Abstract 11637: Using Dielectric Sensing in Making Discharge Decision for Patients With Heart Failure

Abstract

Introduction: Decompensated heart failure is a leading cause of hospitalization in the United States and is driven by high readmission rates. Despite numerous advancements in diagnosing and treating advanced heart failure, there remains a lack of validated measures to predict patient volume status and readmission rates. Remote dielectric sensing (ReDS) is a non-invasive, radiation-free measure of lung fluid content. Hypothesis: We aim to assess the utility of ReDS in guiding clinical management and as a quantitative metric to predict 30-day readmission rates. We predict normalization of lung fluid value before discharge will have a lower 30-day readmission rate than patients with no normal value before discharge. Methods: Lung fluid values were collected using ReDS vests for 1,224 patients with advanced heart failure. Our retrospective analysis compared patients with normalization of lung fluid values during admission to patients without normalization of these measures to determine a difference in readmission rates between groups. Secondary analyses were performed to assess the effects of other clinical markers, including ejection fraction and age, on lung fluid values and 30-day readmission rates for individual patients. Results: The primary comparison between patients with normal lung fluid values at discharge and patients without normalization of lung fluid values revealed a trend towards lower readmission for those with normalization of lung fluid values (16.6% vs. 29.2%, p=0.07). No differences were observed in patients who had non-normal repeat lung fluid values before discharge and those who did not receive a repeat measure (29.2% vs. 29.8%, p=0.88). Readmitted patients had a higher initial lung fluid value (45±6.6 versus 44 ± 5.9, p=0.02). No differences in lung fluid values at discharge or changes during hospitalization were observed. Conclusions: Remote dielectric sensing may predict lower readmission rates in patients whose values normalize during admission compared to patients with non-normalized values. Future prospective studies are necessary to determine the full utility of ReDS in the clinical setting.