Predictive Validity and Inter-Rater Reliability of the Persian Version of Full Outline of Unresponsiveness Among Unconscious Patients with Traumatic Brain Injury in an Intensive Care Unit

Abstract

The Glasgow Coma Scale (GCS) has some limitations when evaluating the unconscious patient. This study aims to validate the Persian version of the FOUR (Full Outline of Unresponsiveness) score as a proposed substitute. Two nurses, two nursing students, and two physicians scored the prepared Persian version of the FOUR and GCS in 84 patients with acute brain injury. The inter-rater agreement for the FOUR and the GCS scores was evaluated by the weighted kappa (kappa (w)). The outcome prediction power of the scales was assessed by the area under the curve (AUC) in the ROC curve. The inter-rater agreement of the FOUR was excellent (kappa (w) = 0.923, 95 CI, 0.874-0.971) and comparable with the one of the GCS (kappa (w) = 0.938, 95 CI, 0.889-0.987). The area under the curve (AUC) for predicting in-hospital mortality (modified Rankin Scale: 6) was 0.835 for the FOUR (95 CI, 0.739-0.907) and 0.772 for the GCS (95 CI, 0.668-0.856) (P = 0.01). AUC for predicting poor outcome (modified Rankin Scale: 3-6) for the total FOUR score was 0.983 (95 CI, 0.928-0.999), which is comparable with 0.987 for the total GCS score (95 CI, 0.934-1.000). The researchers conclude that the Persian version of the FOUR score is a reliable and valid scale to assess unconscious patients with traumatic brain injury and can be substituted for the GCS.