Objective
To test the hypothesis that antibiotics leading to greater endotoxin release are associated with greater mortality in septic trauma patients.
Design
Post hoc analysis of data from a previously conducted prospective, randomized, multicenter study designed to evaluate the efficacy of interferon gamma in preventing infection and death in trauma patients.
Setting
Nine level I trauma centers.
Patients
Severely injured trauma patients at high risk for sepsis. Eighty percent (N=334) of the enrolled patients developed some manifestation of gram-negative sepsis, defined by the administration of gram-negative specific antibiotics.
Main Outcome Measures
The in-hospital mortality rate of patients who received penicillin-binding protein 3/tumor necrosis factor (PBP3/TNF)–specific antibiotics associated with the greatest degree of endotoxin release and TNF production (PBP3/TNF group, n=78: aztreonam. ceftazidime, and cefotaxime sodium) was compared with that of patients not receiving these agents (non-PBP3/TNF group, n=256).
Results
Mortality in the PBP3/TNF group (17%) was higher than in the non-PBP3/TNF group (8%, P=.02). The two groups were similar in their mean (±SD) Injury Severity Scores (34±9), ages (31 ± 12 years), and initial degree of bacterial contamination.
Conclusions
Antibiotics that are associated with greater release of endotoxin and production of TNF are also associated with greater mortality in septic trauma patients. Decisions regarding antibiotic administration may need to consider the endotoxin-releasing properties of antibiotics in addition to their antibacterial sensitivity spectrum. Prospective studies of the effect of endotoxin-releasing properties of antibiotics on mortality are warranted.
(Arch Surg. 1995;130:1234-1241)
