BACKGROUND: Interatrial block (IAB; P waves >/= 110 milliseconds), the conduction delay between the right (RA) and left atrium (LA), is depicted on the electrocardiogram (ECG) as prolonged, often bifid ("notched"), P waves with distinguishable RA and LA components. Although electrophysiologic (EP) studies give some insight on how RA and LA components are depicted on the surface ECG in normal conduction, few if at all any, have conclusively demonstrated this correlation with IAB. Using existing EP knowledge, we investigated if such P-wave markers on bedside ECGs exist in IAB and appraised their utility in IAB recognition. METHODS: We reviewed the medical records of patients admitted to a general hospital from December 1, 2004, to December 15, 2004. Of those, 151 patients had been admitted for nonacute presentations and were screened with 12-lead ECGs. Thirty-eight ECGs were excluded for nonsinus and paced rhythms, severe motion artifact, errors in lead placement, absence of adequate patient identification, and duplicate patient admissions after discharge. The remaining 113 ECGs were then evaluated for IAB. Sixty-three patients who did not have IAB formed the control (group A), whereas of the remaining 50 patients with IAB, 24 who had past ECGs for comparison formed group B1 and 26 without past ECGs formed group B2. Groups were compared for common clinical comorbidities, whereas sensitivity and specificity were calculated for significant P-wave markers. P values were also calculated, with a value of <.05 considered significant. RESULTS: Clinical characteristics of patients in all groups were statistically comparable. Overall, almost all P waves in patients with IAB (groups 1 and 2) appeared "notched" (94%, P < .0001; sensitivity, 75%; specificity, 94% for IAB recognition; positive predictive value, 94%). P-wave RA components were commonly depicted as "domes," whereas their LA counterparts formed "spikes" (48%, P < .0001; sensitivity 96%; specificity, 70% for IAB recognition). When groups B1 and B2 were compared with increased accuracy, more P waves in group B1 were noted to have notches and had easily discernible RA and LA components; often, the RA duration is longer than the LA duration. In addition, more "dome-and-spike" complexes could be determined when past ECGs were present for comparison. These markers could be found on any bedside ECG lead in IAB but were predominant on leads II and V3 to V6. CONCLUSIONS: Specific noninvasive surface markers such as P-wave "dome-and-spike" complexes and "notches" in any lead (predominantly leads II and V3-V6) on the bedside ECG could alert clinicians to measure P waves and so identify IAB.
BACKGROUND: Interatrial block (IAB; P waves >/= 110 milliseconds), the conduction delay between the right (RA) and left atrium (LA), is depicted on the electrocardiogram (ECG) as prolonged, often bifid ("notched"), P waves with distinguishable RA and LA components. Although electrophysiologic (EP) studies give some insight on how RA and LA components are depicted on the surface ECG in normal conduction, few if at all any, have conclusively demonstrated this correlation with IAB. Using existing EP knowledge, we investigated if such P-wave markers on bedside ECGs exist in IAB and appraised their utility in IAB recognition. METHODS: We reviewed the medical records of patients admitted to a general hospital from December 1, 2004, to December 15, 2004. Of those, 151 patients had been admitted for nonacute presentations and were screened with 12-lead ECGs. Thirty-eight ECGs were excluded for nonsinus and paced rhythms, severe motion artifact, errors in lead placement, absence of adequate patient identification, and duplicate patient admissions after discharge. The remaining 113 ECGs were then evaluated for IAB. Sixty-three patients who did not have IAB formed the control (group A), whereas of the remaining 50 patients with IAB, 24 who had past ECGs for comparison formed group B1 and 26 without past ECGs formed group B2. Groups were compared for common clinical comorbidities, whereas sensitivity and specificity were calculated for significant P-wave markers. P values were also calculated, with a value of <.05 considered significant. RESULTS: Clinical characteristics of patients in all groups were statistically comparable. Overall, almost all P waves in patients with IAB (groups 1 and 2) appeared "notched" (94%, P < .0001; sensitivity, 75%; specificity, 94% for IAB recognition; positive predictive value, 94%). P-wave RA components were commonly depicted as "domes," whereas their LA counterparts formed "spikes" (48%, P < .0001; sensitivity 96%; specificity, 70% for IAB recognition). When groups B1 and B2 were compared with increased accuracy, more P waves in group B1 were noted to have notches and had easily discernible RA and LA components; often, the RA duration is longer than the LA duration. In addition, more "dome-and-spike" complexes could be determined when past ECGs were present for comparison. These markers could be found on any bedside ECG lead in IAB but were predominant on leads II and V3 to V6. CONCLUSIONS: Specific noninvasive surface markers such as P-wave "dome-and-spike" complexes and "notches" in any lead (predominantly leads II and V3-V6) on the bedside ECG could alert clinicians to measure P waves and so identify IAB.
Authors: Vignendra Ariyarajah; Mark Kranis; Sirin Apiyasawat; David H Spodick Journal: Ann Noninvasive Electrocardiol Date: 2007-01 Impact factor: 1.468