Symptom clusters in COVID-19: A potential clinical prediction tool from the COVID Symptom Study app

Sudre, Carole H.; Lee, Karla A.; Lochlainn, Mary Ni; Varsavsky, Thomas; Murray, Benjamin; Graham, Mark S.; Menni, Cristina; Modat, Marc; Bowyer, Ruth C. E.; Nguyen, Long H.; Drew, David A.; Joshi, Amit D.; Ma, Wenjie; Guo, Chuan-Guo; Lo, Chun-Han; Ganesh, Sajaysurya; Buwe, Abubakar; Pujol, Joan Capdevila; du Cadet, Julien Lavigne; Visconti, Alessia; Freidin, Maxim B.; El-Sayed Moustafa, Julia S.; Falchi, Mario; Davies, Richard; Gomez, Maria F.; Fall, Tove; Cardoso, M. Jorge; Wolf, Jonathan; Franks, Paul W.; Chan, Andrew T.; Spector, Tim D.; Steves, Claire J.; Ourselin, Sébastien

As no one symptom can predict disease severity or the need for dedicated medical support in coronavirus disease 2019 (COVID-19), we asked whether documenting symptom time series over the first few days informs outcome. Unsupervised time series clustering over symptom presentation was performed on data collected from a training dataset of completed cases enlisted early from the COVID Symptom Study Smartphone application, yielding six distinct symptom presentations. Clustering was validated on an independent replication dataset between 1 and 28 May 2020. Using the first 5 days of symptom logging, the ROC-AUC (receiver operating characteristic – area under the curve) of need for respiratory support was 78.8%, substantially outperforming personal characteristics alone (ROC-AUC 69.5%). Such an approach could be used to monitor at-risk patients and predict medical resource requirements days before they are required.