Neal G. Ravindra, PHD,a David P. Kao, MDb
Pulmonary congestion is the principal admission reason in patients with acute decompensated heart failure (ADHF). Tracking fluid retention over the course of HF may prevent ADHF hospitalization and improve the quality of inpatient care. However, objectively tracking congestion cheaply, frequently, and accurately remains challenging.
Machine learning (ML) has accelerated “nonclinical” adoption of automated and functional speech recognition systems broadly, but it remains to be seen whether these systems can be used to track physiologic variables like pulmonary congestion for clinical use. A number of studies have taken advantage of the smartphone, a pervasive and noninvasive device for patient data collection, and studies have also shown a relationship between fluid retention and vocal cord vibration. Amir et al explore the possibility of monitoring pulmonary congestion through voice analysis using a smartphone app. In this commentary, we highlight implications this study may have on HF management, discuss potential alternative approaches, and highlight some cautionary notes regarding equity in use of such technology with respect to vulnerable populations.