Abstract
Introduction: COVID-19 mainly manifests as a respiratory disease, and cough is a major symptom. Age and certain comorbidities are recognized risk factors for severe disease and hospitalization. Mobile technology could help to more precisely predict the course of disease.
Aims and objectives: To detect cough frequencies in hospitalized patients with COVID-19 and non-COVID-19 pneumonia and correlate these data to a variety of clinical parameters.
Methods: Smartphone-enabled detection of coughs technically based on a convolutional neural network-based model was used in 33 patients with COVID-19 and 12 patients with non-COVID-19 pneumonia in a non-ICU setting. Clinical data were extracted from medical records and correlated to cough frequencies.
Results: The technology reliably detected coughing events in all COVID-19 and non-COVID-19 patients over extended periods of time. In contrast to non-COVID-19, significant positive correlations between hourly cough counts and blood ferritin levels, FiO2, and breathing rate were found in COVID-19 pneumonia (Figure 1), and hourly cough counts decreased significantly with hospitalization length.
Conclusions: Automated, smartphone-based quantification of cough is feasible in an in-patient setting. Cough counts correlated with surrogate markers of COVID-19 disease activity and decreased towards hospital discharge. Although a low sample size limits the generalizability of our study, results are encouraging and warrant further investigation of cough as a COVID-19 digital biomarker.
Footnotes
Cite this article as: European Respiratory Journal 2021; 58: Suppl. 65, PA3865.
This abstract was presented at the 2021 ERS International Congress, in session “Prediction of exacerbations in patients with COPD”.
This is an ERS International Congress abstract. No full-text version is available. Further material to accompany this abstract may be available at www.ers-education.org (ERS member access only).
- Copyright ©the authors 2021
