Primary care REFerral for EchocaRdiogram (REFER) in heart failure: a diagnostic accuracy study

Clare J Taylor, Andrea K Roalfe, Rachel Iles, FD Richard Hobbs, The REFER investigators, P Barton, J Deeks, D McCahon, Cowie, G Sutton, RC Davis, J Mant, T McDonagh, L Tait

Symptoms of breathlessness, fatigue, and ankle swelling are common in general practice but deciding which patients are likely to have heart failure is challenging. To evaluate the performance of a clinical decision rule (CDR), with or without N-Terminal pro-B type natriuretic peptide (NT-proBNP) assay, for identifying heart failure. Prospective, observational, diagnostic validation study of patients aged >55 years, presenting with shortness of breath, lethargy, or ankle oedema, from 28 general practices in England. The outcome was test performance of the CDR and natriuretic peptide test in determining a diagnosis of heart failure. The reference standard was an expert consensus panel of three cardiologists. Three hundred and four participants were recruited, with 104 (34.2%; 95% confidence interval [CI] = 28.9 to 39.8) having a confirmed diagnosis of heart failure. The CDR+NT-proBNP had a sensitivity of 90.4% (95% CI = 83.0 to 95.3) and specificity 45.5% (95% CI = 38.5 to 52.7). NT-proBNP level alone with a cut-off <400 pg/ml had sensitivity 76.9% (95% CI = 67.6 to 84.6) and specificity 91.5% (95% CI = 86.7 to 95.0). At the lower cut-off of NT-proBNP <125 pg/ml, sensitivity was 94.2% (95% CI = 87.9 to 97.9) and specificity 49.0% (95% CI = 41.9 to 56.1). At the low threshold of NT-proBNP <125 pg/ml, natriuretic peptide testing alone was better than a validated CDR+NT-proBNP in determining which patients presenting with symptoms went on to have a diagnosis of heart failure. The higher NT-proBNP threshold of 400 pg/ml may mean more than one in five patients with heart failure are not appropriately referred. Guideline natriuretic peptide thresholds may need to be revised.