[Objective] The goal of this study was to investigate the reliability and validity of the score used to estimate the ability to obey commands (the Obedience Score : OS). [Subjects and methods] The study subjects comprised 138 (35 male and 103 female) patients hospitalized in Sakamoto Hospital Annex. We estimated the OS by scoring the degree of the response to each command of 1) phonation, 2) throat clearing, 3) tongue movement, 4) dry swallowing, 5) eye opening and closing and 6) shaking hands. The responses were scored according to a 3-point scale and then were added together to obtain the OS. Chronbach's α co-efficient, Kappa indices among 3 estimators, and test-retest reliability were investigated as indicators of reliability. In regard to the validity, the association between OS and the subjective estimation by 3 nurses working in the wards was evaluated and the cut-off level to predict the possibility of command obedience was calculated. [Results] There was a significant association (r=0.86,
p < 0.0001) between the sum of the scores for the upper aero-digestive tract motion and the sum of the scores for the other motions;furthermore, a high Cronbach's α co-efficient (0.94) was demonstrated. Kappa indices among the 3 estimators of the scores for all of the commands except dry swallowing were more than 0.8. Kappa indices between the test-retest were lower than those among the 3 estimators. An apparent difference was observed in the OS between the cases that were subjectively estimated as “possible to obey commands” and “impossible to obey commands.” Lastly, the best balance of sensitivity and specificity was demonstrated when the cut-off level was set at 1 to predict the possibility of command obedience. [Conclusion] These data demonstrate the sufficient reliability and validity of the OS. Furthermore, a single individual may estimate the OS but the test time should be selected carefully.
View full abstract