An Initial Psychometric Analysis of the Brain Gauge Tactile-Based Test Battery and Its Potential for Clinical Use Assessing Patients With Acute Mild Traumatic Brain Injury

Objective A new brief computerized test battery that uses tactile stimulation, Brain Gauge (BG), has been proposed as a cognitive assessment aid and its developers have reported an almost perfect ability to distinguish acute mild traumatic brain injury (mTBI) patients from healthy controls. This investigation attempted to replicate those results and serve as an initial psychometrically and clinically focused analysis of BG. Methods BG scores from 73 military service members (SM) assessed within 7 days after having a clinically diagnosed mTBI were compared to 100 healthy SMs. Mean scores were compared, score distributions were examined, and univariate and multivariate base rate analyses of low scores were performed. Results SMs with mTBI had statistically significantly worse performance on both BG Reaction Time (RT) tests and the Sequential Amplitude Discrimination test as reflected by higher mean RT and RT variability and higher minimum detectable amplitude difference. SMs with mTBI also had a significantly lower whole-battery composite (i.e., Cortical Metric Symptom Score). Larger proportions of SMs with mTBI had lower overall performance than controls. However, at most only 26.9% of those with mTBI performed at potentially clinically meaningful cutoffs that were defined as various numbers of low scores that were prevalent in no more than 10% of the control group, which is equivalent to specificity >= 90% and sensitivity <= 26.9% for mTBI. Conclusion Our analysis did not replicate the high level of classification accuracy reported by BG's developers. Pending further psychometric development, BG may have limited clinical utility for assessing mTBI patients.

Automated summary

This study aimed to replicate the results of a new computerized test battery, Brain Gauge (BG), which uses tactile stimulation, for cognitive assessment of patients with mild traumatic brain injury (mTBI). The analysis did not replicate the high level of classification accuracy reported by BG's developers, suggesting limited clinical utility for assessing mTBI patients.