Department   School of Medicine(Tokyo Women's Medical University Hospital), School of Medicine
   Position   Endowed Professor
Article types Original article
Language English
Peer review Peer reviewed
Title Evaluation of thoracic impedance trends for implant-based remote monitoring in heart failure patients - Results from the (J-)HomeCARE-II Study.
Journal Formal name:Journal of electrocardiology
Abbreviation:J Electrocardiol
ISSN code:0022-0736/1532-8430
Domestic / ForeginForegin
Publisher Elsevier Inc.
Volume, Issue, Page 53,pp.100-108
Author and coauthor Maier Sebastian K G, Paule Stefan, Jung Werner, Koller Marcus, Ventura Rodolfo, Quesada Aurelio, Bordachar Pierre, García-Fernández F Javier, Schumacher Burghard, Lobitz Nicole, Takizawa Kaname, Ando Kenji, Adachi Kazumasa, Shoda Morio
Authorship Last author
Publication date 2019/03
Summary Aims: Remote monitoring by implantable devices substantially improves management of heart failure (HF) patients by providing diagnostic day-to-day data. The use of thoracic impedance (TI) as a surrogate measure of fluid accumulation is still strongly debated. The multicenter HomeCARE-II study evaluated clinically apparent HF events in the context of remote device diagnostics, focusing on the controversial role of TI.

Methods and results: We followed 497 patients (66.6 ± 10.1 years, 77% male, QRS 139.8 ± 36.0 ms, ejection fraction 26.8 ± 7.0%) implanted with a CRT-D (67%) or an ICD (33%) for 21.4 ± 8.1 months. An independent event committee confirmed 171 HF events of which 82 were used to develop a TI-based algorithm for the prediction of imminent cardiac decompensation. Highly inter-individual variations in patterns of TI trends were observed. The algorithm resulted in a sensitivity of 41.5% (50.0%) with 0.95 (1.34) false alerts per patient year, and a positive predictive value of 7.9% overall and 27.9% in the HF event group of patients. Averaged ratio statistics showed a significant pre-hospital decrease and a highly significant in-hospital increase in TI after intensified diuresis. Recurrent decompensations turned out to be preceded by a significantly stronger decrease of TI compared to first events with a higher chance for detection (63.6% sensitivity, p b 0.05).

Conclusions: Overall performance in predicting imminent decompensation by monitoring TI alone is limited due to its high inter-patient variability. TI stand-alone applications should be redirected towards a target population with more advanced symptoms where post-hospital observation aimed to maintain the patient's discharge status might be the most valuable approach.

Clinical trial registration: Identifier NCT00711360 (HomeCARE-II) and NCT01221649 (JHomeCARE-II).
DOI 10.1016/j.jelectrocard.2019.01.004
PMID 30739055