Prospective alterations in therapy for penetrating abdominal trauma.

In a double-blind, randomized study, 170 patients with traumatic perforation of the gastrointestinal tract were administered an advanced-generation cephalosporin. Patients were divided into infection risk groups (< or = 40%, low; 40% to 70%, mid; and > 70%, high) at surgical closure using a logistic regression formula based on four proved risk factors--age, blood replacement, ostomy, and the number of organs injured. Patients in the low group received 2 days of antibiotic therapy; those in the mid to high group received 5 days of antibiotic therapy. Those patients in the low to mid group had primary wound closure; those in the high group had their wounds packed open and closed later. Most of the patients (144 [85%]) were in the low group. Their major and minor infection rates (10% and 12%, respectively) were not significantly different from 145 historic control subjects receiving 5 days of antibiotic therapy (9% major; 14% minor). Patients in the mid to high group showed a greater incidence of major infections (46%) but a similar incidence of minor infections (12%). The results indicate that risk factors can be used to identify low-risk patients who require only short-term antibiotic therapy and primary wound closure. The remaining patients are at greater risk for infection despite prolonged antibiotic therapy and delayed wound closure.