Siemens AI-based COVID-19 Severity Algorithm Becomes Available to Help Clinicians Predict Progression to Severe Disease

Through a year-long collaboration with a number of leading healthcare institutions across the globe, Siemens has developed the Atellica COVID-19 Severity Algorithm, a model designed to help predict the potential likelihood of progression to severe disease and life-threatening multi-organ dysfunction in COVID-19 patients. Leveraging de-identified COVID-19 patient data from more than 14,000 COVID-19 patients from multiple healthcare institutions worldwide, nine clinically significant lab parameters were identified and selected for inclusion in the algorithm. In addition to patient age, D-dimer, Lactate dehydrogenase (LDH), Lymphocyte %, Eosinophil %, Creatinine, C-reactive protein (CRP), Ferritin, PT-INR, and high-sensitivity Cardiac Troponin-I are used to help predict the likelihood of disease progression to severe disease endpoints.

By entering a potential patient’s lab values and age, the algorithm will generate a COVID-19 clinical severity score, including projected probability of progression to ventilator use, end-stage organ damage, and 30-day in-hospital mortality. The AI-based algorithm has been interfaced to the Atellica Data Manager software and is currently being evaluated as Investigational Use Only to help assess potential benefit to patient care. With integration into the existing physician order/sample processing/result reporting workflow, a later generation of the algorithm could provide clinical decision support capabilities to support standardized testing protocols for patients positive for COVID-19.

“We want healthcare providers to have access to a predictive tool that leverages our expertise in artificial intelligence and helps advance the understanding of how to improve patient care for those affected by COVID-19,” said Deepak Nath, PhD, President of Laboratory Diagnostics, Siemens Healthineers. "With much of the world still in the throes of the pandemic, facing strict resource allocations, the ability to quickly identify patients at risk of progressing to severe disease in real-time could potentially assist in expediting triage. Early introduction of the most appropriate state-of-the-art treatments has been demonstrated to improve survival in high-risk patients."