A critical re-appraisal of different ways of selecting ambulatory patients with suspected heart failure for echocardiography

Background: ECG and BNP have been assessed as screening tests for LVSD and heart failure. However, echocardiography also provides information about valvular disease and LVH. We assessed how good these screening tests are in identifying whether the subsequent echocardiogram will have any significant abnormality. Aims: To re-appraise the ECG and BNP as screening tests for echocardiography since there are important practical deficiencies in our current knowledge in this area. Methods: General practitioners referred suspected heart failure patients for clinical assessment, echocardiography, electrocardiography, and BNP measurement. The accuracy of each screening test and combinations of screening tests were calculated for LVSD, heart failure, valvular disease, and LVH. Results: The sensitivities of the ECG for LVSD, heart failure, LVH and valvular disease were 97%, 95%, 76%, and 69%, respectively. The corresponding figures for BNP were 86%, 82%, 59%, and 48%, respectively. When patients with atrial fibrillation and murmurs were excluded, the values for ECG were 94%, 87%, 53%, and 55%, while for BNP they were 83%, 73%, 50%, and 32%. Conclusions: ECG interpretation and BNP are adequate screening tests to detect LVSD or heart failure but fail to screen for other echocardiographic abnormalities, like valvular disease and LVH. This remains the case even if patients with atrial fibrillation or heart murmurs are excluded on the basis that they require echocardiography anyway.