ISSN: 1223-1533

< Return to the list of articles

HOLTER ECG DATA COULD IMPROVE MORTALITY RISK PREDICTION FOR PERMANENT ATRIAL FIBRILLATION PATIENTS


Authors: Irena Kurcalte, Oskars Kalejs, Renars Erts, Artis Kalnins, S. Abdulah, Aivars Lejnieks




 

Introduction: Practical use of most Holter ECG (HM)-based risk predictors is limited in permanent atrial fibrillation (PAF) patients due to marked irregularity of ventricular contractions. The aim of this analysis is to study Holter-based predictors of mortality in PAF patients.

 

Methods: Two PAF patient groups were compared: patients who died – 74 pts (34 male, mean age (SD) - 78 (7,8) , and patients  were  alive - 181 (79 male,  age - 74  (10,1), minimum 3 year follow-up – 40 (16) month. Primary end-point: total mortality. Parameters of HR, HR circadian changes, incidence of ventricular arrhythmias obtained from archived in 2007 - 2010 HM recordings were in comparison. After univariate analysis of HM findings the logistic regression was performed to include HM parameters in patient risk assessment model with history of diseases, CHA2DS2-VASc score, QRS complex width.

 

Results: The univariate analysis showed that in the dead patient group was higher incidence of coronary artery disease,  diabetes,  hypertension,  chronic  kidney  disease (CKD), cancer, wide QRS complex, ventricular arrhythmias, higher CHA2DS2-VASc score, lower maximal HR (p-.002), circadian HR changes (day night average HR ratio (d/n HRr), p<.001, maximum minimum HR difference (delta HR), p-.001). Results of the multivariate analysis indicated that the eight predictors model including clinical (CKD, cancer, CHA2DS2-VASc score, wide QRS complex) and HM predictors (nonsustained VT, Lown class 2 VE, d/n HRr, delta HR) provided a statistically significant (p<.05) improvement over the clinical only model. The Nagelkerke R2indicated that the model accounted for 41% of the total variance. The correct prediction rate was 81%.

 

Conclusions:

  1. Inclusion of Holter ECG data in permanent AF patient mortality risk assessment in addition to clinical risk factors could noticeably increase predictive value of risk modeling.
  2. Decrease of HR circadian changeability is associated with higher mortality risk in permanent AF patients.