The trusted source for
healthcare information and
Sources: Li SF, et al. Electrocardiographic diagnosis of acute myocardial infarction in the presence of left bundle branch block. Ann Emerg Med 2000;36:561-565; and Sokolove PE, et al. Interobserver agreement in electrocardiographic diagnosis of acute myocardial infarction in patients with left bundle branch block. Ann Emerg Med 2000;36:566-571.
In the study by Li and associates, the authors performed a retrospective cohort trial investigating the use of the Sgarbossa criteria1 in emergency department (ED) patients with acute myocardial infarction (AMI) and electrocardiographic left bundle-branch block (LBBB) pattern. All patients admitted with suspected acute coronary ischemia and LBBB were eligible. Traditional definitions of AMI (using CPK-MB isoenzyme elevation) and LBBB (electrocardiographic criteria) were employed; the chronicity of the LBBB also was determined (new, old, or unknown). Three physicians, blinded to the clinical diagnosis, interpreted the LBBB ECGs using the Sgarbossa criteria as follows: 1) concordant ST segment elevation (STE) = 1 mm; 2) discordant STE = 5 mm; and 3) concordant ST segment depression (STD) in leads V1-V3. Of the 306 patients eligible for study entry, 116 were excluded because of immediate need for therapeutic intervention prior to diagnosis (22); incomplete cardiac enzyme set or medical records (25 and 8, respectively); or presentation not consistent with acute coronary ischemia (61). Data analysis was performed on the remaining 190 patients; 25 (13%) individuals ultimately experienced AMI.
All three criteria demonstrated extremely low sensitivity and extremely high specificity for AMI. Only two indices—concordant STE or having any of the three criteria—had positive likelihood ratios (LR+) greater than 1; concordant STE had a LR+ of 16, and having any of the three criteria demonstrated a LR+ of 3. Eleven patients (6%) had new LBBB, of which six had AMI. Inter-observer agreement for application of this rule among the study physicians was 91% or higher for each of the criteria among all subgroups. The authors concluded that the Sgarbossa criteria lacked sensitivity to exclude AMI in the ED.
In the Sokolove and colleagues study, inter- and intra-observer agreement were tested between emergency physicians (EPs) and cardiologists with regard to the Sgarbossa criteria for the diagnosis of AMI in the LBBB patient. Four EPs and four cardiologists interpreted 224 LBBB ECGs—100 with AMI confirmed by enzyme analysis and 124 without either electrocardiographic or clinical evidence of AMI; a subset of 25 ECGs was re-evaluated by each physician to determine intra-observer reliability. Inter-rater agreement using the Sgarbossa criteria was high, with kappa (k) coefficients of 0.81 for cardiologists, EPs, and all readers combined. The intra-observer reliability was stronger for cardiologists (k = 0.81) than for EPs (k = 0.71), but this difference was not statistically significant. The median sensitivity for EP diagnosis of AMI was 67% compared to a value of 73% for cardiologists. The authors concluded that the study physicians were able to use the criteria with excellent inter- and intra-observer reliability, and speculated that EPs should be able to use the criteria reliably in the evaluation of the chest pain patient with LBBB.
LBBB markedly reduces the diagnostic power of the ECG. In fact, until recently, common medical opinion stated that the electrocardiographic diagnosis of AMI was impossible in the presence of LBBB. The introduction of the Sgarbossa criteria in 1996 has led to a re-evaluation of this position. Li and associates concluded that the Sgarbossa criteria are not useful in excluding AMI in the ED. Other literature support this conclusion. The first investigation applied the criteria to patients with chest pain and LBBB, reporting a very low sensitivity for electrocardiographic diagnosis of AMI coupled with poor interobserver reliability.2 A second study investigated the diagnostic and therapeutic impact of these criteria, and found that none effectively distinguished patients with AMI from patients with noncoronary diagnoses; the authors concluded that electrocardiographic criteria are poor predictors of AMI in the presence of LBBB and suggested that all patients suspected of AMI with LBBB should be considered for thrombolysis.3 A third investigation followed just this recommendation.4 A thrombolytic agent was used in all patients with LBBB presumed to have AMI, and they reported an alarmingly high rate of inappropriate thrombolysis. The authors also retrospectively investigated the impact of the criteria on diagnosis and management.4 These investigators, in contrast to the previously noted reports, found significant accuracy using the Sgarbossa criteria—approximately an 80% correct diagnosis rate using the prediction rule. The authors suggested that, had the prediction rule been employed, inappropriate thrombolysis would have been avoided in many instances.
These ECGs are complicated and the patient care scenarios are challenging; thus, no easy answer is available to the clinician. A LBBB pattern does not entirely invalidate ECG use, it merely reduces the utility of the ECG in arriving at the correct diagnosis. As long as the EP recognizes this limitation, the ECG can be used in the ED evaluation of the potential AMI patient with LBBB. The study by Li et al only reinforces this statement. The clinician must realize that when considering patients with chest pain, LBBB pattern, and clinically suspected AMI, only a minority will manifest obvious electrocardiographic abnormalities as reported by the Sgarbossa group. As in all patients with suspected coronary ischemia, a nondiagnostic ECG in no way excludes the diagnosis of AMI. The Sokolove et al study suggests that these criteria can be reliably used by EPs to evaluate patients with LBBB on ECG for coronary ischemia.
1. Sgarbossa EB, et al. Electrocardiographic diagnosis of evolving acute myocardial infarction in the presence of left bundle-branch block. N Engl J Med 1996;334: 481-487.
2. Shapiro NI, et al. Validation of electrocardiographic criteria for diagnosing acute myocardial infarction in the presence of left bundle branch block. Acad Emerg Med 1998;5:508.
3. Shlipak MG, et al. Should the electrocardiogram be used to guide therapy for patients with left bundle-branch block and suspected myocardial infarction? JAMA 1999;281:714-719.
4. Edhouse JA, et al. Suspected myocardial infarction and left bundle branch block: Electrocardiographic indicators of acute ischaemia. J Accid Emerg Med 1999;16:331-335.