A newly-developed algorithm that has been trained with machine learning methods uses COVID-19 as an example to predict patients’ individual mortality risk.
An international team led by researchers at the Max Planck Institute for Intelligent System (Tübingen, Germany) has developed the algorithm and trained it to predict individual mortality risk for patients with COVID-19 by drawing on the data of thousands of patients around the world. The algorithm that aims to help medical professionals with mortality predictions for COVID-19 patients can also be trained to predict mortality risk for other diseases, and thus support physicians in decision-making processes.
The algorithm called Covews, which is short for COVID-19 Early Warning System, draws on medical data to reliably predict a patient’s risk of dying up to eight days in advance with a sensitivity of more than 95%. This means that in 95 out of 100 cases, the algorithm can detect whether a patient will die unless preventative measures are taken. At the same time, Covews works with a specificity of just under 70% for a prediction eight days in advance, meaning that in about 70 out of 100 cases in which death is predicted, the patients ultimately die. In other words, the algorithm sounds a false alarm in only 30 out of 100 cases and is significantly better for shorter time horizons. The algorithm can also be trained to make less sensitive, but more specific predictions.
To develop and especially to train Covews, the researchers used 33,000 anonymized data records from a cohort called Optum, which tracks patients in various hospitals in the US. They fed the algorithm information about how several routinely collected patient health parameters evolved over the course of the disease, and whether or not the person died from COVID-19. As a result, Covews learned to identify patterns in the data sets that indicated a high risk of mortality. The international team then tested how accurately Covews estimated this risk on about 14,000 other data sets from the Optum cohort. By testing Covews on data from the TriNetX global health network, which includes about 5,000 patients with positive COVID tests in the US, Australia, India, and Malaysia, the researchers showed that the algorithm not only predicts mortality risk with a high degree of certainty with data sets from this cohort, but also with data from other hospitals.
Although Covews makes reliable predictions, it will likely take quite some time before it is used in practice. This is partly because at many hospitals, the available data are not sufficiently structured, making the development of suitable software based on the algorithm particularly challenging. In any case, by making Covews freely available on the internet, the researchers are laying the groundwork for putting the algorithm into practice quickly. Not only could it be used for COVID-19 patients; with the right training, it could also predict mortality risk for other diseases.
“Doctors must thus always decide on treatment measures,” said Stefan Bauer of the Max Planck Institute for Intelligent Systems who led the international team of researchers. “However, our algorithm can provide insights that people can’t derive from the data, and that can help with medical decisions.”