Effect of Two Glasgow Outcome Scale-Extended Scoring Methods on Traumatic Brain Injury Clinical Trial Design: A...

Area of Science:

• Neuroscience
• Clinical Trials
• Biostatistics

Background:

• The Glasgow Outcome Scale-Extended (GOSE) is a standard outcome measure in traumatic brain injury (TBI) research.
• GOSE can be administered to assess overall functional limitations (GOSE-All) or TBI-specific limitations (GOSE-TBI).
• The method of GOSE administration can influence clinical trial design, particularly sample size and statistical power.

Purpose of the Study:

• To assess the impact of using GOSE-All versus GOSE-TBI on sample size and power in TBI clinical trials.
• To determine how different TBI severity and extracranial injury criteria affect these estimations.
• To evaluate the influence of assessment timing (2 weeks to 12 months post-injury) on study design parameters.

Main Methods:

• Utilized data from 2,288 participants in the prospective observational TRACK-TBI study.
• Simulated the effects of GOSE-All and GOSE-TBI administration on sample size and power using statistical tests (e.g., Mann-Whitney).
• Examined scenarios with dichotomized outcomes and varying Glasgow Coma Scale (GCS) severity levels.

Main Results:

• Using GOSE-All instead of GOSE-TBI necessitates an 8-158% increase in sample size for Mann-Whitney tests, or power decreases significantly if sample size is not adjusted.
• Dichotomizing the GOSE outcome requires a 6-165% increase in sample size.
• Sample size requirements are highest for milder TBI populations and decrease over time post-injury for GCS 13-15 patients.

Conclusions:

• Researchers must pre-define the GOSE administration method (GOSE-All vs. GOSE-TBI) based on study aims to ensure appropriate power and sample size calculations.
• Clear instructions for respondents and examiners regarding the attribution of disability are essential.
• The chosen assessment method must be explicitly stated in protocols, publications, and data documentation, and emphasized in examiner training.