OBJECTIVES: Interpretation of stress echocardiography is subjective, and highly dependent on the experience of the interpreter. We sought to evaluate whether a cardiologist without any previous experience in stress echocardiography could adequately learn the skills of interpreting dobutamine stress echocardiograms (DSE) in post-infarct patients, after a period of systematic training. METHODS: A trainee in cardiology blindly reported 51 consecutive DSEs from a database of post-infarction studies, after 2 and 4 months of systematic training. We compared his interpretation with that of an expert. RESULTS: Agreement between the trainee and the expert improved significantly from 2 to 4 months of training in the left anterior descending artery territory for the overall scan interpretation (from kappa = 0.58 to kappa = 0.73; p = 0.03), wall thickening assessment in individual segments (from kappa = 0.40 to kappa = 0.55; p < 0.01) and the diagnosis of viable myocardium (from kappa = 0.11 to kappa = 0.43; p = 0.01). Similar improvement was observed in left circumflex, but not in the right coronary artery territory. Agreement in identifying inducible ischaemia also remained poor. CONCLUSION: This study suggests that systematic training can significantly reduce interobserver variability in a short time frame (4 months) and may improve the interpretation of DSE by a trainee. But improvements in image quality and use of predefined reading criteria are necessary to improve interobserver agreement further in myocardial regions where conformity in dobutamine stress echocardiographic interpretation is low.
OBJECTIVES: Interpretation of stress echocardiography is subjective, and highly dependent on the experience of the interpreter. We sought to evaluate whether a cardiologist without any previous experience in stress echocardiography could adequately learn the skills of interpreting dobutamine stress echocardiograms (DSE) in post-infarctpatients, after a period of systematic training. METHODS: A trainee in cardiology blindly reported 51 consecutive DSEs from a database of post-infarction studies, after 2 and 4 months of systematic training. We compared his interpretation with that of an expert. RESULTS: Agreement between the trainee and the expert improved significantly from 2 to 4 months of training in the left anterior descending artery territory for the overall scan interpretation (from kappa = 0.58 to kappa = 0.73; p = 0.03), wall thickening assessment in individual segments (from kappa = 0.40 to kappa = 0.55; p < 0.01) and the diagnosis of viable myocardium (from kappa = 0.11 to kappa = 0.43; p = 0.01). Similar improvement was observed in left circumflex, but not in the right coronary artery territory. Agreement in identifying inducible ischaemia also remained poor. CONCLUSION: This study suggests that systematic training can significantly reduce interobserver variability in a short time frame (4 months) and may improve the interpretation of DSE by a trainee. But improvements in image quality and use of predefined reading criteria are necessary to improve interobserver agreement further in myocardial regions where conformity in dobutamine stress echocardiographic interpretation is low.