System to Detect Severe Respiratory Illness by Vocalization Analysis of Cough

DESCRIPTION (provided by applicant): Cough is a prominent symptom of severe cough-transmissible respiratory illness (SCTRI), e.g. influenza, tuberculosis (TB), and pertussis. The acoustic characteristics of cough have not been quantified previously as the basis for a disease-screening tool. We propose to use existing acoustic vocalization analysis software to characterize the acoustic parameters (APs) of cough. In the long term we intend to build a screening tool that will distinguish individuals with cough signatures due to potential SCTRI from those with cough signatures related to less serious conditions. We expect to demonstrate that a measurement tool based on APs is able to screen coughs for SCTRI well and quickly. Thus, cough analysis could constitute a new, non-invasive screening paradigm for control of epidemic respiratory illness. We have chosen to conduct this study in the healthcare setting, because hospitals are a high-leverage point to intervene in the spread of SCTRI. Close contact among healthcare workers (HCWs) and patients can lead to rapid spread within and between the two groups, widespread illness, severe staffing shortages and even deaths. To simplify the experimental design, we will enroll only HCWs as study participants. In Aim I, we will collect voluntary cough data from HCWs, who will be asked to cough into a recording microphone and to answer a brief list of questions about their health status. It is anticipated that in a large sample (500-1,000) some HCWs will have background coughs i.e., coughs associated with common, less- severe etiologies, such as allergies, asthma, smoker's cough, environmental irritants and mild upper respiratory infections. Recordings of healthy voluntary coughs and background coughs will be analyzed separately and compared. In Aim II, we will record and analyze the coughs of HCWs with SCTRI. In Phase II, we will undertake an experimental study to distinguish between coughs due to SCTRI and coughs due to less serious respiratory conditions on the basis of cough APs. Our short-term goal is the identification of the APs that characterize human coughs. In the long term, we will commercialize an electronic screening system that will enable hospitals and public health officials to distinguish those individuals with cough signatures due to potential SCTRI and those with cough signatures related to less serious conditions. We will conduct the following specific aims: I. Identify and quantify the APs that can be used to characterize coughs. II. Determine the APs that can be used to characterize SCTRI. Two or three sentence narrative for the lay reader: We propose to use vocalization analysis software to characterize coughs. In the long term we intend to build a screening tool that will distinguish individuals whose coughs are related to serious illnesses such as tuberculosis, pertussis and influenza from those with coughs related to less serious conditions such as the common cold. PUBLIC HEALTH RELEVANCE: We propose to use vocalization analysis software to characterize coughs. In the long term we intend to build a screening tool that will distinguish individuals whose coughs are related to serious illnesses that can cause epidemics such as tuberculosis and influenza from those with coughs related to less serious conditions such as the common cold.