The REFER (REFer for EchocaRdiogram) study: a prospective validation and health economic analysis of a clinical decision rule, NT-proBNP or their combination in the diagnosis of heart failure in primary care

Background Heart failure is a treatable condition but making a diagnosis can be challenging. Objective To evaluate the performance of a clinical decision rule (CDR) with or without a natriuretic peptide assay for identifying heart failure in symptomatic patients presenting to primary care. Design Prospective, observational, diagnostic validation study and economic evaluation. Setting Twenty-eight general practices in central England, UK. Participants Primary care patients aged ≥ 55 years presenting with recent new-onset shortness of breath, lethargy or peripheral ankle oedema of > 48 hours’ duration. Instrument The CDR included a clinical element (male, history of myocardial infarction, crepitations at the lung bases and oedema) and N-terminal pro-B-type natriuretic peptide (NT-proBNP) blood test. The reference standard was an expert consensus panel of three cardiology specialists. Main outcome measure The main outcome measure was test performance of the CDR and the natriuretic peptide test alone, and in combination, in estimating sensitivity and specificity, positive predictive value (PPV) and negative predictive value (NPV) for a diagnosis of heart failure. Economic evaluation of a decision tree with a NHS/Personal Social Services perspective determined the cost per quality-adjusted life-year (QALY) gained. Results In total, 304 participants were recruited to the validation cohort. The mean age was 73.9 years (standard deviation 8.8 years) and 124 (40.8%) participants were male. In total, 104 [34.2%, 95% confidence interval (CI) 28.9% to 39.8%] had a confirmed diagnosis of heart failure. The CDR had a sensitivity of 90% (95% CI 83% to 95%), specificity of 46% (95% CI 39% to 53%), PPV of 46% (95% CI 39% to 53%) and NPV of 90% (95% CI 83% to 95%). NT-proBNP level alone with a cut-off point of < 400 pg/ml had a sensitivity of 77% (95% CI 68% to 85%) and specificity of 92% (95% CI 87% to 95%). At the lower cut-off point of 125 pg/ml, sensitivity was 94% (95% CI 88% to 98%) and specificity was 49% (95% CI 42% to 56%). The economic model results suggest that referring a patient for a confirmatory diagnosis if they have had a previous myocardial infarction or have a NT-proBNP level that is greater than a 400 pg/ml threshold (current practice in England) is the most cost-effective option, with a cost of £4400 per QALY gained compared with a do nothing strategy. The base-case results were robust to deterministic and probabilistic sensitivity analyses. Conclusions Natriuretic peptide testing alone performed as well as the validated CDR in determining which patients presenting with symptoms went on to have a diagnosis of heart failure. The current NT-proBNP cut-off point of 400 pg/ml used in the UK is too high and means that one in five patients with heart failure may not be appropriately referred for further investigation and diagnosis, but this threshold was cost-effective in the REFer for EchocaRdiogram (REFER) trial. The study found only three patients with heart failure with reduced ejection fraction (HFREF), which might limit the benefits of early detection. The other diagnostic strategies with lower NT-proBNP referral levels become more cost-effective as the proportion of HFREF patients increases. International consensus on the optimal cut-off point for natriuretic peptide testing in patients with symptoms suggestive of heart failure should be sought. Trial registration Current Controlled Trials ISRCTN17635379. Funding This project was funded by the Efficacy and Mechanism Evaluation (EME) programme, a MRC and NIHR partnership.