Screening to prevent heart failure (STOP-HF): Expanding the focus beyond asymptomatic left ventricular systolic dysfunction

Abstract

Aims We evaluated the extent to which left ventricular diastolic dysfunction (LVDD) contributes to the high false-positive rates observed when natriuretic peptides (NPs) are used to screen for left ventricular systolic dysfunction (LVSD), and the use of NPs in combination with electrocardiogram (ECG) to screen for pre-clinical ventricular dysfunction (PCVD). Methods and results Eight hundred and fourteen patients over 40 years of age and with at least one cardiovascular risk factor were recruited. Screening strategies for LVSD included brain natriuretic peptide (BNP) alone at cut-offs of 20, 50, and 100 pg/mL, and BNP and abnormal ECG combined. Systolic and diastolic function was assessed by Doppler echocardiography. A left ventricular ejection fraction (LVEF) of <50% was present in 33 (4.1%) of subjects, while 11 (1.4%) had LVEF <40%. At a cut-off of 20, 50, and 100 pg/mL, sensitivity for BNP alone when screening for LVSD was 88, 70, and 45%, and specificity 46, 77, and 90%, respectively. Of those labelled 'false positive' in the 20, 50, and 100 pg/mL cut-off groups, 26, 46, and 65%, respectively, were found to have significant LVDD (left atrial volume index >34 mL/m2). Optimal sensitivity (80%) and specificity (72%) for PCVD was obtained when BNP at a cut-off of 50 pg/mL or an abnormal ECG were defined as a positive screen so that only this group would be sent for Doppler echocardiography. Conclusions A significant number of patients at risk for LVSD and labelled false positive with screening were found to have LVDD. Identifying this at-risk cohort may improve outcomes, but the clinical and economic benefit of this screening strategy requires formal assessment. © 2012 The Author.