Drug intervention trials in sepsis: divergent results.

CONTEXT: Important advances have been made in our understanding of severe sepsis. Outcome can be improved by targeted interventions, including early and appropriate antibiotic therapy and goal-directed resuscitation, and might be further improved by selective decontamination of the digestive tract, tight control of glucose, and possibly by giving corticosteroids to selected patients. Drugs that target specific steps in the septic cascade include cytokine inhibitors, anti-endotoxins, and the three naturally occurring anticoagulants. Only one of these trials, which assessed the efficacy of activated protein C, reported significant improvements in outcome. Translation of these results into practice has been hampered by high drug costs, and by apparent discrepancies between interim results and final outcomes in two of the trials with natural anticoagulants. STARTING POINT: Recently, Steven Opal and colleagues (Crit Care Med 2004; 32: 332-41) reported a randomised trial with platelet-activating-factor acetylhydrolase to suppress the inflammatory response in septic patients. No effects on outcome were observed (mortality 24% with placebo vs 25% with the intervention). By contrast, Jose Garnacho-Montero and colleagues, in a cohort study (Crit Care Med 2003; 31: 2742-51), saw large mortality reductions with initially appropriate choice of antibiotics in septic patients (19.8% reduction overall and 43.4% in patients with septic shock). These benefits were higher than those even in the most successful trial with an antisepsis agents, underscoring the importance of basic measures in severe sepsis. WHERE NEXT? Initial management in severe sepsis should include early goal-directed fluid resuscitation, appropriate antibiotic treatment, and surgical-site control. Intensive-care units should be run by specialists, with adequate medical and nursing staffing. Tight regulation of glucose, selective decontamination of the digestive tract, and moderate-dose corticosteroids in selected cases should be considered. Expensive new drugs, such as activated protein C, might further improve outcome, but should be considered only when organisational aspects and supportive care have been optimised.