Classification of the surgical wound: a time for reassessment and simplification.

Classification of the surgical wound in the operating room by surgeons and nurses is a time-honored routine that has been practiced for at least 30 years, since the time of the National Academy of Sciences National Research Council study on the influence of ultraviolet irradiation on surgical wound infection. This traditional method uses four classes of wounds based on the risk level and type of contamination expected or observed at operation.2-3 Clean surgical wounds (Class I) are those in which only exogenous (airborne) contamination is expected or observed and the predicted wound infection rate is approximately 2%, largely due to gram-positive microorganisms such as Staphylococcus aureus. Clean-contaminated (Class II) wounds are those in which generally both exogenous and endogenous (aerobic-anaerobic) bacterial contamination occur during elective operations. The infection rate in this category is estimated at 5% to 15% and is usually due to the polymicrobic endogenous flora. Contaminated wounds (Class III) are those with early endogenous leakage or delayed exogenous contamination in the absence of established clinical infection and carry a greater than 15% infection rate. In dirty wounds (Class IV) where active infection is encountered during operation, a postoperative infection rate of greater than 30% is anticipated. During the last decade, there have been problems identified with the use of this traditional wound classification system and the accuracy of the predicted infection rates in each category. The major limitation lies in the lack of attention to the varying risk of infection among subjects in each class of wound.3 Haley et al4 at the Centers for Disease Control and Prevention were among the first to publish on the importance of identifying the varying individual risks for infection among patients in each of the traditional four categories of wounds. Using stepwise multiple logistic regression in nearly 59,000 patients, they developed a new predictive index using four risk factors that, when studied in an equally large group of surgical patients, was able to accurately predict the incidence of wound infection. They identified three different risk groups (low, medium, and high) in both the clean and clean-contaminated wound classes and only two risk groups (medium and high) in the contaminated and dirty classes. There were surprisingly close similarities of predicted infection rates among Class I and II and in Class III and IV wounds. The infection rates in Class I wounds were reported as 1.1% (low risk), 3.9% (medium risk), up to 15.8% (high risk) with an overall rate of 2.9%, while they were predicted as 0.6% (low), 2.8% (medium), up to 17.7% (high), with an overall 3.9% in Class II wounds. The infection rates in Class III wounds, where no low-risk group was identified, were 4.5% (medium) and up to 23.9% (high) with an overall rate of 8.5%, while in Class IV wounds, also without a low-risk group, they were predicted at 6.7% (medium), up to 27.4% (high), and 12.6% overall. It seems apparent from the above study that