Distribution of the probability of survival is a strategic issue for randomized trials in critically ill patients

Background: Many randomized clinical trials in trauma have failed to demons trate a significant improvement In survival rate. Using a trauma patient da tabase, we simulated what could happen in a trial designed to improve survi val rate in this setting. Methods: The predicted probability of survival was assessed using the TRISS methodology in 350 severely injured trauma patients. Using this probabilit y of survival, the authors simulated the effects of a drug that may increas e the probability of survival by 10-50% and calculated the number of patien ts to be included in a triad, assuming alpha = 0.05 and beta = 0.10 by usin g the percentage of survivors or the individual probability of survival. Ot her distributions (Gaussian, J shape, uniform) of the probability of surviv al were also simulated and tested. Results: The distribution of the probability of survival was bimodal with t wo peaks (< 0.10 and > 0.90), There were major discrepancies between the nu mber of patients to be included when considering the percentage of survivor s or the individual value of the probability of survival: 63,202 versus 2,8 48 if the drug increases the probability of survival by 20%. This discrepan cy also occurred in other types of distribution (uniform, J shape) but to a lesser degree, whereas it was very limited in a Gaussian distribution. Conclusions: The bimodal distribution of the probability of survival in tra uma patients has major consequences on hypothesis testing, leading to overe stimation of the power. This statistical pitfall may also occur in other cr itically ill patients.