Response:

We appreciate Drs. Giacino's and Whyte's letter and their interest in our study. Many of the issues they raise were addressed within the article, including the limitations of the study. The study was not designed to be a definitive phase III trial, as implied by Drs. Giacino and Whyte. Rather it was performed as a phase II trial. These trials evaluate safety, dosage, and obtain some data on efficacy. In this trial, we substituted dose timing for the more usual dose escalation, as this was the major concern of the original review panel at the National Institute on Disability and Rehabilitation Research (NIDRR). The findings were a surprise to the authors. Patients clearly improved their rate of recovery while on the medication for several weeks, albeit only a trend within some parameters. The study does demonstrate that there may be effects up to 12 weeks after injury.

The article used the most commonly accepted outcome measures for TBI trials. 1 As most of the accepted outcome measures for TBI rely on ordinal scales, there is no way to address the statistical limitations of these outcome measures. We chose to use nonparametric analyses to be more statistically rigorous.

It is true that most of the recovery in all drug trials will be spontaneous. In the early stages of TBI, one is attempting to affect the rate of recovery. However, we disagree with regard to the "limited improvement expected in group 2 over the second 6 weeks." Clearly after the first 6 weeks on all of the outcome measures, there was room to improve before there would be a "ceiling" effect.

We agree that the "rate of change" on the DRS only demonstrated a trend. 2 We pointed this out within the article as well as the difference in the starting points between groups 1 and 2.2 The DRS has been criticized for insensitivity in many TBI trials, but it has been recommended by the National Institutes of Health and the US Food and Drug Administration for drug trials 1 and has been used in numerous TBI trials, including by the authors of the letter. 3 However, we also used other established outcome measures that had some obviously provocative findings.

We disagree with regard to the issue that "minor differences in prognosis" in acute TBI patients can be completely controlled for. In fact, this is the most significant issue with these trials. TBI patients are a heterogeneous group of patients. Furthermore, we do not clearly know all the variables that may affect recovery in this patient population. A recent review pointed out that definitive studies will be difficult and will require large numbers of TBI patients. 5 Because early intervention is likely to have the most profound effect, the design of the trials will, of necessity, look at the rate of recovery as well as a final endpoint, 4 particularly for phase II trials. 5

It is hoped that trial design evaluating recovery from TBI will improve with studies conducted in the National Center for Medical Rehabilitation Research (NCMRR) recently designated Cooperative Multicenter Traumatic Brain Injury Clinical Trials Network, of which we are a part. In the final paragraph of the article, we noted "a larger multicenter double-blind trial of amantadine is now indicated." 2 To that end, that particular study was proposed in our application to the NCMRR for the Clinical Trials Network.

REFERENCES