Abstract: Objective: To explore the application value of the combined delirium prediction model in ICU patients. Methods: The E-PRE-DELIRIC model and PRE-DELIRIC model were implanted into the critical information system to integrate into a combined delirium prediction model. The performance of the combined delirium prediction model was evaluated using the statistical indicators of the evaluation model. Results: The area under the ROC curve of the combined delirium prediction model was 0.815, the area under the ROC curve of the PRE-DELIRIC model was 0.800, and the area under the ROC curve of the E-PRE-DELIRIC model was 0.777. The Z test showed that the validity of the combined delirium prediction model was statistically different from that of the E-PRE-DELIRIC model and the PRE-DELIRIC model (P<0.05). Conclusion: The combined delirium prediction model has a better prediction value, which can effectively screen out different risk groups of delirium, and provide an important reference for the delaminating management of delirium.

Key words: delirium; prediction model; Intensive Care Unit; E-PRE-DELIRIC model; PRE-DELIRIC model