Screening for left ventricular dysfunction may have less value than thought

June 11, 2009

The value and cost-effectiveness of screening for left ventricular (LV) dysfunction remains unclear, particularly since specific, evidence-based treatments are not available for the majority of patients with preserved systolic dysfunction, reports a study in the June issue of the Journal of Cardiac Failure.

In the study,1012 primary care patients with hypertension and/or diabetes without signs or symptoms of heart failure were screened for asymptomatic left ventricular dysfunction (ALVD), using measurements of NT-pro-BNP and echocardiography. Diastolic dysfunction was found in 368 subjects (36%) and was categorized as mild in 327 and moderate-severe in 41. Systolic dysfunction was present in only 11 (1.1%)l. NT-proBNP levels were 170±206 and 859±661 pg/mL respectively in diastolic and systolic dysfunction and 92±169 in normal subjects (p<.0001).

For the 52 subjects (5.1%) with moderate to severe diastolic dysfunction or systolic dysfunction, a NT-proBNP of < 125 pg/ml had a negative predictive value (npv) >99% and a positive predictive value of 33% in patients < 67 years. for older patients, npv was 100%, but ppv was somewhat lower in women (23%) than in men (33%).

This study suggests that the evaluation of NT-proBNP in asymptomatic patients with type 2 diabetes or hypertension may lead to very early exclusion of LV dysfunction. As a perspective of this study, general practitioners could use NT-proBNP determination to rule out heart failure (HF) in these patients, a much more cost effective measure than the use of an echocardiographic one. The study was partially supported by an unrestricted grant from Roche Diagnostics.

"This study demonstrates that low values of NT-proBNP usually exclude significant LV dysfunction, but elevated values are relatively non-specific, especially in a population where systolic dysfunction is rare," comments Barry M. Massie, M.D., Editor-in-Chief of the Journal of Cardiac Failure.

Source: Elsevier