The project ’CDSS’ predicts risk of emergency hospitalization

The project ’CDSS’ predicts risk of emergency hospitalization

Intro
Clinical Decision Support System (CDSS) is a research project that predicts whether citizens are in danger of being hospitalized within the next 100 days. Predictions are based on data from citizens from municipalities, hospitals and GPs in Horsens, Skanderborg, Hedensted and Odder. The data used in CDSS is from the database Cross-tracks.

Challenge
An emergency hospitalization in connection with, e.g., a heart attack frequently occurs when the patient has already been in contact with the public healthcare sector – for instance, after having consulted a physician or home care. Data has already been recorded then in the form of events, symptoms and other factors that may indicate that a heart attack is imminent. The problem is that this data is simply recorded; it is not used to predict if there is a heart attack looming, even though you potentially could predict it.

Solution
An algorithm that, based on data from municipalities, hospitals and general practitioners, can calculate the risk of acute hospitalization so that you can intervene before it goes wrong. In the project CDSS, many stakeholders are involved, and it is Enversion who has built the database from which it is founded. It is also Enversion who has developed the algorithm that calculates the risk of an acute hospitalization with about 90 percent accuracy.

Results
CDSS is, as mentioned, a research project. It runs until 2027, and so far, it is register-based research with anonymous data. However, we will see great socio-economic gains and personal gains for the individual citizen if it becomes possible to use data actively for prevention and predictions rather than treatments when damage has been done.

To sum up
CDSS is a great example of the possibilities that exist for revolutionizing healthcare through artificial intelligence.

Curious?
Call Christian Bang at +45 2277 7044.

Skills

Posted on

September 4, 2018

Skills

Posted on

September 4, 2018