Combining physiological and biomedical data into a novel computer-aided risk score to support near real-time clinical decision-making Insight 2014

Currently, only intensive care clinicians routinely use statistically calculated scores to predict a patient’s risk of death. The University of Bradford project will create a system for giving all hospital clinicians near real-time access to risk of death scores, to support more effective clinical decision making. The team’s aim is to support a move away from measuring past harm to focusing on reducing harm.

The computer-aided risk score (CARS) combines biomedical data from routine blood tests with physiological data from National Early Warning Scores (NEWS). Clinicians will receive a patient’s CARS within 24 hours of an emergency admission. Previous work by the team has demonstrated that it is feasible to produce a risk score and this work will now introduce the score into real clinical settings.

The researchers will use historical NEWS and blood test data to develop a statistical model for each hospital, which they will then validate through a case note review process. They will use plan, do, study, act (PDSA) cycles and co-production methods to determine the most effective ways of incorporating risk scores into care pathways and gathering patient views on the new approach.

Expected outcomes

The project will evaluate the impact of introducing CARS on the quality of care, including mortality rates, cardiac arrests and unplanned admissions to ITU. The project is due to be completed in early 2017.

Contact details

For more information, email the Research team at the Health Foundation or Mohammed A Mohammed, Professor of Healthcare Quality and Effectiveness at the University of Bradford.