Inter-observer agreement in assessing comatose children.

Inter-observer agreement was evaluated for twelve items used in the neurological assessment of comatose children. Data were obtained prospectively on fifteen patients examined independently by two observers in a double-blind fashion. Observer variability was measured by using the Disagreement Rate and Kappa statistic. The Disagreement Rate ranged from 0.01 to 0.12 for all items. Values for Kappa statistic were generally in accordance with those for Disagreement Rate. The data suggest fair to almost perfect inter-observer agreement for the items used to assess comatose children in this study.