Interobserver reliability of Glasgow Coma Scale scores for intensive care unit patients

Abstract

Background: intensive care units frequently use the Glasgow Coma Scale to objectively assess patients’ levels of consciousness. Interobserver reliability of Glasgow Coma Scale scores is critical in determining the degree of impairment. Objective: to evaluate interobserver reliability of intensive care unit patients’ Glasgow Coma Scale scores. Methods: this prospective observational study evaluated Glasgow Coma Scale scoring agreement among 21 intensive care unit nurses and 2 independent researchers who assessed 202 patients with neurosurgi-cal or neurological diseases. Each assessment was completed independently and within 1 minute. Partici-pants had no knowledge of the others’ assessments. Results: agreement between Glasgow Coma Scale component and sum scores recorded by the 2 researchers ranged from 89.5% to 95.9% (P = .001). Significant agreement among nurses and the 2 researchers was found for eye response (73.8%), motor response (75.0%), verbal response (68.1%), and sum scores (62.4%) (all P = .001). Significant agreement among nurses and the 2 researchers (55.2%) was also found for sum scores of patients with sum scores of 10 or less (P = .03). Conclusions: although the study showed near-perfect agreement between the 2 researchers’ Glasgow Coma Scale scores, agreement among nurses and the 2 researchers was moderate (not near perfect) for sub-component and sum scores. Accurate Glasgow Coma Scale evaluation requires that intensive care unit nurses have adequate knowledge and skills. Educational strategies such as simulations or orientation practice with a preceptor nurse can help develop such skills.