What question did this study set out to answer?

To determine how the underlying distribution of survival probability in critically ill trauma patients affects sample size requirements and statistical power in clinical trials.

July 1, 2001Anesthesiology

Distribution of the Probability of Survival Is a Strategic Issue for Randomized Trials in Critically Ill Patients

View Full Paper

Ask AI

Bookmark

Authors

BRBruno RiouCross-Cutting Cardiology PLPaul LandaisUniversité de Montpellier BVBenoît VivienHôpital Necker-Enfants Malades

Discussion

Member takes

Implication

Simulation study reveals bimodal survival probability skews sample size calculations in trauma patients, suggesting standard clinical trial designs severely overestimate statistical power.

Key Points

To determine how the underlying distribution of survival probability in critically ill trauma patients affects sample size requirements and statistical power in clinical trials.
Assessed predicted survival probability using TRISS methodology in N=350 severely injured trauma patients from a patient database.
Simulated drug effects yielding a 10% to 50% relative increase in survival probability, calculating required sample sizes at alpha = 0.05 and beta = 0.10 based on aggregate survivor percentage versus individual probabilities across bimodal, uniform, J-shaped, and Gaussian distributions.
Identified a distinctly bimodal distribution of survival probabilities among trauma patients, showing prominent peaks at <0.10 and >0.90.
Demonstrated marked sample size discrepancies for an intervention raising survival probability by 20%, requiring 63,202 patients using survivor percentages compared to 2,848 using individual survival probabilities.
Showed that while discrepancies also appeared in uniform and J-shaped distributions, power overestimation was minimal in Gaussian distributions.

Cite This Study

Riou et al. (2001) studied this question.

synapsesocial.com/papers/6a909c2a6e7e048e0989e107 https://doi.org/10.1097/00000542-200107000-00014

View Full Paper

Ask AI

Bookmark

Implication

Simulation study reveals bimodal survival probability skews sample size calculations in trauma patients, suggesting standard clinical trial designs severely overestimate statistical power.

Distribution of the Probability of Survival Is a Strategic Issue for Randomized Trials in Critically Ill Patients

Discussion

Member takes

Implication

Key Points

Cite This Study

Distribution of the Probability of Survival Is a Strategic Issue for Randomized Trials in Critically Ill Patients

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider