Abstract 19477: Transthoracic Bioimpedance Monitoring Predicts Readmission After Hospitalization for Heart Failure: Preliminary Results From the SENTINEL-HF Study

• Published in: Circulation (New York, N.Y.) Vol. 132; no. Suppl_3 Suppl 3; p. A19477
• Main Authors: Darling, Chad E, Dovancescu, Silviu, Riistama, Jarno, Saczynski, Jane, Kini, Nisha, Sert Kuniyoshi, Fatima, Rock, Joseph, Meyer, Theo E, Goldberg, Robert J, McManus, David D
• Format: Journal Article
• Language: English
• Published: by the American College of Cardiology Foundation and the American Heart Association, Inc 10.11.2015
• ISSN: 0009-7322; 1524-4539
• DOI: 10.1161/circ.132.suppl_3.19477

IntroductionPatients and health systems are focused on reducing readmissions for patients with acute decompensated heart failure (ADHF). Readmission after hospitalization is often secondary to HF decompensation, but it remains challenging to identify patients at-risk. Bioimpedance is a validated marker of thoracic fluid accumulation. We examined whether changes in bioimpedance, measured using a Fluid Accumulation Vest (FAV), predicted subsequent HF decompensation in patients discharged after ADHF.MethodsParticipants included 83 patients hospitalized for ADHF. Subjects were trained on the use of a FAV-smartphone dyad to obtain and transmit a 5-minute bioimpedance measurement once daily for 45-days after discharge.(see Figure) The outcome of interest, HF-related readmission was assessed using participant report and medical records. Sensitivity, specificity, negative and positive predictive values were calculated to describe the efficacy of the bioimpedance alert algorithm as a predictor of HF readmission.ResultsSubject characteristicsmean age 68 ± 11 years, 36% female, 92% white, mean ejection fraction of 44 ± 19%. 49 participants completed the 45-day follow-up and had sufficient, daily FAV data for analysis. Our main outcome of HF-related rehospitalization occurred in 8% of patients during follow-up. The decompensation detection algorithm demonstrated a sensitivity of 75%, specificity of 47%, positive/negative predictive values of 11% and 96%, respectively.ConclusionsThe preliminary results of this ongoing study suggest that HF readmissions may be predicted with modest sensitivity by our current decompensation detection algorithm. Further refinement of our transthoracic bioimpedance system may offer possibilities for reducing HF readmissions by enabling identification and treatment of outpatients at risk for readmission.