Search:

Mission & Aims

The global aim is to optimize clinical trial methodology in the field of TBI to maximize the chance of demonstrating benefit of effective new therapies.

The IMPACT project will critically examine the methodological challenges posed by TBI trials, and investigate the application of conventional and innovative methods for design and analysis of trials in TBI. Data sets from completed randomized controlled trials (RCTs) and observational studies will be used as "culture media" in which to develop and test these methods. The insight obtained from these investigations will lead to informed recommendations for future clinical trials in TBI, and are expected to also be of relevance to RCT's in other fields.

IMPACT was initially funded from 2003-2006 (IMPACT I). IMPACT I focused on methodology to deal with the heterogeneity of the population and included extensive prognostic analyses.

IMPACT I established the IMPACT database and development of prognostic models necessary for relating final outcome to initial prognostic risk. We found that a relative trial size reduction of up to 50% can be achieved with covariate adjustment and by applying innovative statistical approaches, which exploit the ordinal nature of the Glasgow Outcome Scale (GOS). These include proportional odds analysis and applying the concept of the sliding dichotomy, in which the split for dichotomizing the GOS is differentiated according to baseline prognostic risk (Murray et al 2005).

Continuation funding (IMPACT II) was obtained for the period 2007-2011. In IMPACT II, we expanded the IMPACT database, including data from a mega trial and from more recent studies which contain the Extended GOS (GOSE), an endpoint of presumed increased sensitivity.

Impact II focuses on:

  • center effects and variations in patient management (specific aim 2)
  • sensitivity of outcome measures (specific aim 3) as related to statistical power
  • the choice (specific aim 4) between a mega trial (with large numbers of patients, substantial heterogeneity, and simple outcome measures) and a conventional trial (with fewer patients, less heterogeneity among centers, and more complex outcome measures)