Determinants of Wound Infection Incidence after Isolated Coronary Artery Bypass Surgery in Patients Randomized to Receive Prophylactic Cefuroxime or Cefazolin

Abstract

In this open-label, randomized drug study, we compared two cephalosporin prophylactic regimens, one using cefazolin and one using cefuroxime, in 100 patients having coronary bypass surgery. Additional epidemiological data were collected to identify the patient at higher risk for acquiring an infection. Patients were categorized into four groups: (1) no infection; (2) clinically determined infection without a culture or prescription of additional antibiotics; (3) clinical infection with no or negative wound culture and prescription of additional antibiotics; and (4) clinical infection with positive culture and need for additional antibiotics. Seven cefuroxime patients (13.5%) and 9 cefazolin patients (18.8%; p = 0.471) had a wound that became clinically infected (Groups 2–4). In a univariate analysis, 11 variables were statistically associated with the development of a wound infection. A logistic regression model defined 3 variables at an α level of 0.05 and 3 at an α level of 0.10 that predicted a wound infection. Patients were identified at high risk of wound infection if they had postoperative weight gain, long operative hospitalization, prolonged use of a Foley catheter, postoperative use of blood products, and operation performed by two specific surgeons. Our results indicated that closer observation of the high-risk patients and a definition of the mechanism of the infections are needed.