OBJECTIVE: To compare clinical characteristics and risk factors of suppurative postsurgical mediastinitis according to its etiology. METHODS: Suppurative postsurgical mediastinitis developed in 45 (2.5%) of 1779 patients who underwent open-heart surgery at the Hospital Virgen del Rocío in Seville, Spain, from 1986 to 1996. Microbiological diagnosis was available in 42 patients. RESULTS: Gram-negative rods were isolated in 19 cases and Gram-positive cocci in 23 cases. Seventeen isolates (38%) were sensitive to the antimicrobial agent used perioperatively. Patients with Gram-negative rod infection had a longer duration of bypass (127plus minus36 min versus 96plus minus34 min, p<0.01), and a worse postoperative condition. Longer mechanical ventilation (4plus minus7 days versus 1plus minus2 days, p<0.05) and concomitant infection in a remote site (pulmonary and/or urinary infection) were more frequently observed in this group than in patients with Gram-positive infections (58% versus 22%, p<0.05). Twenty patients (51%) were bacteremic. The mortality rate was 20% (five of 45). CONCLUSIONS: Preventable postoperative remote-site infection may lead to mediastinitis, especially if Gram-negative rods are involved.
OBJECTIVE: To compare clinical characteristics and risk factors of suppurative postsurgical mediastinitis according to its etiology. METHODS: Suppurative postsurgical mediastinitis developed in 45 (2.5%) of 1779 patients who underwent open-heart surgery at the Hospital Virgen del Rocío in Seville, Spain, from 1986 to 1996. Microbiological diagnosis was available in 42 patients. RESULTS: Gram-negative rods were isolated in 19 cases and Gram-positive cocci in 23 cases. Seventeen isolates (38%) were sensitive to the antimicrobial agent used perioperatively. Patients with Gram-negative rod infection had a longer duration of bypass (127plus minus36 min versus 96plus minus34 min, p<0.01), and a worse postoperative condition. Longer mechanical ventilation (4plus minus7 days versus 1plus minus2 days, p<0.05) and concomitant infection in a remote site (pulmonary and/or urinary infection) were more frequently observed in this group than in patients with Gram-positive infections (58% versus 22%, p<0.05). Twenty patients (51%) were bacteremic. The mortality rate was 20% (five of 45). CONCLUSIONS: Preventable postoperative remote-site infection may lead to mediastinitis, especially if Gram-negative rods are involved.