PRE-DELIRIC Prediction Model Plus SMART Care to Reduce the Incidence of Delirium in ICU Patients

Delirium is a severe acute brain dysfunction characterised by sudden confusion, inattention and fluctuating level of consciousness, which mainly affects intubated intensive care patients. It increases the risk of self-extubation, prolongs ICU stay and increases mortality. The incidence of delirium in ICUs varies, with approximately 33.3% of patients affected, and rates of new-onset and pre-existing delirium range from 4% to 89%. Accurate diagnosis is challenging, with 60-80% of patients remaining undiagnosed. Early detection is critical for intervention and improved outcomes.

To address these issues, the PREdiction of DELIRium (PRE-DELIRIC) model incorporates 10 risk factors and predicts delirium within 24 hours of ICU admission, allowing risk stratification into low to very high risk categories. It recalibrates predictive values with a sensitivity of 91.3% and specificity of 64.4% using a cut-off score of 27%. However, its integration into delirium management is underexplored. Delirium risk stratification supports efficient resource allocation, cost control, workload reduction and ethical care, while promptly identifying high-risk patients.

In this study, Investigators evaluate the integration of the PRE-DELIRIC model into a comprehensive delirium management approach called PRE-DELIRIC-guided SMART/SmART care. SMART care includes improving familiarity, assessing pain and anxiety, reducing equipment discomfort and cognitive stimulation. Patients with PRE-DELIRIC scores >30% receive SMART care and multidisciplinary involvement, based on the American Delirium Society.