A total of 337 patients undergoing coronary artery bypass grafting or cardiac valve replacement were randomly assigned to receive cefazolin (1 g every 8 h [q8h]), cefamandole (2 g q6h), or cefuroxime (1.5 g ql2h) as an intravenous antibiotic prophylaxis. All drugs were administered within 60 min before the initial incision and were continued for 48 h postoperatively. No adverse effects related to the study drugs were observed. The percentage of patients with postoperative infection was 9% for the cefazolin group, 6% for the cefamandole group, and 5% for the cefuroxime group or 6.5% overall. There were more infection sites in patients treated with cefazolin than in those treated with cefuroxime (P = 0.05) or cefamandole (P = 0.06). Fewer wound infections ocwcurred with cefuroxime (P < 0.01) and cefamandole (P = 0.06) than with cefazolin. Analyses of the prophylactic regimens used in this study showed cefazolin and cefuroxime to be less costly than cefamandole.

Two cases in which hyperbilirubinemia produced falsely elevated aminoglycoside concentrations, using the Emit assay, are described. Patients' sera were analyzed by both the Emit and a microbiologic assay. The first patient had a 26-percent greater tobramycin concentration with the Emit assay, as compared with the microbiologic assay, when the total bilirubin was 20.1 mg%. For the same patient, there was essentially no difference between the two assays when the total bilirubin was 4 mg%. The Emit concentrations in the other patient were 34-percent elevated with a total bilirubin of 9.7 mg%. Pharmacokinetic analysis also was performed on serum levels reported by each method. Calculated kinetic parameters and dosage regimens varied greatly. Patients with total bilirubin concentrations greater than 4 mg% may need to have aminoglycoside serum levels assayed by a method other than the Emit assay.

No abstract