Knee prosthesis infection is a severe complication. The use of a tourniquet during surgery impairs the efficacy of the antibiotic prophylaxis. We hypothesize that the antibiotic administration before tourniquet release decrease the infection rate. Methods: patients who undergo a total knee arthroplasty will be randomized to receive one of the following regimens of antibiotic prophylaxis:Standard: cefuroxime 1.5 g i.v. 10 min before tourniquet + placebo 10 min before tourniquet release + cefuroxime 1'5 g i.v. 6h after closing surgical wound.Experimental: cefuroxime 1.5 g i.v. 10 min before tourniquet + cefuroxime 1.5 g i.v. 10 min before tourniquet release + cefuroxime 1'5 g i.v. 6h after closing surgical wound.

Objective: when surgery is performed under ischemia, the ideal timing for the administration of antibiotic prophylaxis has not been well defined. Currently, antibiotics are given 10 minutes before ischemia and the infection rate is about 3-5%. In 2005 we performed a study where the efficacy of the standard prophylaxis was compared with a new regimen that consisted in administering the antibiotic 10 min. before tourniquet release. The results were positive although the difference was not statistically significant. In the present study we are proposing a new protocol where the antibiotic is given 10 min. before the inflation of tourniquet and 10 min. before its release. Methodology: all patients who undergo a total knee arthroplasty will be randomized to receive one of the following regimens of antibiotic prophylaxis:Standard: cefuroxime 1.5 g i.v. 10 min before tourniquet + placebo 10 min before tourniquet release + cefuroxime 1'5 g i.v. 6h after closing surgical wound.Experimental: cefuroxime 1.5 g i.v. 10 min before tourniquet + cefuroxime 1.5 g i.v. 10 min before tourniquet release + cefuroxime 1'5 g i.v. 6h after closing surgical wound.The infection rate with standard regimen is 3.4%. We believe that the new regimen will decrease the rate of infection to 1%. To demonstrate differences between both arms, it will be necessary to include 666 patients per arm.

Inclusion Criteria:

• patients undergoing knee arthroplasty

Exclusion Criteria:

• patients allergic to penicillin