The ability to predict which COVID-19 patients will need intensive treatment could help physicians make better therapeutic decisions and help alleviate strain on hospitals during a surge in infections.

In a paper published online on January 18 in PLOS Computational Biology, Albert Einstein College of Medicine and Montefiore Health System researchers have developed a computational model that uses immunological and standard clinical biomarkers to predict patient outcomes. Clinicians have traditionally relied on demographic and clinical data obtained at the time of hospital admission to predict the mortality risk for COVID-19 patients. But the team found that data collected during hospitalization—including spike-protein antibody levels as well as white blood cell, neutrophil, and lymphocyte counts—more accurately predicted whether or not COVID-19 patients would live or die.

The team was led by Kartik Chandran, Ph.D., professor of microbiology & immunology and the Harold and Muriel Block Faculty Scholar in Virology at Einstein; Jonathan Lai, Ph.D., professor of biochemistry at Einstein; Libusha Kelly, Ph.D., associate professor of systems & computational biology and of microbiology & immunology at Einstein; Johanna Daily, M.D., M.S., professor of medicine and of microbiology & immunology at Einstein and an infectious disease physician at Montefiore; and Olivia Vergnolle, Ph.D., formerly at Einstein and currently at Memorial Sloan Kettering Cancer Center. The co-first authors from Einstein are Gorka Lasso, Ph.D., Saad Khan, and Stephanie Allen.