Diagnostic accuracy of the CAM-ICU and ICDSC in detecting intensive care unit delirium: A bivariate meta-analysis

Background: Delirium is a critical and highly prevalent problem among critically ill patients. The Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) and the Intensive Care Delirium Screening Checklist (ICDSC) are the most recommended assessment tools for detecting intensive care unit (ICU) delirium. Objectives: To synthesize the current evidence and compared the diagnostic accuracy of the two tools in the detection of delirium in adults in ICUs. Design: Systematic review and meta-analysis. Data source: A comprehensive search of the following electronic databases was performed using PubMed, Embase, CINAHL and ProQuest Dissertations and Theses A&I. The date range searched was from database inception to April 26, 2019. Review methods: Two researchers independently identified articles, systematically abstracted data and evaluated the sensitivity and specificity of the CAM-ICU or the ICDSC against standard references. Bivariate diagnostic statistical analysis with a random-effects model was performed to summarize the pooled sensitivity and specificity of the two tools. Results: In total, 29 CAM-ICU and 12 ICDSC studies were identified. The pooled sensitivity was 0.84 and 0.83 and pooled specificity was 0.95 and 0.87 for the CAM-ICU and the ICDSC, respectively. The CAM-ICU had higher summary specificity than the ICDSC did (p = 0.04). The percentage of hypoactive delirium, ICU type, use of mechanical ventilation, number of participants, and female percentage moderated the accuracy of the tools. Most of the domains of patient selection, index test, reference standards, and flow and timing were rated as having a low or unclear risk of bias. Conclusions: Although both the CAM-ICU and the ICDSC are accurate assessment tools for screening delirium in critically ill patients, the CAM-ICU is superior in ruling out patients without ICU delirium and detecting delirium in patients in the medical ICU and those receiving mechanical ventilation. Further investigations are warranted to validate our findings. The study protocol is registered at PROSPERO (CRD42020133544).