Vocal Biomarker Predicts Mortality and prior hospitalization among Heart Failure Patients

Home telemedicine holds the potential to reduce costs and improve outcome among patients with congestive heart failure (CHF). This study aimed to developed a vocal biomarker and evaluate its association with adverse outcome among CHF patients. METHODS: Study cohort included 8,630 patients who were registered to a call center of patients with chronic conditions including CHF. Acoustic features were extracted from 20 seconds of speech for each patient. A biomarker was developed based on a training cohort of non-CHF patients (N=6,646) using machine learning techniques. The biomarker was tested on a mutually exclusive CHF study cohort (N=1,984) and was evaluated as a continuous and dichotomized variable: high (upper quartile) and low. RESULTS: Mean age of the CHF study population was 75±11, 64% were men. During a median follow-up of 18 months (IQR 33-12), 676 (34%) patients died. Kaplan Meier survival analysis showed that the cumulative probability of death among high biomarker group was significantly higher (36%±17% vs. 23%±19%, p<.001). Multivariate Cox regression analysis of the vocal biomarker with adjustment to known predictors of poor survival, demonstrated that each standard deviation increase in the biomarker was associated with a significant 63% increased risk of death during follow-up (95% CI 1.39 - 1.92, p < .0001). The model also demonstrated an independent association of the biomarker with hospitalizations during follow-up (p=.007). CONCLUSIONS: This is the first study to document a relationship between a voice and adverse outcome among CHF patients. The results have important clinical implications for telemedicine and CHF patient care.