: Traumatic brain injury (TBI) constitutes a major health and socio-economic problem throughout the world, accounting for a third of trauma deaths and for a much larger proportion of life-long disability after trauma. Despite enthusiasm generated from positive pre-clinical investigations, and of phase II trials on new therapies for reducing secondary brain damage and improving outcome, the results of phase III trials have been extremely disappointing, none of them showing convincing evidence of efficacy in the overall population of TBI. Problems in clinical trial design and analysis, specific to the field of TBI, have contributed to this failure. Methodologic challenges are related to the heterogeneity of the population, outcome assessment, detection of treatment effects, the use of surrogate markers and interim analysis. Addressing these problems requires a multidisciplinary effort involving expertise from the fields of clinical research, biostatistics and epidemiology. This effort is realized in the proposed study, organized as an international collaboration, based on extensive analysis of data available from studies (n=3) and trials (n=8) in TBI, encompassing approximately 10,000 patients. The efficiency of mechanistic and/or prognostic targeting, as well as covariate adjustment for dealing with heterogeneity, will be investigated (spec.
aim 1), influence of outcome distribution and observer variation on statistical power, differentiated analysis of outcome according to prognostic risk (spec.
aim 2), and development/validation of prognostic models with conventional statistical approaches, as well as with approaches new to the field of TBI (spec.
aim 3). Additional design issues, considered relevant to TBI trials, focusing on recruitment, time-windows, and sequential design are addressed (spec.
aim 4). The insight obtained from these investigations will lead to informed recommendations for the design and analysis of future clinical trials. Results may also be relevant to trials in other fields of medicine, in which similar problems have been recognized, such as stroke, sepsis and ARDS. The global aim is to optimize chances for demonstrating benefit of an effective new therapy or therapeutic agent in the field of TBI.
Showing the most recent 10 out of 49 publications
