Nosocomial infections develop in 2 to 5 percent of patients undergoing clean extra-abdominal surgeries and after approximately 20 percent of intra-abdominal procedures. Patients who develop surgical-site infections have mortality rates double those of other surgical patients. Surgical-site infections also may lead to admission to intensive care units, extended hospital stays and readmission after discharge, and substantially increased health care costs. The National Surgical Infection Prevention Project aims to reduce mortality and morbidity from surgical-site infections by promoting appropriate use of prophylactic antimicrobials in surgical practice. National experts have developed performance measures to monitor the appropriate initiation, selection, and discontinuation of prophylactic antimicrobials in surgery. The project monitors Medicare patients undergoing procedures for which antimicrobials are clearly indicated, such as open-chest and vascular surgery, abdominal and colorectal procedures, hip and knee arthroplasty, and vaginal or abdominal hysterectomy. The project’s initial recommendations (made in 2002) recently have been revised by an expert panel.

After reviewing the timing of antimicrobial prophylaxis, the panel recommends that the agent be administered as near as possible to the time of incision. The infusion of the first dose should begin 60 minutes or less before the incision, with the exception of fluoroquinolones and vancomycin (Vancocin); with these agents, the infusion should be initiated 120 minutes before incision to minimize antibiotic-associated reactions. Administration at the time of anesthesia induction appears to result in safe and adequate drug levels in tissues. If a tourniquet is required, the antimicrobial infusion should be completed before the tourniquet is inflated.

Studies have not shown added benefit from multidose prophylaxis or prophylaxis given after wound closure. In addition, because prolonged use of antimicrobials may contribute to bacterial resistance, the guidelines recommend that prophylaxis end within 24 hours of surgery. The only exception may be in cardiothoracic surgery, where experts variously recommend administration up to 24 or 72 hours after surgery. Control of hyperglycemia reduces rates of surgical-site infection, and new evidence supports supplemental oxygenation and aggressive fluid resuscitation.

The choice of antimicrobial agent is based on drug effectiveness and patient-related factors, particularly potential adverse effects. Drug allergies may be poorly documented. The patient’s medical history should identify true allergy or severe adverse drug reaction to avoid denying a patient the agent of choice because of false evidence of adverse reaction. Skin testing may be necessary to establish allergic status. Whenever possible, a single antimicrobial agent is recommended, with the choice based on the type of surgery. The use of vancomycin has not been demonstrated to reduce the incidence of methicillin-resistant Staphylococcus aureus (MRSA). Nasal mupirocin (Bactroban) can reduce nasal carriage of S. aureus but has not been shown to reduce rates of surgery-related infection. Postoperative antibiotic use for more than 24 hours and discharge to a long-term care facility are the only factors associated with development of MRSA infection at the surgical site. Vancomycin is indicated, however, for surgical prophylaxis in patients already colonized with MRSA.

Although general recommendations have been made for common surgeries, the authors stress that recommendations must be individualized for each situation and will change over time as patterns of bacterial resistance change and the results of more research become available. The enduring principles are that adequate concentrations of the most appropriate antimicrobial agent should be present in the tissues at the time of maximal risk of bacterial contamination. In addition to pharmacologic properties, patient safety and cost are factors in the choice of agent.