Résumé : Purpose: To enhance the understanding of severe sepsis, a database of patients from multiple clinical trials spanning a 6-year period was constructed. Initial analyses evaluated the 28-day survival in the placebo group and further assessed the treatment effect of drotrecogin alfa (activated) (DrotAA). Methods: Five severe sepsis studies with similar entry criteria were combined, and baseline characteristics and 28-day mortality were evaluated (4459 severe sepsis patients; placebo, n = 1231; DrotAA, n = 3228). An integrated data analysis with propensity score adjustment was performed. Twenty-one variables selected by stepwise logistic regression were included in a propensity score of differences between the 2 groups of patients. Results: Over the 6-year period of these trials, there was no change in placebo mortality rates overall (P = .67), nor in subgroups of Acute Physiology and Chronic Health Evaluation score ≥25 (P = .73) or multiple organ dysfunction (P = .38). The adjusted relative hazard risk for DrotAA patients was 0.84 (95% confidence interval, 0.73-0.95; P = .007). Serious bleeding (0.8% in placebo vs 3.5% in DrotAA, P < .0001) was increased during the DrotAA infusion period. Conclusions: Initial analyses indicate that placebo mortality remained unchanged over a recent 6-year period. These analyses also further substantiate that treatment with DrotAA is associated with improved survival. © 2007 Elsevier Inc. All rights reserved.