important for the Department of Defense and the Department of Veterans Affairs to work with the research community and provide acute clinical data on human blast-induced neurotrauma to enable refinement of the animal models.
The studies of TBI evaluated by the committee had numerous limitations. A primary limitation results from the nature of the control or comparison group assembled by the investigator. In an attempt to improve the quality of future TBI studies, the committee has described what it considers to be appropriate control groups.
Evaluating whether TBI in service members is associated with particular outcomes requires comparison groups of service members who have experienced injuries other than TBI and service members who have been deployed but not injured. Comparing outcomes of TBI with outcomes in those reference groups is the only means of identifying which outcomes are due solely to TBI and not to deployment or to injury in general.
The committee recommends that the Department of Veterans Affairs include, in the development of the Traumatic Brain Injury Veterans Health Registry (hereafter referred to as “the registry”), other service members who could provide a valid comparison for the analysis of outcomes. Comparison groups should be made up of injured persons without traumatic brain injury or blast exposure, uninjured deployed veterans, and uninjured nondeployed but previously active-duty veterans. Those groups could be compared with persons who have received a diagnosis of traumatic brain injury and with those who have possible or probable traumatic brain injury. The three comparison groups should have samples large enough to provide reference rates of outcomes of interest. Furthermore, the registry needs to be representative of the traumatic brain injury population to be able to determine associations between such injury and various outcomes. There should be no exclusions on the basis of sex, race, geographic region, or rank.
Access to medical records is essential to ensure the validity of a recommended research design. Neurologic status, computed tomographic or magnetic resonance imaging, electroencephalography, associated nonbrain injuries, and durations of impaired consciousness and PTA are important for the accurate classification of service members into appropriate groups.
For the registry to have the greatest benefit, predeployment information on all groups mentioned above should be made available to the injury-research community. Complete medical information on outcomes of each person (stripped of personal identifiers) in the registry should be available whether or not care is sought at or covered by the VA system.
In considering the question of long-term outcomes of TBI, questions arise that are very seldom addressable in current studies: What was the predeployment cognitive ability of the person? How did the TBI affect the baseline functioning? The answers to those questions are important in isolating and understanding the effects of TBI itself on long-term outcome. Most