Validity of the Veterans Health Administration's traumatic brain injury screen

Objective: To compare the results of Veterans Affairs' (VA's) initial traumatic brain injury (TBI) Clinical Reminder Screen with the more extensive second-level Comprehensive TBI Evaluation in a national sample.

Design: Criterion-standard.

Setting: Veterans Health Administration system of polytrauma care.

Participants: The data were from VA's centralized database Patient Care Services on the TBI Clinical Reminder Screen and Comprehensive TBI Evaluation results of veterans (N=48,175).

Interventions: None.

Main outcome measures: Sensitivity, specificity, positive predictive value, and negative predictive value of the TBI Clinical Reminder Screen were calculated by using the Comprehensive TBI Evaluation findings as the comparative standard for TBI confirmation.

Results: The TBI Clinical Reminder Screen has generally good sensitivity (.87-.90) but poor specificity (.13-.18). In addition, the TBI Clinical Reminder Screen, when compared with the Comprehensive TBI Evaluation by a clinician, has generally poor negative predictive power (.31-.49) in this sample. However, negative predictive power is good with an estimated Veterans Health Administration system-wide TBI prevalence rate of 15% (.89). Positive predictive power was acceptable (.77) in this sample. The screen performs comparably across patient demographic and symptom severity characteristics, as well as across level of polytrauma care. Systematic evaluations by clinicians primarily reveal mental health-perceived causes of ongoing symptoms.

Conclusions: In summary, VA's Clinical Reminder Screen, when evaluated against the follow-up Comprehensive TBI Evaluation, has good sensitivity but poor specificity.