Applicability of current diagnostic algorithms in geriatric patients suspected of new, slow onset heart failure

BACKGROUND: referral for echocardiography for all geriatric outpatients suspected of heart failure (HF) is not feasible. Diagnostic algorithms could be helpful. OBJECTIVE: to investigate whether available diagnostic algorithms accurately identify (older) patients (aged 70 years or over) eligible for echocardiography, with acceptable numbers of false-negatives. METHODS: algorithms (European Society of Cardiology (ESC)) guideline, National Institute for Health and Clinical Excellence (NICE) guideline, multidisciplinary guideline the Netherlands (NL) and algorithm by Mant et al. were validated in 203 geriatric patients (mean age 82 ± 6 years, 30% men) suspected of new, slow onset HF. HF was adjudicated by an outcome panel. Applicability of algorithms was evaluated by calculating proportion of patients (i) referred for echocardiography, (ii) with HF among referred patients and (iii) without HF in the non-referred. RESULTS: ninety-two (45%) patients had HF. Applying algorithms resulted in referral for echocardiography in 52% (normal NT-proBNP; ESC), 72% (normal ECG; ESC), 56% (NICE), 93% (NL) and 70% (Mant) of all patients, diagnosing HF in 78, 56, 76, 49 and 62% of those referred, respectively. In patients not referred for echocardiography HF was absent in 90, 82, 93, 100 and 95%, respectively. CONCLUSION: the ESC NT-proBNP (<400 pg/ml)-based algorithm combines the lowest number of referrals for echocardiography (of whom 78% has HF) with a limited number (10%) of false negatives in the non-referred.