Cardiac surgery in a high-risk group of patients: is prolonged postoperative antibiotic prophylaxis effective?

In a prospective, randomized study, postoperatively prolonged antibiotic prophylaxis is evaluated in a high-risk group of patients undergoing cardiac operations. These patients had postoperative low cardiac output necessitating inotropic support and intraaortic balloon pumping.

Between January 1991 and 1994, 53 patients were enrolled in the study (42 men, mean age 65 years). All patients received the usual perioperative (24 hours) cefazolin prophylaxis. In the study group (n = 28) a prolonged regimen of prophylaxis with ticarcillin/clavulanate was performed for 2 days and vancomycin was added in a low dose until removal of the intraaortic balloon pump. The control group (n = 25) did not receive a prolonged regimen of prophylaxis. Follow-up ended at hospital discharge.

Early mortality was 7 of 28 patients (25%) in the prophylaxis group and 8 of 25 patients (32%) in the control group (p = 0.397). Defined infections (pneumonia, n = 22; sepsis, n = 8; deep sternal wound infection, n = 2) occurred in 50% of the study group and 68% of the control group (p = 0.265). In all patients with septicemia, only coagulase-negative staphylococci could be isolated from the bloodstream (5 patients in the prophylaxis group vs 3 in the control group). Infectious parameters were controlled daily and did not differ significantly between groups. A total of 1158 bacteriologic tests were performed (blood cultures, n = 389; intravascular catheters, n = 208; bronchial aspirates, n = 411; intraaortic balloon pumps, n = 42; wound secretions, n = 108) showing bacterial growth in 322 (28%) without a significant difference between the groups. In the prophylaxis group, 13 intravascular catheters and intraaortic balloon pumps showed bacterial growth versus 11 in the control group. No side effects were seen.

In a high-risk group of patients undergoing cardiac operations, infectious outcome could not be effectively influenced by an additional and prolonged postoperative prophylaxis regimen with low-dose vancomycin and ticarcillin/clavulanate. Low-dose vancomycin did not reduce the rate of infections or colonizations of intravascular catheters with gram-positive organisms.