Sgarbossa NEJM 334:8 481-7 Feb 1996
Rule of appropriate discordance:
In LBBB with out STEMI the major terminal portion of QRS and initial upsloping portion of the ST/T segment should lie on opposite sides of isoelectric baseline = "discordance"
Sgarbossa Criteria for MI in presence of LBBB
Scores ≥ 3 have 90% specificity for AMI (but sensitivity of 20%)
Fig 1: ECG meeting all three independent criteria of Sgarbossa and colleagues for the diagnosisof acute MI with LBBB. The ECG shows at least 1-mm concordant ST-elevation in lead II, at least 1-mm ST depression in leads V2 and V3,as well as discordant ST-elevation of at least 5 mm in leads III and aVF (Reproduced from Sgarbossa EB, Pinski SL, Barbagelata A, et al, for the GUSTO-1 investigators.Electrocardiographic diagnosis of evolving acute myocardial infarction in the presenceof left bundle branch block. N Engl J Med 1996;334:481–7
cf Normal LBBB:
Fig 2: In uncomplicated LBBB, ECG leads with a predominantly negative QRS complex show ST-segment elevation, those with a positive SRS show ST-segment depression ("discordance") and there is no ST depression anteriorly
Sgarbossa criteria — A large trial of thrombolytic therapy for acute MI (GUSTO-1) provided an opportunity to revisit the issue of the electrocardiographic diagnosis of evolving acute MI in the presence of LBBB . Among 26,003 North American patients who had a myocardial infarction confirmed by enzyme studies, 131 (0.5 percent) had LBBB. A scoring system, often called the Sgarbossa criteria, was developed from the coefficients assigned by a logistic model for each independent criterion, on a scale of 0 to 5.
The three ECG criteria with an independent value in the diagnosis of acute infarction and the score for each were:
ST segment elevation of 1 mm or more that was in the same direction (concordant) as the QRS complex in any lead — score 5.
ST segment depression of 1 mm or more in any lead from V1 to V3 — score 3.
ST segment elevation of 5 mm or more that was discordant with the QRS complex (ie, associated with a QS or rS complex) — score 2.
A minimal score of 3 was required for a specificity of 90 percent. The first two criteria are similar to those described above since the ST segment is concordant rather than discordant with the QRS complex. However, the third finding requires further validation, since a high take-off of the ST segment in leads V1 to V3 has been described with uncomplicated LBBB, particularly if there is underlying left ventricular hypertrophy. In a substudy from the ASSENT 2 and 3 trials, the third criteria added little diagnostic or prognostic value .
A number of the other criteria, including those of Wackers mentioned above , were not found to be useful. A difference may be that the GUSTO criteria were applied within hours of the MI as part of the GUSTO-1 trial, while those of Wackers were often applied much longer after the acute event. Furthermore, only the Wackers study attempted to localize the infarct.
A Sgarbossa score of ≥3 was highly specific (ie, few false positives) but much less sensitive (36 percent) in the validation sample in the original report . Similar findings were noted in a subsequent meta-analysis of 10 studies of 1614 patients in which a Sgarbossa score of ≥3 had a sensitivity of 20 percent and a specificity of 98 percent . The sensitivity may increase if serial or previous ECGs are available .
In addition to their utility in diagnosis, the Sgarbossa criteria may also predict prognosis in patients with acute MI.