Résumé : Purpose: The objective of this study was to develop a model to identify patients in whom drotrecogin alfa (activated) (DAA) might be administered for periods shorter than the recommended 96 hours. Methods: We did a retrospective chart review of all 124 patients treated with a standard 96-hour infusion of DAA in a 31-bed department of intensive care. Using a stepwise approach, we identified and combined parameters that could help predict outcomes to achieve the best sensitivity associated with 100% specificity. Results: Twenty-one (17%) of the 124 patients had a favorable outcome (left the intensive care unit within 5 days of DAA initiation); of these, 11 had an increase in arterial pH in the first 24 hours of treatment compared with 22 (21%) of the 103 patients with intermediate (intensive care unit stay >5 days after DAA initiation) or unfavorable (died within 5 days of DAA initiation) outcomes (P = not significant). Eight (72.7%) of these 11 patients and no other patient showed a decrease in sequential organ failure assessment score of at least 50% during the first 24 hours (P < .001). By combining these 2 variables, we could identify, with 100% specificity, 8 of the patients with a favorable outcome (38%) who made a prompt recovery. Conclusions: A simple model based on sequential organ failure assessment score and arterial pH can help identify patients with a rapid favorable course in whom a shorter duration of DAA treatment may be justified. © 2009 Elsevier Inc. All rights reserved.