Abstract
BACKGROUND: Nurse-led rounding checklists are a common strategy for facilitating evidence-based practice in the intensive care unit (ICU). To streamline checklist workflow, some ICUs have the nurse or another individual listen to the conversation and customize the checklist for each patient. Such customizations assume that individuals can reliably assess whether checklist items have been addressed. OBJECTIVE: To evaluate whether 1 critical care nurse can reliably assess checklist items on rounds. METHODS: Two nurses performed in-person observation of multidisciplinary ICU rounds. Using a standardized paper-based assessment tool, each nurse indicated whether 17 items related to the ABCDEF bundle were discussed during rounds. For each item, generalizability coefficients were used as a measure of reliability, with a single-rater value of 0.70 or greater considered sufficient to support its assessment by 1 nurse. RESULTS: The nurse observers assessed 118 patient discussions across 15 observation days. For 11 of 17 items (65%), the generalizability coefficient for a single rater met or exceeded the 0.70 threshold. The generalizability coefficients (95% CIs) of a single rater for key items were as follows: pain, 0.86 (0.74-0.97); delirium score, 0.74 (0.64-0.83); agitation score, 0.72 (0.33-1.00); spontaneous awakening trial, 0.67 (0.49-0.83); spontaneous breathing trial, 0.80 (0.70-0.89); mobility, 0.79 (0.69-0.87); and family (future/past) engagement, 0.82 (0.73-0.90). CONCLUSION: Using a paper-based assessment tool, a single trained critical care nurse can reliably assess the discussion of elements of the ABCDEF bundle during multidisciplinary rounds.