Algorithm developed to Predict Patient’s Chance of Survival in ICU through Individual’s Medical History

An individual patient’s risk of mortality in the ICU can be predicted more effectively using a newly developed algorithm by researchers in Denmark. The algorithm that they have developed outperforms current non- computational methods of estimating mortality, as has been demonstrated by their study published in the journal, “Digital Health”. An opportunity to improve patient outcomes and catching problems early on can be achieved through these algorithms, as they direct resources where they are needed the most.
Various metrics meant to estimate an individual’s chance of survival are already being used in the ICU by doctors and nurses, for a better understanding of the course of treatment to be implemented to deliver optimal care. However, in practice these metrics may or may not the accurate.
The data which was used in the latest research was taken from over 230,000 ICU patients, along with 23 years of medical history. The various factors in a patient’s medical history are weighed by neural networks by this new algorithm. Though the algorithm was not very predictive at first, there were significant improvements to accuracy when the researchers included measurements and tests made in the first 24 hours of admission to the ICU. The risk of a patient dying in the hospital, which could be any number of days after admission, the risk of the patient dying within 30 days of admission, and the risk of the patient dying within 90 days of admission are the three predictions that the algorithm makes.

The age and length of previous hospital visits were two of the most important variables their algorithm identified, according to an observation made by the team of researchers from Denmark. Additionally, the length of medical history was also crucial according to their findings because they found diagnoses from many years earlier had an important effect on predicting survival after admission to the ICU.

Leave a Reply

Your email address will not be published. Required fields are marked *

   Contact: +14079154157 | +14089009135 | Email:

Scroll to Top