Sternal or mediastinal infections after heart operation occurs infrequently but carries a high cost in money, morbidity, and mortality. At our hospital, Staphylococcus species cause most of these infections and is uniformly sensitive to vancomycin. However, the relative low concentration of vancomycin in sternal bone was noted given intravenously. Topical vancomycin was therefore considered to apply on the cut edege of sternal bone before closure in patients receiving coronary artery bypass graft to evaluate the effect on reducing the infection rate of staphylococcus related mediastinitis (6). In addition, we try to observe that prophylactic use of topical vancomycin will increase the risk for acquiring resistant pathogen such as VISA (vancomycin-intermediate S. aureus), VRSA (Vancomycin-resistant S. aureus), VRE (vancomycin-resistant enterococcus)

Background Postoperative sterno-mediastinitis is a life-threatening complication. Although the incidence rate is about 0.75 to 1.4% in all open heart operations, the association with morbidity, mortality and cost is unacceptably high (1, 2). Some risk factors should be minimized as possible, such as preoperative days in the hospital, and poor controlled blood sugar and other chronic underlying conditions. In addition, patients should shower the night and morning before surgery with chlorhexadine and remove hair with clipper or depilatory crème instead of a shave. These basic rules of sterile technique can't be emphasized enough and may reduce the occurrence of postoperative sterno-mediastinitis form 7.5 to 0.8% (3, 4).

The matter of prophylactic antibiotics is controversial. Although everybody agrees the necessity of using prophylactic antibiotics but which to use and how long it should be administered remain a debate. The most commonly occurring bacterial offenders in the institution always should be considered in selecting of prophylactic antibiotics.

Staphylococcus spp. is most common pathogens in most institutions and also in our hospitals. It is currently recommended that the use of vancomycin for perioperative prophylaxis should be considered only in patients with a life-threatening allergy toβ-lactam antibiotics or at institutions with a high rate of infections caused by MRSA or methicillin-resistant S. epidermidis (5). However, the relative low concentration of vancomycin in sternal bone was noted when given intravenously (5). According to the study performed by Vander et al. in 1989, we also try to apply vancomycin over the cut sternal edges in patients receiving CABG (coronary artery bypassing graft) and to evaluate the effect on reducing the infection of staphylococcus related mediastinitis (6). In addition, we try to observe that prophylactic use of topical vancomycin will increase the risk for acquiring resistant pathogen such as VISA (vancomycin-intermediate S. aureus), VRSA (Vancomycin-resistant S. aureus), VRE (vancomycin-resistant enterococcus).

Aim We try to apply vancomycin over the cut sternal edges in patients receiving CABG (coronary artery bypassing graft) and to evaluate the effect on reducing the infection of staphylococcus related mediastinitis. In addition, we try to observe the use of topical vancomycin will induce the collateral effect on increasing the infections of resistant pathogen such as VISA (vancomycin-intermediate S. aureus), VRSA (Vancomycin-resistant S. aureus), VRE (vancomycin-resistant enterococcus) or other infections caused by GNB or yeast.

Materials and methods

Patients Patients who aged over 20 and receive CABG in our hospital, a 1000-beds medical center, from June 1, 2007 to Dec 31, 2007 will be included in our study. Patients who ever received vancomycin or teicoplanin or linezolids within one week will be excluded. Other exclusion criteria are patients in pregnant, allergic to vancomycin, and hesitate to this study plan. A register of inclusions will be established to avoid inclusion of the same patient twice.

Data collection

Preoperation A standard case report form was designed to collect patents' demographic data, laboratory data and medical conditions. The demographic data include age, gender, hospital unit (ward or ICU) before operation, and the TISS score if patients reside in ICU. The medical conditions include BMI (body mass index), smoking, medical disease, steroid or other immunosuppressants usage, left ventricular ejection fraction, cardiac index, prior hospital staying days, previous cardiac surgical procedure, previous radiotherapy over operation site, other infection diseases with antibiotics usage, catheters in place. Laboratory data including WBC, platelet counts, hemoglobulin, BUN, Cr, albumin and aminotransferase are recorded.

Operation During operation, ASA score (American Society of Anethesiologists risk score), main operators should be recorded before operation. In addition, overall operation time, which vessels be harvested if extracorporeal bypass used and the used time, if IABP (intraaortic balloon pumping), ECMO (extracorporeal membrane oxygenation), external pacemaker used during operation will be recorded.

