Automatic delirium prediction system in a Korean surgical intensive care unit

Background: In Korea, regular screening for delirium is not considered essential. In addition, delirium is often associated with vague concepts, making it harder to identify high-risk patients and impeding decision-making. Aims: To assess the impact of the Automatic PREdiction of DELirium in Intensive Care Units (APREDEL-ICU) system on nursing-sensitive and patient outcomes for surgical ICU patients and to evaluate nurse satisfaction with the system and its usability. Methods: A pre-post research design was adopted. Our study included 724 patients admitted before the implementation of the APREDEL-ICU (January to December 2010) and 1111 patients admitted after the system was installed (May 2011 to April 2012). The APREDEL-ICU uses a pop-up window message to inform the nursing staff of patients at risk for delirium, allowing evidence-based nursing interventions to be applied to the identified patients. A total of 42 nurses were surveyed to determine the system's usability and their level of satisfaction with it. Results: After the implementation of APREDEL-ICU, high-risk patients, determined using a prediction algorithm, showed a slight decrease in the incidence of delirium, but the changes were not significant. However, significant decreases in the number and duration of analgesic/narcotic therapies were observed after the implementation of the system. Nurse self-evaluation results showed an improvement in all categories of knowledge regarding delirium care. Conclusion: The use of a prediction and alerting system for ICU patients at high risk of delirium showed a potential increase in the quality of delirium care, including early detection and proper intervention.