Auto-DelRAS — Automated Delirium Risk Assessment System

What this tool is and how to interpret its evidence

Auto-DelRAS is an electronic system designed to identify ICU patients at increased risk of developing delirium. It uses 11 predictors already available in the health record, covering demographic and neurological information, physiological measurements, clinical service, infection, laboratory data and aspects of treatment or care. The implemented system displayed high, moderate or low risk on a clinical screen so that nurses could consider preventive action. It does not perform a bedside delirium assessment, and its risk category must not be interpreted as a current diagnosis.

Moon and colleagues developed the model using 3,284 medical and surgical ICU patients in Seoul. A further 325 patients were used for external validation, and performance after clinical implementation was examined in 694 patients. At one year the authors reported sensitivity of 0.88, specificity of 0.72, positive predictive value of 0.53 and negative predictive value of 0.94. These values depend on delirium prevalence, predictor availability and the source hospitals' workflow.

Auto-DelRAS is notable because the investigators assessed sustained performance after integration rather than reporting only model development. Even so, it was built within two Korean university hospitals, and a local EHR implementation would require technical mapping, governance, validation and monitoring for drift. This page gives only the number and broad nature of predictors; the paper is the authoritative source for definitions. The system can prompt prevention and repeated screening, but it cannot rule delirium in or out.