How many patients with severe sepsis are needed to confirm the efficacy of drotrecogin alfa activated? A Bayesian design
Authors:
- 27 May 2008
- journal article
- research article
- Published by Springer Nature in Intensive Care Medicine
- Vol. 34 (10) , 1804-1811
Abstract
Objectives Controversies concerning the pivotal trial and negative results from other Drotrecogin alfa activated (DAA) trials have raised questions about its efficacy in severe sepsis patients. Based on all available evidence, our study aimed to respond to: (1) What is the current probability that DAA is not better than the control? If the current probability is not small, e.g. greater than 0.05, then (2) How many patients will be needed for the DAA confirmatory trial? Design We applied the Bayesian methodology to all randomized trials which tested anti-thrombotic therapies for severe sepsis. Prior distributions were defined as severe skeptic (all anti-thrombotic trials or DAA trials); moderate skeptic (anti-thrombotic trials with high risk of death); mild skeptic (DAA trials only); enthusiastic (DAA trials with high risk of death). Results The total study sample includes 7,683 patients: DAA (N = 3,143); Anti-thrombin-III (N = 2,581); tissue factor pathway inhibitor (N = 1,959). Answer 1: All current probabilities that DAA is not better than the control (except for the enthusiastic analysis) range from 0.14 to 0.48, which strongly suggest the need for a confirmatory trial. Answer 2: The number of patients necessary for the DAA confirmatory trial ranges up to 8,350 for the severe skeptic; 730–810 for moderate skeptic; 550–685 for mild skeptic; and zero for enthusiastic analysis. Conclusions A confirmatory trial with approximately 600 patients with severe sepsis and high risk of death can provide a convincing answer for both the mild and moderate skeptic physicians concerning the efficacy of DAA in severe sepsis.Keywords
This publication has 40 references indexed in Scilit:
- Drotrecogin Alfa (Activated) in Severe SepsisNew England Journal of Medicine, 2006
- Bayesian clinical trialsNature Reviews Drug Discovery, 2006
- Struggle for implementation of new strategies in intensive care medicine: anticoagulation, insulin, and lower tidal volumesJournal of Critical Care, 2005
- Why clinicians are natural bayesiansBMJ, 2005
- Prior convictionsJACC, 2004
- Risks and Benefits of Activated Protein C Treatment for Severe SepsisNew England Journal of Medicine, 2002
- Severe Sepsis — A New Treatment with Both Anticoagulant and Antiinflammatory PropertiesNew England Journal of Medicine, 2001
- Efficacy and Safety of Recombinant Human Activated Protein C for Severe SepsisNew England Journal of Medicine, 2001
- Sifting the evidence---what's wrong with significance tests? Another comment on the role of statistical methodsBMJ, 2001
- ANTITHROMBIN III SUPPLEMENTATION IN SEVERE SEPSISShock, 1997