BACKGROUND: Pulmonary embolism (PE), a major cause of morbidity and mortality, remains an elusive diagnosis. Recently investigators have found a new electrocardiographic (ECG) finding, simultaneous T-wave inversions in the anterior and inferior leads, which may distinguish PE from acute coronary syndrome (ACS). OBJECTIVES: Our primary objective was to estimate the prevalence of this finding in PE. We also estimate the inter-rater reliability of this finding, its test characteristics, and assess ECG findings traditionally associated with PE. METHODS: In this unmatched case-control study, we selected electrocardiograms from patients diagnosed with PE, ACS, and non-cardiac chest pain. Two emergency physicians, blinded to diagnoses, reviewed electrocardiograms for explicitly defined ECG findings. We calculated kappa (K) for inter-rater agreement and estimated prevalence differences (PD) for findings in the PE group vs. pooled control groups. RESULTS: We included 97 patients with PE, 89 with ACS, and 105 with non-cardiac chest pain. A 1-mm T-wave inversion was seen in both III and V(1) in 11/97 (0.113) of patients with PE vs. 9/194 (0.046) controls (PD 0.07 [95% confidence interval (CI) -0.01-+0.14]; K = 0.7). Other criteria for anterior and inferior T-wave inversions were less common in PE (0.04-0.05). Among several other ECG abnormalities tested, only sinus tachycardia (PD 0.20 [95% CI 0.09-0.31]; K = 0.7) and the classic S(I)Q(III)T(III) pattern (PD 0.05 [95% CI -0.01-+0.11]; K = 0.5) statistically distinguished PE and were noted with fair or better inter-rater agreement. CONCLUSION: In our study, simultaneous T-wave inversions in anterior and inferior leads were associated with PE but are seen in only 4-11% of cases.
BACKGROUND:Pulmonary embolism (PE), a major cause of morbidity and mortality, remains an elusive diagnosis. Recently investigators have found a new electrocardiographic (ECG) finding, simultaneous T-wave inversions in the anterior and inferior leads, which may distinguish PE from acute coronary syndrome (ACS). OBJECTIVES: Our primary objective was to estimate the prevalence of this finding in PE. We also estimate the inter-rater reliability of this finding, its test characteristics, and assess ECG findings traditionally associated with PE. METHODS: In this unmatched case-control study, we selected electrocardiograms from patients diagnosed with PE, ACS, and non-cardiac chest pain. Two emergency physicians, blinded to diagnoses, reviewed electrocardiograms for explicitly defined ECG findings. We calculated kappa (K) for inter-rater agreement and estimated prevalence differences (PD) for findings in the PE group vs. pooled control groups. RESULTS: We included 97 patients with PE, 89 with ACS, and 105 with non-cardiac chest pain. A 1-mm T-wave inversion was seen in both III and V(1) in 11/97 (0.113) of patients with PE vs. 9/194 (0.046) controls (PD 0.07 [95% confidence interval (CI) -0.01-+0.14]; K = 0.7). Other criteria for anterior and inferior T-wave inversions were less common in PE (0.04-0.05). Among several other ECG abnormalities tested, only sinus tachycardia (PD 0.20 [95% CI 0.09-0.31]; K = 0.7) and the classic S(I)Q(III)T(III) pattern (PD 0.05 [95% CI -0.01-+0.11]; K = 0.5) statistically distinguished PE and were noted with fair or better inter-rater agreement. CONCLUSION: In our study, simultaneous T-wave inversions in anterior and inferior leads were associated with PE but are seen in only 4-11% of cases.