[Detection of atrial late potentials with P wave signal averaged electrocardiogram among patients with paroxysmal atrial fibrillation].

UNLABELLED: The analysis of the QRS-complex with signal averaged ECG (SAECG) has been evaluated for patients affected by ventricular tachycardia for a long time. A longer filtered QRS-complex was a marker of a slower ventricular conduction velocity and reentry tachycardia. This method was modified for an analysis of the P wave (P-SAECG). Different filter methods were evaluated for the analysis of atrial late potentials.

METHOD: We measured the bidirectional P wave signal averaged ECG of 45 consecutive patients with (group A) and without (group B) paroxysmal atrial fibrillation (PAF) and 15 young volunteers without a cardiac disease (group C).

RESULTS: As a result patients with PAF had a significantly lower root mean square voltage of the last 20 ms (RMS 20) (2.59 +/- 0.89) vs 4.08 +/- 1.45 microV, p < 0.0003) and a significantly longer filtered P wave duration (FPD) than patients of the control collective (139.2 +/- 17.5 vs 115.1 +/- 17.7 ms, p < 0.0001) and the young volunteers (3.44 +/- 0.95 microV, p < 0.0001/101.9 +/- 14.2 ms, p < 0.009). Furthermore we found an age-dependent relationship of FPD between group B and C (115.1 +/- 17.7 vs 101.9 +/- 14.2 ms, p < 0.05) but not an age-dependent relationship of the RMS 20 (4.08 +/- 1.45 vs 3.44 +/- 0.95 microV, p = n.s.). A specificity of 80% and a sensitivity of 78% was achieved for identifying patients with atrial fibrillation by using a definition of atrial late potentials as FPD > 120 ms and a RMS 20 < 3.5 microV.

CONCLUSIONS: The analysis of the P-SAECG can be used as a non-invasive method for identifying atrial late potentials. Atrial late potentials might be a reason for PAF. The predictive power of atrial late potentials has to be examined by prospective investigations of a larger patient population.