AI Insight
Researchers tested a wrist-worn device that uses machine learning to estimate pulmonary capillary wedge pressure (a measure of fluid congestion in heart failure) through non-invasive venous waveform analysis. In both hospitalized heart failure patients undergoing diuretic treatment and a controlled porcine model, the device's NIVA Score decreased significantly after fluid removal, demonstrating it can track changes in volume status. The discharge NIVA Score also showed promise in predicting 30-day hospital readmission with an area under the curve of 0.85.
Why it matters
This technology could provide a practical tool for continuously monitoring fluid status in heart failure patients at home, potentially allowing earlier intervention before symptoms worsen and reducing hospital readmissions. Current methods for assessing congestion require invasive procedures or clinical examination, making frequent monitoring impractical.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
Residual congestion is the principal driver of heart failure readmission, and reliable serial assessment of volume status remains an unmet clinical need. This study asked whether a wrist-worn, machine-learning-based device for non-invasive venous waveform analysis in heart failure (the NIVAHF device), which produces an integer-scaled estimate of pulmonary capillary wedge pressure termed the NIVA Score, responds to acute changes in volume status. Agreement between the NIVA Score and invasively measured pulmonary capillary wedge pressure at single time points has been established in a separate prospective, multi-site study; however, such static agreement does not establish whether the measure tracks dynamic decongestion. We therefore evaluated the directional responsiveness of the locked NIVA Score in two prespecified cohorts: hospitalized adults with acute decompensated heart failure undergoing routine intravenous diuresis, and a controlled porcine model of volume overload followed by diuresis. In eleven patients contributing thirteen paired measurements (mean net fluid balance -2.1 {+/-} 1.0 L), NIVA Scores decreased significantly after diuresis (paired t-test, P = 0.04). In five pigs contributing twenty-four paired measurements, NIVA Scores decreased significantly after intravenous furosemide following crystalloid loading (P < 0.01), and the direction of change was concordant with measured urine output in every animal. Statistical significance was reached in both cohorts despite modest sample sizes, indicating a measurable NIVA Score reduction with volume removal. In an exploratory analysis, the discharge NIVA Score yielded an area under the receiver-operating-characteristic curve of 0.85 (95% confidence interval 0.575-1.00; P = 0.04) for thirty-day readmission. Together, the significant, directionally concordant NIVA Score reductions across independent clinical and preclinical cohorts demonstrate that the device tracks acute decongestion and support its use for serial, non-invasive congestion monitoring; an adequately powered prospective study is the planned next step.
Source: Early Feasibility of NIVA Score Decongestion Responsiveness: A Pilot Clinical and Preclinical Study