QT variability improves risk stratification in patients with dilated cardiomyopathy. (23rd March 2015)

• Record Type: Journal Article
• Authors: Fischer, C; Seeck, A; Schroeder, R; Goernig, M; Schirdewan, A; Figulla, H R; Baumert, M; Voss, A
• Is Part Of: Physiological measurement. Volume 36:Number 4(2015:Apr.)
• Page Start: 699
• Page End: 713
• Publication Date: 2015-03-23
• Subjects: QT variability -- heart rate variability -- blood pressure variability -- dilated cardiomyopathy -- risk stratification; Physiology -- Measurement -- Periodicals; Patient monitoring -- Periodicals; 612
• DOI: 10.1088/0967-3334/36/4/699
• Languages: English
• ISSNs: 0967-3334

Abstract

Abstract: Recently it could be demonstrated that systolic and diastolic blood pressure variability (BPV) as well as segmented Poincare plot analysis (SPPA) contribute to risk stratification in patients suffering from dilated cardiomyopathy (DCM). The aim of this study was to improve the risk stratification applying a multivariate technique including QT variability (QTV). We enrolled and significantly separated 56 low risk and 13 high risk DCM patients by nearly all applied BPV and QTV methods, but not with traditional heart rate variability analysis. The optimum set of two indices calculating the multivariate discriminate analysis (DA) included one BPV index calculated by symbolic dynamics method (DBPShannon ) and one index calculated from QTV (QTVlog ) achieving an area under the receiver operating characteristics curve (AUC) of 92%, sensitivity of 92.3% and specificity of 89.3%. Performing only electrocardiogram analysis, the optimum multivariate approach including indices from segmented Poincaré plot analysis and QTV still achieved a remarkable AUC of 88.3%. Increasing the number of indices for multivariate DA up to three, we achieved an AUC of 95.7%, sensitivity of 100% and specificity of 85.7% including one clinical, one BPV and one QTV index. Summarizing, we identified DCM patients with an increased risk of sudden cardiac death applying QTV analysis in a multivariate approach.