Postoperation After operation, the units of blood product transfusion, catheter in placed, wound condition will be recorded. The condition of wound should be evaluated and be recorded everyday until postoperative 30th days. The thoracic wound condition was stratified into three grades: "0" = clean wound, "1"= local swelling, redness or heat and tenderness, "2" = purulent discharge. In addition, serum Vancomycin level will be checked everyday for one week in study group. The date and pathogens of concomitant infections, such as ventilator-associated pneumonia, catheter-related infection, urinary tract infection, donor site wound infection will be also recorded if develop. Once methicillin-resistant staphylococcal species yield, MIC (minimal inhibit concentration) of vancomycin and molecular typing of this pathogen will be performed.

Furthermore, the complications of sternal wound infection, including acute renal failure with or without long term hemodialysis, ARDS (acute respiratory distress syndrome), septic shock, disseminated intravascular coagulopathy, heptic dysfunction and mortality will be also recorded. The overall hospitalization days will also be recorded.

Pre - and Intra-operative preparation protocol Nasal swab and anal swab should be performed in all patients. All patients should shower the night and morning before surgery with chlorhexadine. Hair removal should be used clipper or depilatory crème instead of a shave just before operation. Prophylactic antibiotics with cefazolin 1g should be administered intravenous 30 minutes before operation, and one extra dose of 1g cefazolin will be administered if operation time over 2 hours. The gloves and instruments used to harvest saphenous vein will be changed before working in intrathoracic field. Before the sternum been closed, a hemostatic paste containing 1gm of powdered absorbable gelatin (Gelfoam, Upjohn Co., Kalamazoo, Mich.) and topical thrombin (1000 units/ml) (Armour Pharmaceutical Co., Kankakee, Ill.) will be applied to the cut edges of the sternum. Vancomycin mixture is prepared by 500mg of powdered vancomycin (Vancocin; Eli Lilly & Co.) been added to above mixture before application. Patients with odd-numbered medical record will be applied with vancomycin mixture. In either cases, the resultant material was made by adding sufficient thrombin to achieve a consistency some what like that of fresh pie dough (5).

Study endpoints Primary endpoint is to compare the infection rate of methicillin-resistant staphylococcal pathogen between the group of prophylactic systemic cefazolin used only and the group of topical vancomycin added.

Secondary endpoints are, first, to identify the associated risk factors and the consequent complications of sternal wound infections and then compare the cost-effective in these two groups. Secondary, we compare the pathogens, either inducing infection or colonization, between these two groups.

Definitions Surgical site infection is defined according to the CDC's suggestion (6). Patients with an intravascular device-related infection was defined as (1) at least one positive blood culture obtained from a peripheral vein, clinical manifestations of infections (i.e., fever, chills, and/or hypotension), and no apparent source for the bacteremia except catheter or (2) one positive semi-quantitative (>15 CFU/catheter segment) whereby the same organism (species and antibiogram) is isolated form the catheter segment and peripheral blood (6). Ventilator associated pneumonia is defined Patients had new or progressive pulmonary infiltrates 48 hours after endotracheal intubation plus 1 of the following: positive blood culture results, radiographic evidence of cavitation, or histologic evidence of pneumonia, or 2 or more of the following: new fever, new leukocytosis, or grossly purulent tracheal aspirates (7).

Statistical analysis For the analysis, means and standard deviation (SD) were calculated as summaries of continuous variables. The comparison was performed with Mann-Whitney U method for continuous variables, and with Chi-square test, Fisher's exact tests, or ANOVA for categorical variables. Factors associated with death in univariate analysis (p<0.1) were further evaluated with multivariate analysis. Data were collected in a Microsoft Excel database (Microsoft Excel 2001; Microsoft Corporation, Seattle, WA) and analyzed with SPSS software for Window (Release 10.0; SPSS, Inc., Chicago, IL).

Inclusion Criteria:

• Patients with coronary artery disease underwent aoronary artery bypass grafting

Exclusion Criteria:

• Patients who ever received vancomycin or teicoplanin or linezolids within one week will be excluded. Other exclusion criteria are patients in pregnant, allergic to vancomycin, and hesitate to this study plan.