The provocative study by Stein et al titled “Electrocardiogram in Pneumonia” merits additional clinical commentary. It serves as a reverse review of the investigators’ previous discussions of electrocardiographic (ECG) findings in acute pulmonary embolism. Unfortunately, the “clinical clouds” remain undissipated with regard to offering assistance in using electrocardiography in the differential diagnosis between pneumonia and pulmonary embolism. As indicated, the strength of the study is predominantly provided by the formidable list of exclusion criteria. Its weakness includes the small number of patients classified as having pneumonia and the lack of classification of subsets of patients comparing age with the ECG abnormalities. This could perhaps indicate the presence of undiagnosed heart disease, preferentially that of coronary artery disease.

The investigators list pulmonary embolism as 1 of the exclusion criteria but fail to reveal how this diagnosis was excluded. Also, a quantitative and qualitative description of the pulmonary parenchymal infiltrates could have added important data in comparing the ECG abnormalities with clinical status to differentiate the ECG changes ascribed to pneumonia from other significant factors of causation.

It seems logical that significant ECG changes would be unlikely in patients with minimal or insignificant pulmonary infiltrates, especially in the absence of underlying cardiac or pulmonary disease or other co-morbid conditions.

Is it a correct assumption that the 4 ventilator-dependent patients listed in Table 2 were the patients who died? Also of importance would be the presence of diabetes mellitus as a subset. This was not noted as to its exclusion or presence.

In Table 4, did the patients who had new findings within 1 month also have ECG abnormalities other than those listed in the table, and were they discharged with the same ECG abnormalities other than the new abnormalities believed to be associated with the diagnosis of pneumonia?

The electrocardiogram obtained 12 hours after admission (Figure 1) reveals an increase in the PR interval from 0.16 to 0.20 seconds, probably of little significance. The electrocardiogram obtained 1 day after admission (Figure 2) reverts to a minimal inferior repolarization abnormality which may not be related to the axis shift. Of greater significance, the electrocardiogram obtained 1 day after admission (Figure 3) changed from a normal repolarization pattern to one of a minimal generalized repolarization abnormality in addition to the stated reversion of the right atrial abnormality to normal.

The investigators stated that 3 patients had complete right bundle branch block and the S ₁ Q ₃ T ₃ pattern (4.8% of patients). It is accepted that the S ₁ Q ₃ T ₃ pattern may not be a common presentation in acute pulmonary embolism, but I suspect that it would be of lesser presentation in uncomplicated pneumonia.

The presence of S ₁ Q ₃ T ₃ , right bundle branch block, and P-pulmonale are ECG abnormalities usually associated with cor pulmonale and pulmonary hypertension. Their presence may be indicative of significant underlying pathology consistent with cardiac or pulmonary disease. The S ₁ Q ₃ T ₃ ECG pattern, when combined with right ventricular dilatation has been stated as having a 23% to 69% positive predictive value for the presence of pulmonary hypertension caused by acute pulmonary embolism.

In conclusion, I find it difficult to accept ECG abnormalities as being due solely to pneumonia, given our present state of knowledge and limited research on the subject. Stein et al’s continued interest in this matter is greatly appreciated and certainly warranted.