Wound site infections are a major source of postoperative illness, accounting for approximately a quarter of all nosocomial infections. National studies have defined the patients at highest risk for infection in general and in many specific operative procedures. Advances in risk assessment comparison may involve use of the standardized infection ratio, procedure-specific risk factor collection, and logistic regression models. Adherence to recommendations in the 1999 Centers for Disease Control and Prevention guidelines should reduce the incidence of infection in surgical patients.

We conducted a prospective, randomized, double-blind trial comparing ertapenem (1 g once daily) with piperacillin-tazobactam (3.375 g every 6 h) as parenteral treatment for 540 adults with complicated skin and skin-structure infections. The most common diagnoses were skin or soft-tissue abscesses and lower-extremity infections associated with diabetes. The mean duration (+/- standard deviation) of therapy was 9.1+/-3.1 days for ertapenem and 9.8+/-3.3 days for piperacillin-tazobactam. At the assessment of primary efficacy end point, 10-21 days after treatment, 82.4% of those who received ertapenem and 84.4% of those who received piperacillin-tazobactam were cured. The difference in response rates, adjusting for the patients' assigned strata, was -2.0% (95% confidence interval, -10.2% to 6.2%), indicating that the response rates in the 2 treatment groups were equivalent. Cure rates for the 2 treatment groups were similar when compared by stratum, diagnosis, and severity of infection. The frequency and severity of drug-related adverse events were similar in the treatment groups.

The effectiveness of short-term, low-dose, preoperative oral administration of neomycin and erythromycin base combined with vigorous purgation in reducing the incidence of wound infections and other septic complications of elective colon and rectal operations has been studied in a prospective, randomized, double-blind, clinical trial. One hundred and sixteen patients completed the study; all received mechanical preparation; 56 received neomycin-erythromycin base while 60 received an identical appearing placebo. The two patient groups were comparable in age distribution, clinical diagnoses, associated systemic diseases, types of operation performed and similar clinical features. The overall rate of directly related septic complications was 43 per cent in the placebo group and 9% in the group receiving neomycin and erythromycin base. The wound infection rates were 35% in placebo and 9% in antibiotic treated patients. Oral administration of neomycin and erythromycin base together with vigorous mechanical cleansing reduces the risk of septic complications after elective colo-rectal operations.

To identify the risk factors for the development of postoperative septic complications in patients with intestinal perforation after abdominal trauma, and to compare the efficacies of single-drug and dual-drug prophylactic antibiotic therapy, we studied 145 patients who presented with abdominal trauma and intestinal perforation at two hospitals between July 1979 and June 1982. Logistic-regression analysis showed that a higher risk of infection (P less than 0.05) was associated with increased age, injury to the left colon necessitating colostomy, a larger number of units of blood or blood products administered at surgery, and a larger number of injured organs. The presence of shock on arrival, which was found to increase the risk of infection when this factor was analyzed individually, did not add predictive power. Patients with postoperative sepsis were hospitalized significantly longer than were patients without infection (13.8 vs. 7.7 days, P less than 0.0001). Both treatment regimens--cefoxitin given alone and clindamycin and gentamicin given together--resulted in similar infection rates, drug toxicity, duration of hospitalization, and costs.