The Diagnosis

This patient had a large pulmonary embolism (PE). His chest radiograph demonstrated a positive Westermark's sign, and his ventilation/perfusion scan was read as having a high probability for pulmonary embolism. The initial ECG is suggestive of the diagnosis of PE, featuring both sinus tachycardia and an S1Q3T3 pattern, meaning a prominent S wave in lead I, a Q wave in lead III, and an inverted T wave in lead III. Pulmonary embolism can be an elusive diagnosis. Despite important advances in the diagnosis and treatment of patients with suspected PE, mortality and recurrence rates remain high.1 The 12-lead ECG is an easy, noninvasive bedside test that is frequently employed when evaluating patients with cardiopulmonary complaints. The ECG has inadequate sensitivity and specificity for the diagnosis of PE.2 It should be used to investigate the causes of chest pain, dyspnea, and related symptoms, and only as an adjunct during the assessment of patients in which the diagnosis of PE is entertained, never to exclude the diagnosis of PE. This review describes some of the traditional electrocardiographic findings associated with PE, beginning with S1Q3T3. S1Q3T3 Pattern This finding is perhaps the most common reflex answer to the ED teaching rounds question, “What do you see on the ECG in patients with PE?” This classic finding has not been found with such reassuring frequency in many subsequent case series since it was first described more than 65 years ago in seven patients with acute cor pulmonale secondary to PE; this electrocardiographic pattern results from acute right heart strain.2,3 Reviewing three major case series and comparing the results with their own data, Ferrari and colleagues found the incidence of S1Q3T3 in PE to range from 12 to 50 percent.4 However, all three components of S1Q3T3 may not appear simultaneously.5 Moreover, the dynamic nature of this finding (as well as that of others seen with PE) further makes reporting a fixed incidence a difficult endeavor. Some have emphasized the lack of established criteria with regard to the threshold amplitude of the S and Q waves seen with S1Q3T3.6 Others4 have employed the criteria advanced by McGinn and White: S wave in lead I and Q wave in lead III, with an amplitude of more than 1.5 mm, associated with inversion of the T wave in lead III.3 However, in a study looking at ECG changes in patients with suspected PE (not confirmed PE, as most studies have used), S1Q3T3 was found with equivalent frequency (approximately 12%) in patients both with and without PE.7 Other ECG Findings Although commonly associated with myocardial ischemia, precordial T wave inversion has been linked to PE. In a series of 80 patients with confirmed PE, precordial T wave inversion was the most common electrocardiographic abnormality, seen in 68 percent of cases.4 This finding exceeded the frequency of sinus tachycardia (26%) and S1Q3T3 (50%) in that series.Figure: This ECG shows sinus tachycardia at 117/minute with a right axis deviation; the latter is due to the prominent S wave in lead I. A small Q wave is seen in lead III, along with an inverted T wave in the same lead. There is a “clockwise” shift in the normal precordial transition zone, as R does not exceed S in amplitude until V5.Using right heart catheterization data, the authors found precordial T wave inversion to correlate directly with the severity of PE, a finding that has been reported by others8 using nuclear scintigraphic findings. There was a high frequency of this finding in the three other case series of confirmed PE that Ferrari et al reviewed, ranging from 46 to 89 percent.4 A recent review by Chan and colleagues found sinus tachycardia to be present in eight to 69 percent of patients with PE (range derived from six studies).2 The wide range is disconcerting, considering the traditional teaching of sinus tachycardia being a common finding in PE. Pulmonary embolism can be an elusive diagnosis, and despite important advances, mortality and recurrence rates are high Looking again at patients with suspected PE, Rodgers and associates found sinus tachycardia and incomplete right bundle branch block to be the only electrocardiographic findings significantly associated with confirmed PE; however, they cautioned that no electrocardiographic finding has sufficient positive or negative predictive value to confidently predict or exclude PE.7 To our knowledge, no study has looked at the incidence of sinus tachycardia in PE patients taking heart rate-controlling medications (e.g., beta-adrenergic blockers, calcium channel blockers, digoxin). Intuitively it seems that those medications might mask an underlying tachycardia. Chan et al reported right atrial strain in two to 31 percent of patients with PE (data from nine cases series).2 Also known as “P pulmonale,” it is manifested by peaking of the P wave in lead II with an amplitude of greater than or equal to 2.5 mV. Right bundle branch block, both incomplete and complete, occurs in six to 67 percent of PE cases.2 Recently, 16 of 20 (80%) patients with autopsy-proven main trunk pulmonary artery embolism featured either incomplete (n=6) or complete (n=10) right bundle branch block. In contrast, 30 cases of autopsy-proven peripheral PE did not demonstrate any new right bundle branch block in the same series.6 Other findings linked with PE include the following: normal ECG, atrial fibrillation and flutter, new frontal plane QRS axis changes, shift in precordial transition zone to the left (so-called clockwise rotation), displacement of the PR segment, low voltage QRS, ST segment depression, and S1S2S3.2,9 These findings do not appear to be significantly more common among patients with confirmed PE than in those with unconfirmed PE, although in one report, the same was true for right atrial strain, right bundle branch block, and sinus tachycardia.9 No ECG finding, or combination of ECG findings, adequately predicts or excludes the diagnosis of PE. Studies that report the incidence of a given electrocardiographic finding in PE should be interpreted with caution in that these populations often consist of patients with confirmed PE. Extrapolation of positive and negative predictive values to our ED population — patients with suspected PE — is flawed. Electrocardiographic findings suggestive of PE should be considered as complementary data along with other clinical evidence (e.g., risk factors, history and physical examination, chest x-ray findings, blood gas results, nuclear or CT imaging) when trying to confirm or exclude the diagnosis of PE. The principal utility of the ECG in suspected PE is in the consideration or exclusion of other diagnoses that may cause similar symptoms (e.g., cardiac ischemia).