An algorithm to identify people most at risk of dying from COVID when being treated in a hospital has been developed by scientists.
The COVID-19 Disease Outcome Predictor (CODOP) will help to give early warning to doctors about who needs critical care.
The system uses 12 blood molecules which are normally collected in hospital when a patient is admitted – meaning the algorithm should be easy to adopt by health care systems across the world.
CODOP uses artificial intelligence and can identify which patients face a poor prognosis, according to the study published in eLife.
It found that the algorithm can predict the survival or death of hospitalised patients with high accuracy until nine days before either outcome occurs.
Along with helping doctors make decisions on those patients needing the most critical care and the route of treatment, it could also help countries who have access to fewer resources.
The study’s senior author and leader of the international project said new variants, waning immunity and the relaxation of safety measures “means we are likely to continue seeing surges of infections and hospitalisations”.
COVID-19: Kim Jong Un slams North Korea’s ‘immature’ response to outbreak as death toll grows
COVID-19: Scientists receive £15m boost to tackle deadly respiratory illnesses following COVID study
COVID-19: Dogs can be trained in weeks to detect infection
He added: “There is a need for clinically valuable and generalisable triage tools to assist the allocation of hospital resources for COVID-19, particularly in places where resources are scarce.
“But these tools need to be able to cope with the ever-changing scenario of a global pandemic and must be easy to implement.”
Read more:
Dogs can be trained to detect COVID
Data was used from routine blood samples taken from nearly 30,000 patients in more than 150 hospitals in Spain, the US, Bolivia, Honduras and Argentina between March 2020 and February 2022.
This allowed researchers to collect information from people with vaccinated, unvaccinated and natural immunity status and covered the emergence of different variants, including Omicron.
