Test-retest reliability and minimal detectable change of the computerized dynamic posturography PROPRIO for adults with chronic traumatic brain injury

Purpose: Balance deficits after brain injury, including reactive recovery from unexpected perturbations, can persist well after rehabilitation is concluded. While traditional clinical assessments are practical, the anticipatory nature of the tasks may mask perceptible balance control. Computerized dynamic posturography can directly quantify capacity to respond to unexpected, external perturbations. This study examined the reliability of the computerized dynamic posturography assessment with the device PROPRIO (R) 4000 in adults with traumatic brain injury and created the minimal detectable change for its standardized test. Methods: Ten adults (ages 21-55 years) with chronic (average 10 +/- 6 years post-injury) severe (loss of consciousness 2-75 days) brain injury performed three trials of the Propriotest (R) on two separate days. The average of three trials and the best scores were used separately for analysis. Test-retest reliability was verified using Intraclass Correlation Coefficients with 95% confidence interval and standard error of measurement in relation to the Intraclass Correlation Coefficients at 95%. The minimal detectable change was calculated at 95% confidence level (minimal detectable change(95)) and Bland-Altman plots were created to express agreement between measurement days. Results: The results exhibited excellent reliability for both average (Intraclass Correlation Coefficient of 0.969, standard error of measurement 50.9 points) and best (Intraclass Correlation Coefficient of 0.985, standard error of measurement 31.3 points) scores, with average and best minimal detectable change(95) of 141.0 and 86.7 points, respectively. Conclusions: Clinicians and rehabilitation researchers can use these findings to determine if a Propriotest (R) change score represents a true post-treatment effect with adults with chronic brain injury.