Introduction In patients presenting with chest pain, the presence of Left Bundle Branch Block (LBBB) on the ECG may obscure the diagnosis of acute myocardial infarction (AMI). Patients with LBBB caused by AMI are shown to benefit significantly from rapid provision of reperfusion therapy, yet evidence suggests this is often underprovided. Difficulties in the identification of AMI in these patients is the most commonly cited reason.

Aims To determine whether the application of the simplified Sgarbossa criteria to undifferentiated chest pain patients presenting with ECG changes of LBBB, transmitted during the prehospital phase of care, will be positively predictive of a discharge diagnosis of AMI.

Methods 785 ECGs were reviewed retrospectively and categorised based on their features. Those displaying LBBB and complete patient data were evaluated using the simplified Sgarbossa criteria and a decision made as to whether the patient had suffered AMI, a process which was blind peer reviewed. Results of cardiac biomarkers were traced and the diagnosis of AMI confirmed or refuted.

Results Of the 785 ECGs reviewed and categorised 56 (7.1%) displayed the features of LBBB and 18 were eligible for inclusion in the study. The sensitivity and specificity of the Sgarbossa criteria were 50% and 75% respectively. Inter-observer reliability was κ=0.93 (almost perfect).

Conclusion Despite excellent inter-observer agreement, the Sgarbossa criteria is too insensitive to be implemented in clinical practice as it cannot be relied upon to identify correctly, patients who have suffered AMI. Instead, diagnosis and management should proceed according to clinical judgement with reperfusion therapy administered following evaluation of risk verses benefit.