To the Editor: Rosenberger et al. (1) described the limited utility of intraoperative transesophageal echocardiography (TEE) as a primary diagnostic tool for obtaining a definitive diagnosis of pulmonary embolism (PE). We wish to highlight the role of the electrocardiogram (ECG) in diagnosing PE. A 51-yr-old patient presented to the operating room for reduction of a hip dislocation. The patient's preoperative ECG, obtained earlier the same day, revealed sinus bradycardia (Fig. 1). Immediately before a planned induction of general anesthesia, we attached an ECG monitor and noted a new right bundle branch block (RBBB). We postponed induction and obtained a 12-lead ECG in the operating room (Fig. 2). The ECG revealed new findings, including normal sinus rhythm, (RBBB), multiple ST segment abnormalities, S1Q3T3 pattern, and T wave inversion V1-V3. A subsequent computerized tomography (CT) scan revealed multiple, deep-venous thrombi and massive PE.Figure 1.: Preoperative electrocardiogram obtained 12 hours prior to planned surgery. Sinus bradycardia.Figure 2.: Electrocardiogram obtained in the operating room. Normal sinus rhythm, right bundle branch block, multiple ST segment abnormalities, S1Q3T3 pattern, and T wave inversion V1-V3.The usefulness of ECG findings in the diagnosis of PE is controversial (1). As with clinical presentation and physical examination, an ECG is of limited value in diagnosing PE: findings are often inconsistent and nonspecific (1). ECG changes may vary with the size of the embolus, impact upon hemodynamics, and cardiopulmonary reserve (1,2). Classically described findings are often absent, or they are equally prevalent in patients suspected of having PE in whom an alternative diagnosis is ultimately made (2). Although approximately 90% of patients with PE have at least one abnormality on ECG, many of these findings are nonspecific and therefore limited in their diagnostic utility in cases of suspected PE (3). Given the widespread use of imaging modalities for diagnosis of PE, the primary function of the ECG in modern clinical practice is to raise or reinforce suspicion (4). In our case it was a new RBBB, indicative of impairment in right-sided cardiac conduction, that prompted further evaluation. Its significance for diagnosing PE is unclear. One study found new RBBB in 80% of patients with massive trunk embolism but in no cases of peripheral embolism in the pulmonary artery (4). Another study found RBBB present in 10% of patients with confirmed PE, although the severity of PE failed to correlate with the presence or absence of RBBB (5). In that same study, a pattern of subepicardial ischemia (T-wave inversions) in the precordial leads was the finding most often noted in PE and also the finding most closely correlated with severity. Inverted T-waves, however, present in many other processes, are thus of limited diagnostic utility. In another study, 28 different ECG abnormalities were analyzed in patients suspected of having PE who presented for imaging studies (2). Only tachycardia (42%) and incomplete RBBB (7%) occurred significantly more often in patients who were confirmed by imaging to indeed have PE than patients with other etiologies. In the same study complete RBBB was found in only 4% of cases of confirmed PE, and it was not diagnostic. Even the classic S1Q3T3 pattern, with a reported prevalence of 12% in massive PE, may occur frequently enough in patients falsely suspected of having PE so as to limit its usefulness as a diagnostic criterion (2,5,6). Statistical analysis is complicated by the numerous possible combinations of findings that may exist on any given ECG. Given that the ECG alone is of limited value in the diagnosis of PE, clinicians must consider the clinical presentation in its entirety (7). Typically, it is prudent to pursue confirmatory imaging studies, such as a CT scan, before proceeding with definitive treatment for PE. The article by Rosenberger et al. indicates that TEE lacks the sensitivity to play this role. The most important role of the ECG is to trigger further evaluation that can lead to definitive diagnosis and prompt treatment of this life-threatening condition. Greg Ginsburg, MD Neelakantan Sunder, MD Priscilla Grace Harrell, MD Department of Anesthesia and Critical Care Massachusetts General Hospital Boston, Massachusetts [email protected]
Read full abstract