Non-invasive vocal biomarker is associated with pulmonary hypertension

Abstract

Abstract Background Heart failure (HF) is a debilitating disease and is associated with significant morbidity and mortality, and costs to health care systems. Monitoring the impact of therapy remotely holds the potential to reduce HF-related hospitalizations, improve quality of life, and optimize the use of the limited resources. Voice signal is an emerging non-invasive biomarker that has been associated with a number of disease states. We previously identified a significant relationship between specific vocal biomarkers and coronary artery disease, and recently extended these observations by showing that a pre-specified voice biomarker was associated with increased mortality and re-hospitalization in patients with HF. Purpose In the current study, we evaluated the association between a vocal biomarker derived from voice signal analysis and invasively measured indices of pulmonary hypertension (PH). We hypothesized that the pre-specified voice biomarker might be associated with hemodynamic indices reflective of PH that are known to be linked to HF-severity, and that predict outcomes such as HF-related hospitalization and death. Methods The study population included patients referred for an invasive cardiac hemodynamic study between January 2017 and December 2018. Subjects had their voice signal recorded to their smartphone on three separate occasions prior to the cardiac study. A pre-established numeric vocal biomarker was derived from each recording, and the mean vocal biomarker calculated for each patient. Patients were a priori divided into two groups: those with high pulmonary arterial pressure (PAP) defined as ≥35 mmHg consistent with moderate or greater PH, versus those with a lower PAP. Results Eighty three patients, mean age 61.6±15.1 years, 37 (44.6%) male, were included in the study. The intraclass correlation coefficient for the vocal biomarker in all patients was 0.83 implying very good agreement between values. Patients with a high mean PAP (≥35 mmHg) had significantly higher values of the voice biomarker compared to those with a lower mean PAP (0.74±0.85 vs. 0.43±0.86, p=0.008). Patients with a high pulmonary vascular resistance (PVR) defined as a PVR ≥1.7 Wood Units had significantly higher values of the voice biomarker compared to those with a lower PVR (0.62±0.83 vs. 0.33±0.90, p=0.026). Multivariable logistic regression showed that an increase in the voice biomarker by 1 unit was significantly associated with a high PAP, odds ratio (OR) 2.31, 95% CI 1.05–5.07, p=0.038, and with borderline significance with a high PVR, OR 2.14, 95% CI 0.94–4.87, p=0.07. Conclusion The current study shows a relationship between a noninvasive vocal biomarker derived from voice signal analysis and invasively derived hemodynamic indices related to PH obtained during cardiac catheterization. These results may have important and practical clinical implications for telemedicine and remote monitoring of patients with HF and PH. Funding Acknowledgement Type of funding source: Private grant(s) and/or Sponsorship. Main funding source(s): Mayo Foundation; Beyond Verbal Communications