Predictive models of long COVIDResearch in context
Blessy Antony,
Hannah Blau,
Elena Casiraghi,
Johanna J. Loomba,
Tiffany J. Callahan,
Bryan J. Laraway,
Kenneth J. Wilkins,
Corneliu C. Antonescu,
Giorgio Valentini,
Andrew E. Williams,
Peter N. Robinson,
Justin T. Reese,
T.M. Murali,
Christopher Chute
Affiliations
Blessy Antony
Department of Computer Science, Virginia Polytechnic Institute and State University (Virginia Tech), Blacksburg, VA, 24061, USA
Hannah Blau
The Jackson Laboratory for Genomic Medicine, Farmington, CT, 06032, USA
Elena Casiraghi
AnacletoLab, Computer Science Department, Dipartimento di Informatica, Università degli Studi di Milano, Milan, 20133, Italy; Division of Environmental Genomics and Systems Biology, Lawrence Berkeley National Laboratory, Berkeley, CA, 94720, USA; ELLIS - European Laboratory for Learning and Intelligent Systems, Milan Unit, Milan, 20133, Italy
Johanna J. Loomba
Integrated Translational Health Research Institute of Virginia, University of Virginia, Charlottesville, VA, 22904, USA
Tiffany J. Callahan
Department of Biomedical Informatics, Columbia University Irving Medical Center, New York, NY, 10032, USA
Bryan J. Laraway
Department of Biomedical Informatics, University of Colorado Anschutz Medical Campus, Aurora, CO, 80045, USA
Kenneth J. Wilkins
Biostatistics Program, Office of the Director, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD, 20814, USA
Corneliu C. Antonescu
Banner Health, University of Arizona, Phoenix, AZ, 85006, USA
Giorgio Valentini
AnacletoLab, Computer Science Department, Dipartimento di Informatica, Università degli Studi di Milano, Milan, 20133, Italy; ELLIS - European Laboratory for Learning and Intelligent Systems, Milan Unit, Milan, 20133, Italy
Andrew E. Williams
Institute for Clinical Research and Health Policy Studies, Tufts University School of Medicine, Boston, MA, 02111, USA
Peter N. Robinson
The Jackson Laboratory for Genomic Medicine, Farmington, CT, 06032, USA; Institute for Systems Genomics, University of Connecticut, Farmington, CT, 06269, USA
Justin T. Reese
Division of Environmental Genomics and Systems Biology, Lawrence Berkeley National Laboratory, Berkeley, CA, 94720, USA
T.M. Murali
Department of Computer Science, Virginia Polytechnic Institute and State University (Virginia Tech), Blacksburg, VA, 24061, USA; Corresponding author.
Summary: Background: The cause and symptoms of long COVID are poorly understood. It is challenging to predict whether a given COVID-19 patient will develop long COVID in the future. Methods: We used electronic health record (EHR) data from the National COVID Cohort Collaborative to predict the incidence of long COVID. We trained two machine learning (ML) models — logistic regression (LR) and random forest (RF). Features used to train predictors included symptoms and drugs ordered during acute infection, measures of COVID-19 treatment, pre-COVID comorbidities, and demographic information. We assigned the ‘long COVID’ label to patients diagnosed with the U09.9 ICD10-CM code. The cohorts included patients with (a) EHRs reported from data partners using U09.9 ICD10-CM code and (b) at least one EHR in each feature category. We analysed three cohorts: all patients (n = 2,190,579; diagnosed with long COVID = 17,036), inpatients (149,319; 3,295), and outpatients (2,041,260; 13,741). Findings: LR and RF models yielded median AUROC of 0.76 and 0.75, respectively. Ablation study revealed that drugs had the highest influence on the prediction task. The SHAP method identified age, gender, cough, fatigue, albuterol, obesity, diabetes, and chronic lung disease as explanatory features. Models trained on data from one N3C partner and tested on data from the other partners had average AUROC of 0.75. Interpretation: ML-based classification using EHR information from the acute infection period is effective in predicting long COVID. SHAP methods identified important features for prediction. Cross-site analysis demonstrated the generalizability of the proposed methodology. Funding: NCATS U24 TR002306, NCATS UL1 TR003015, Axle Informatics Subcontract: NCATS-P00438-B, NIH/NIDDK/OD, PSR2015-1720GVALE_01, G43C22001320007, and Director, Office of Science, Office of Basic Energy Sciences of the U.S. Department of Energy Contract No. DE-AC02-05CH11231.
