Table 3.
Potential Reasons for the Failure of Clinical Trials for Therapies for TBI: Reasons Related to Clinical Trial Design and Methods
| Reason | Summary | References |
|---|---|---|
| Single center | RCTs conducted in single centers may be less reliable than multiple center RCTs | Bafeta et al.69; Bellomo et al.70 |
| Follow-up bias | Patient selection bias may occur due to loss of follow-up in some patient populations (low socioeconomic status, substance abuse, violent injury mechanisms) | Corrigan et al.67 |
| Randomization/Blinding | Intervention effect estimates may be exaggerated due to inadequate or unclear random sequence generation or blinding, especially with subjective end-points | Bragge et al.14; Page et al.68 |
| Sample size | Inadequate sample size to produce adequate statistical power | Burke et al.13; Bragge et al.14; Button et al.77 |
| Therapy dose selection | Translation of doses used in pre-clinical studies should be based on physiological, pharmacokinetic, and toxicology data | Blanchard and Smoliga79 |
| Patient selection | Select patient populations most likely to be helped by therapy | Hawryluk et al.15; Narayan et al.35 |
| Outcome errors | Misclassification of outcomes (e.g., GOS score) can significantly reduce effect size | Lu et al.85 |
| Inadvertent bias | Pressures to publish, industry bias, positive results, government vs. industry support | Fanelli et al.159; Dwan et al.163; Fanelli et al.164; Easterbrook et al.165 |