[Interobserver reliability of the Glasgow coma scale in critically ill patients with neurological and/or neurosurgical disease]

Abstract

Introduction: The Glasgow coma scale (GCS) is a common tool used for neurological assessment of critically ill patients. Despite its widespread use, the GCS has some limitations, as sometimes different observers may value differently the same response.

Objective: To evaluate the interobserver agreement, among intensive care nurses with a minimum of 3 years experience, both in the overall estimate of GCS and for each of its components.

Methods: Prospective observational study including 110 neurological and/or neurosurgical patients conducted in a critical care unit of 18 beds, from October 2010 until December 2012. Registered variables: Demographic characteristics, reason for admission, overall GCS and its components. The neurological evaluation was conducted by a minimum of 3 nurses. One of them applied an algorithm and consensual assessment technique and all, independently, valued response to stimuli. Interobserver agreement was measured using the intraclass correlation coefficient (ICC) for a confidence interval (CI) of 95%. The study was approved by the Ethics Committee for Clinical Trails.

Results: The intraclass correlation coefficient (confident interval) for scale was: Overall GCS: 0.989 (0.985-0.992); ocular response: 0.981 (0.974-0.986); verbal response: 0.971 (0.960-0.979); motor response: 0.987 (0.982-0.991).

Conclusion: In our cohort of patients we observed a high level of consistency in the application of both the GCS as in each of its components.

Keywords: Concordance; Concordancia; Enfermería; Escala del coma de Glasgow; Glasgow coma scale; Neurological patients; Nursing; Paciente neurológico.