Abstract
BACKGROUND: Delirium is a prevalent and serious ICU complication, particularly in elderly or ventilated patients. Accurate assessment is crucial but often inconsistent. Intensive care unit (ICU) nurses' use of the Intensive Care Delirium Screening Checklist (ICDSC) may be limited without structured training. AIM: To evaluate the delirium assessment performance of ICU nurses using ICDSC and assess the effectiveness of a multimodal educational intervention for performance enhancement. STUDY DESIGN: This pre- and post-intervention study was conducted in three medical ICUs in Northern Taiwan. The delirium assessment performance of ICU nurses using ICDSC was evaluated, followed by a three-month multimodal educational intervention aimed at improving assessment performance. Each nurse's ICDSC assessment was paired with an independent assessment by a trained expert nurse. To ensure representation that reflects the true performance level, accounting for variations in nurses' working shifts, weekday and weekend staff ratios and sampling fairness across the three MICU units, the pairings were made using a three-step randomization process, managed by an independent third party. A 3-month multimodal educational intervention, including didactic lectures, difficult scenario reviews and one-to-one bedside mentoring, was implemented afterward. Inter-rater agreement before and after the intervention was assessed using Cohen's kappa and Gwet's AC1 statistics. RESULTS: The baseline agreement between ICU nurses and the expert nurse was suboptimal (kappa = 0.63, 95% CI: 0.57-0.70). The multimodal educational intervention was well-received by the units, particularly among new nurses. Following the intervention, kappa significantly improved to 0.74 (95% CI: 0.69-0.80). Notable improvements were observed in key items of ICDSC, including the assessment of altered level of consciousness, inattention, disorientation, psychomotor agitation or retardation, and sleep-wake cycle disturbances. However, agreement remained poor for certain patient populations, especially those aged over 85 years and those subjected to physical restraint. CONCLUSIONS: A structured, multimodal educational intervention significantly improved the delirium assessment performance of ICU nurses using the ICDSC. One-to-one coaching and scenario-based learning were particularly effective in enhancing clinical assessment skills. However, additional strategies may be required to address persistent challenges in assessing very elderly patients and those subjected to physical restraint. RELEVANCE TO CLINICAL PRACTICE: A structured, multimodal educational intervention can substantially enhance the performance of ICU nurses in delirium screening using the ICDSC. Tailored training strategies may help bridge the knowledge-practice gap, leading to more reliable clinical assessments in critical care settings.