Abstract
HYPOTHESIS: The overall aim was to assess the effects of repetitive testing on performance on a next-generation Sensory Organization Test (NG-SOT). BACKGROUND: The psychometric properties of sensory organization assessment with a next-generation posturography device are not fully characterized. Thus, a better understanding of the reliability of the NG-SOT is required to support its use as an outcome measure. METHODS: Forty adults (20 vestibular-impaired) participated. Test-retest reliability was determined using the interclass correlation coefficient [ICC (3,1)] for the composite, somatosensory, vision, vestibular, and visual preference scores. Learning effects were assessed by analyzing the change in the composite score over time. RESULTS: Analyzing the NG-SOT scores across all participants produced the following interclass correlation coefficients [95% confidence interval (CI)]: composite = 0.95 (0.92, 0.97), somatosensory = 0.20 (-0.06, 0.44), vision = 0.93 (0.88, 0.96), vestibular = 0.91 (0.85, 0.94), and visual preference = 0.19 (-0.07, 0.43). The mean differences (95% CI; p value) for the composite score between administrations were: 2.7 (1.0, 4.4; 0.002) for visits 1 and 2; 1.4 (-0.3, 3.1; 0.099) for visits 2 and 3; 0.7 (-1.0, 2.4; 0.403) for visits 3 and 4; and 0.4 (-1.3, 2.1; 0.653) for visits 4 and 5. CONCLUSIONS: Test-retest reliability was excellent for the composite, vision, and vestibular scores, but it was poor for the somatosensory and visual preference scores. A learning effect for the composite score was identified, but performance was stable after two administrations. Changes in the composite score that are greater than 6.5 points exceed the measurement error for the test.