Comparing the Validity of CAM-ICU and CAM-ICU-7 for Delirium Assessment based on Simulation

This is the authors' abstract. We don't add key points for this paper.

Purpose : This study aimed to compare the validity and user evaluations of the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) and its 7-item version (CAM-ICU-7) for delirium assessment. Methods : Thirty-six intensive care unit (ICU) nurses from four university hospitals participated in this simulation-based methodological study. Four standardized patients with normal cognition, subsyndromal delirium, hypoactive delirium, and hyperactive delirium were recruited. The participants assessed all four standardized patients using the CAM-ICU and CAM-ICU-7. We analyzed the validity, assessment time, and user evaluations of both tools. DeLong’s test and the paired t-test were used for comparisons. Results : The CAM-ICU showed sensitivity, specificity, and accuracy of 80.6%, 94.4%, and 84.0%, respectively, with an Area Under Curve (AUC) of 87.5%. CAM-ICU-7 demonstrated a sensitivity, specificity, and accuracy of 90.7%, 97.2%, and 92.4%, respectively, with an AUC of 94.0%. The CAM-ICU-7 showed a better AUC than the CAM-ICU for detecting subsyndromal delirium (p =.029). However, the CAM-ICU-7 required more time for assessment across all patient types (p＜ .001). ICU nurses reported that the CAM-ICU was faster to administer, whereas the CAM-ICU-7 was easier to use, better at detecting delirium, more effective at reflecting changes in delirium, and had a higher intention to use in clinical practice. Conclusion : The CAM-ICU-7 demonstrated superior validity compared to the CAM-ICU. Despite requiring more time, ICU nurses evaluated the CAM-ICU-7 as more useful. These findings support using the CAM-ICU-7 for delirium assessment in ICU settings.