed to assess clinical significance and develop predictive model.
Author Disclosures: V. Venkataraman: None. A.Y. Sun: Research Grant; Significant; Boston Scientific Corp., Medtronic, Inc.. Other Research Support; Significant; Medtronic, Inc., Biosense Webster, Inc.. R. Yang: None. J.I. Koontz: None. K.B. Campbell: None. Z.S. Finn: None. R. Schumann: None. S. Guerrant: None. C. Henriquez: None. J.P. Daubert: Research Grant; Significant; ARCA Biopharma, Inc., Biosense Webster, Inc., Boston Scientific Corp., NIH, St. Jude Medical, Heart Metabolics. Other Research Support; Modest; ARCA Biopharma, Orexigen, Northwestern University. Other Research Support; Significant; Biosense Webster, Inc., Biotronik, Boston Scientific Corp., St. Jude Medical. Consultant/Advisory Board; Modest; Biotronik. Other; Modest; Biotronik, Boston Scientific Corp., St. Jude Medical, VytronUS, Inc., Zoll Medical Corporation.
- © 2016 by American Heart Association, Inc.