Rehabilitation for Traumatic Brain Injury

To the Editor: The article by Dr Salazar and colleagues¹ on cognitive rehabilitation for traumatic brain injury (TBI) leads to inaccurate conclusions because they fail to place their findings and methods in the appropriate scientific context.

First, Salazar et al state that they compared an in-hospital cognitive rehabilitation program modeled after my own milieu-oriented approach with a limited home-rehabilitation program in a prospective randomized controlled design. I have repeatedly emphasized, however, that my neuropsychological rehabilitation program was intended for postacute patients with TBI who fail to return to work after undergoing traditional rehabilitation procedures.² Typically, patients were enrolled in this program 1 to 2 years after their injury. In contrast, the patients in the study by Salazar et al were enrolled a mean of 38 days beyond their TBI—hardly a group of patients in the postacute stage of injury.

Second, in conducting milieu-oriented neuropsychological rehabilitation programs for patients with moderate-to-severe TBI, it is important that staff members are able to develop a good working alliance. These patients have serious cognitive and personality disturbances. Without a good working alliance with staff members, patients often make poor vocational and interpersonal choices. With a working alliance, they tend to return to a productive lifestyle.³ Salazar et al do not mention this important variable.

Third, Salazar et al report that their rates of patients returning to work after the hospital and home programs were 90% and 94%, respectively. Yet, they state that their patients had moderate-to-severe TBIs. Other statistics indicate that, at best, only about a third of patients with severe TBIs are able to return to work 2 to 3 years after injury.^{4 - 5} Given these known statistics, the authors do not adequately account for the high rate of return to work, which is unprecedented in the literature. Their patients may have been much less severely impaired than most patients with moderate-to-severe TBIs, or perhaps they were returned to work inappropriately.

Fourth, and perhaps most telling, is the authors' observation that aggressive behavior increased in both groups of patients at their 1-year follow-up. An increase in such aggressive behavior is often a sign that patients with TBI have been enrolled in milieu-oriented neuropsychological rehabilitation programs too soon after their injury.³ The authors fail to discuss this important determinant of rehabilitation outcomes adequately. Yet, it likely underlies their own findings, which reflect a potentially useful program applied too soon. In contrast, patients' ability to control their affect improves after appropriate and effectively timed treatment.^{2 ,4 ,6} Unfortunately, the study by Salazar et al will likely be used by managed care interests as a tool to deny patients with TBI access to certain forms of cognitive rehabilitation.