The study, which was published on Thursday by Chinese researchers in a journal called Nature Machine Intelligence, claims to have found a way to determine how sick a patient with COVID-19 might get more than 10 days before they actually do. It also boasts a 90 percent accuracy rate, according to the study.
“This study leverages a database of blood samples from 485 infected patients in the region of Wuhan, China, to identify crucial predictive biomarkers of disease mortality,” the study’s authors wrote.
“For this purpose, machine learning tools selected three biomarkers that predict the mortality of individual patients more than 10 days in advance with more than 90 percent accuracy: lactic dehydrogenase (LDH), lymphocyte and high-sensitivity C-reactive protein (hs-CRP),” researchers said.
High levels of LDH play a role in determining what cases might require immediate care and what patients need to be prioritized, according to the study. High levels of LDH are also associated with tissue breakdown and could help provide clues as to what patients can expect going forward.
This news comes less than a week after the Food and Drug Administration (FDA) authorized a new antigen test that can quickly detect coronavirus proteins from the swabs of infected patients.
The test, developed by Quidel Corp. in San Diego, Calif., examines samples from the patient’s nasal cavity and is the third type of test to be authorized by the FDA, according to The Associated Press.
The FDA said it expects to authorize similar antigen tests in the future. Quidel claimed its test can provide accurate, automated results in about 15 minutes.
A separate research study posted online last week claimed most patients who recover from COVID-19 will make antibodies despite age, gender or how badly they were infected.
As of Saturday afternoon, there were over 1.45 million confirmed cases of coronavirus in the U.S. and over 88,000 deaths, according to Johns Hopkins University.
The Associated Press contributed to this report