Signal-averaged P wave: predictor of atrial fibrillation.

Abnormalities of the P wave seen during sinus rhythm are associated with atrial fibrillation and other supraventricular arrhythmias. Intra-atrial conduction delays can be seen on the surface ECG as P wave prolongation, which is more visible with averaging techniques used in advanced recording devices. Averaging followed by amplification after proper filtering of the electrical signal should allow more precise measurements of duration and amplitude of the P wave. Data on reproducibility, filter settings, required number of beats, and precise definitions of onset and offset are dependent largely on the system used, which can be QRS or, preferably, P wave triggered. This explains conflicting data in the literature. It is clear that P wave duration is one of the best predictors of perioperative atrial fibrillation. For paroxysmal forms, the data are less convincing. Root mean square voltages of the P wave can be useful. Combining the P wave duration with other data often improves the diagnostic value of signal averaging. Standardization of the techniques appears to be necessary.

Analysis of the P wave with signal averaging to assess the risk of atrial fibrillation after coronary artery bypass surgery.

P wave signal averaging was performed in 91 consecutive patients undergoing coronary artery bypass grafting to detect patients at risk of postoperative atrial fibrillation (AF). Sixteen patients (17.5%) developed AF after surgery. The P wave duration on the signal-averaged electrocardiogram (ECG) and on surface ECG was prolonged in AF patients compared to others (respectively 141 +/- 12 vs. 132 +/- 12 ms and 124 +/- 9 vs. 113 +/- 9 ms). The root mean square voltages (RMS) of the total P wave were not different between the two groups; the RMS of the late portion of the P wave (late RMS) was significantly higher (0.25 +/- 0.15 vs. 0.17 +/- 0.10 microV) and the RMS of the first 110 ms of the P wave (early RMS) significantly lower (0.88 +/- 0.28 vs. 1.09 +/- 0.33 microV) in AF. The late/ early RMS ratio was different (0.29 +/- 0.16 vs. 0.17 +/- 0.11). In a multivariate analysis only age and the late/early RMS ratio were predictive for AF.