Abstract
Objective: The authors reviewed the existing literature on the Veterans Health Administration's (VHA's) traumatic brain injury (TBI) screening and evaluation program to provide a qualitative synthesis and critical review of results focusing on the psychometric properties of the screen.
Methods: All studies of the VHA's screening and evaluation process were reviewed, both those involving primary data collection and those relying upon VHA data. Diagnostic statistics were summarized and also recalculated on the basis of a positive screening rate of 20%, the observed rate within the VHA, and an estimated population prevalence of TBI of 15% within the Department of Veterans Affairs (VHA).
Results: The TBI screen within the VHA is administered to nearly every eligible patient. The majority of clinical presentations are deemed to be due to mental health and/or a combination of mental health and TBI factors. The screen has good internal consistency, variable test-retest reliability, and questionable validity, with poor agreement between the TBI screen and criterion standards overall. Studies based on nonrepresentative samples reported high sensitivity. Assuming the VHA's TBI screening rate of 20% in a hypothetical sample, sensitivity is poor (the screen misses 30%-60% of TBI cases). However, specificity remains adequate. Studies based on samples with high rates of TBI reported much higher positive predictive values (and slightly lower negative predictive values) than those observed when a hypothetical TBI prevalence of 15% was used.
Conclusion: Questions remain about the validity of the TBI screen. Future research should address the utility of screening for TBI.