ABSTRACT
BACKGROUND: Hospital-acquired infections have been suggested to affect patients' outcomes and raise health care costs. However, research regarding the change in rates and types of infections over time remains limited. METHODS: All patients who underwent cardiac surgical procedures from 1995 to 2012 at the Queen Elizabeth II Health Sciences Center in Halifax, Canada were identified. The prevalence of superficial surgical site infection (sSSI), deep surgical site infection (dSSI), urinary tract infection, sepsis, pneumonia, and leg site infection was examined to determine trends in infections over time. Nonparsimonious logistic regression models were created to identify independent preoperative predictors of length of stay and infection onset. RESULTS: A total of 19,333 consecutive patients underwent cardiac surgical procedures, of whom 2,726 (14%) contracted at least one postoperative infection. The incidence of infections increased from 8% to 20% during the 17-year span (p < 0.0001). The overall prevalence of infection types, from highest to lowest, was pneumonia (6%), urinary tract infection (6%), sepsis (3%), sSSI (2%), leg infection (2%), and dSSI (1%). After adjusting for clinical differences, postoperative infection was found to be an independent predictor of length of stay longer than 9 days. In turn, independent predictors for contracting a postoperative infection included operative era, advanced age of patients, and complex procedures. CONCLUSIONS: The incidence of infection increased nearly threefold since 1995 independent of patient- or procedure-related variables and was found to affect hospital length of stay significantly. Our findings highlight that efforts to monitor only rates of hospital-acquired infections may not in isolation help affect patient care.