Clinical outcome implications of novel auscultation device of artificial intelligence (NAVI-HF) in predischarge acute heart failure patients

Aim This study evaluated the prognostic significance of a novel auscultation device of artificial intelligence (NAVI-HF) in predicting clinical outcomes in predischarge acute heart failure (HF) patients. Methods This second, prognostic phase of a prospective observational cohort study extends our prior diagnostics phase, in which NAVI-HF demonstrated good performance compared with lung ultrasound (LUS) B-lines as the reference standard. Patients were classified as NAVI-HF positive or negative prior to discharge and followed for 6 months for evaluation. Results A total of 239 patients were included of whom 111 (45.1%) were NAVI-HF positive. NAVI-HF positive patients had a worse clinical profile and higher NT-proBNP ( p < 0.05). During a median follow-up of 6 months, 82 composite events (34.3%) occurred. In univariate Cox analysis, NAVI-HF positive was associated with significant increased risk of composite outcome (HR 1.59, CI 1.03–2.47; p 0.03) and HF hospitalization (HR 1.64, CI 1.05–2.46; p 0.02), but not all-cause mortality (p 0.24). The association in composite and HF hospitalization outcomes remained significant after adjustment for age and sex but was attenuated after adjustment for other prognostic variables. Conclusion NAVI-HF–detected congestion was associated with increased adverse clinical outcomes in predischarge acute HF patients.