A method for classifying patients according to the nosocomial infection risks associated with diagnoses and surgical procedures.

To compare validly the nosocomial infection rates (NIRs) in groups of patients studied from different time periods and/or different hospitals, one must control for the important factors that influence a patient's susceptibility to infection. The authors developed a method for assessing one component of nosocomial infection risk, based on patients' diagnoses and surgical procedures. This method classifies patients according to their risk of developing a nosocomial infection at each of four infection sites and at all four sites combined. Applying the method to data collected on 136,516 patients from 276 hospitals studied in the SENIC Project (Study on the Efficacy of Nosocomial Infection Control), the authors found that NIRs increased according to the predicted ranking of risk categories, even when the analyses were stratified individually by age, sex, hospital service and exposure to urinary catheterization or continuous ventilatory support. Depending on the site of infection, the rate increased as much as 100-fold from low-risk to high-risk categories. The data indicate that infection risk as assessed with this classification method will account for some of the variation in NIRs due to differences in patients' clinical conditions. Further analyses using multivariate techniques must be performed to explore in detail the relative importance of this risk classification in comparison with other risk factors and to determine which factors must be controlled in SENIC analyses.