Screening for left ventricular dysfunction in the community: ready for prime‐time?

Should we be screening for LV dysfunction in the general community? Is there evidence that identifying individuals with asymptomatic dysfunction allows us to slow down the progression of disease and thereby improve such individuals’ quality and length of life—and can this be done at a reasonable return on the necessary investment of time and resources? In this issue of the journal, Mureddu and colleagues report the findings from a population-based study in middle-aged to elderly individuals living in the Lazio region of Italy. They used transthoracic echocardiography to identify LV dysfunction (either systolic or diastolic)—although in 15% they could not obtain adequate images—and then examined which screening strategy might work best to identify such patients. Six per cent of the population had clinically obvious heart failure, as expected in this relatively elderly population, and were excluded from the study. Of the remainder, the majority of patients (65%) had a history of some form of cardiovascular disease. However, the proportion with asymptomatic LV systolic dysfunction (EF ,50%) was low (1.5%). Many individuals had some diastolic dysfunction (42%), but this was graded as moderate or severe in only 6%. Measurement of plasma NT-proBNP provided moderately good diagnostic information, with the area under the receiver operating characteristic (ROC) curve being 0.83 for asymptomatic systolic dysfunction (22 cases) and 0.71 for asymptomatic moderate to severe diastolic dysfunction (88 cases). Adding the ECG did not add predictive value for either group. We are not told if such individuals with asymptomatic underlying LV disease could have been identified by a combination of clinical features alone, nor are we told the incremental value of NT-proBNP over such clinical features. Recent work from the Framingham Heart Study and from Europe suggests that there is a differential clustering of risk factors for individuals developing systolic, or diastolic, heart failure. We also do not know what proportion of the individuals with asymptomatic disease identified by Mureddu et al. were already on potentially disease-modifying therapy with an ACE inhibitor and/or a beta-blocker prescribed because of co-morbidity such as hypertension, AF, diabetes mellitus, stroke, or renal dysfunction. Two other European studies have published similar results for LV systolic dysfunction. Nielsen and colleagues screened 1257 individuals aged 25–74 years, and found 48 (3.8%) with severe systolic dysfunction (EF ,32%), with an area under the ROC curve of 0.86. Galasko and colleagues reported a prevalence of LV systolic dysfunction (EF ,45%) of 3.5% in the general London population aged 45 years or older, but some were already known to their general practitioner as having heart failure, LV systolic dysfunction, or were on loop diuretics. The cheapest strategy to identify such individuals was to pre-screen with either plasma NT-proBNP or a 12-lead ECG, with hand-held echocardiography thereafter. Such an approach would not be able to quantify diastolic abnormalities easily. However, data from the Framingham Heart Study suggested considerably less value for using natriuretic peptides to identify individuals likely to have asymptomatic LV systolic dysfunction (5.6% of the screened community population had an EF ≤50% and/or fractional shortening ,22%). Mureddu and colleagues also attempted to estimate the value for money of various screening strategies, using the intermediate measure of ‘cost per case’ identified. The costs for echocardiography and NT-proBNP measurement used in their calculations appear low compared with other European countries, but they concluded that it would costE1009 to find acase of asymptomatic LV systolic dysfunction, andE90 to find a case of asymptomatic moderate or severe diastolic dysfunction, if an NT-proBNP first strategy was employed, and transthoracic echocardiography performed for those with an abnormal NT-proBNP concentration. Such a strategy would miss more than a third of the cases of asymptomatic LV systolic dysfunction and 65% of the cases of asymptomatic LV diastolic dysfunction. The