Nosocomial infection rates in veterinary referral hospitals : using syndromic surveillance to establish baseline rates

Abstract

Nosocomial, or hospital-acquired, infections are considered to be the most common complication affecting hospitalized human patients, but their impact on hospitalized veterinary patients is less well understood. In fact, the incidence of nosocomial infections that occur in veterinary hospitals has not been established. There is evidence, however, that nosocomial infections are of great consequence in veterinary medicine and can have considerable negative effects on the individual patient as well as on the veterinary hospital as a whole. Due to the increased risk of infection in hospitalized patients, it is anticipated that some nosocomial infections will occur. Establishing a baseline rate of infection using surveillance techniques will allow investigators to ascertain the proportion of infections that can be prevented using infection control measures. The purpose of this study was to establish baseline rates of infection using a syndromic surveillance system created for implementation in small animal and equine referral hospitals. This study included weaned dogs, cats, and horses (n=2248) that were hospitalized in the critical care unit of one of five participating veterinary hospitals during a 12 week period in 2006. Clinicians actively performed syndromic surveillance on hospitalized patients and reported their results no later than the time of the patient’s discharge from the hospital. Adjusted rates of nosocomial events were estimated using Poisson regression, and risk factors associated with an increased risk of developing a nosocomial event were analyzed using multivariable logistic regression. Adjusting for hospital of admission, 19.7% of horses, 16.3% of dogs, and 12% of cats included in this study were reported to have experienced a nosocomial event occur during hospitalization. The only risk factor found to have a positive association with the development of a nosocomial event in all three species was placement of a urinary catheter. Surgical site inflammation and intravenous catheter site inflammation were two of the most commonly reported events across all species. Results of this study suggest that nosocomial event rates can be established using syndromic surveillance systems in multiple hospitals. Data pertinent to risk factors for the occurrence of nosocomial events can also be effectively collected using the same technique. Further research is warranted in order to evaluate how generalizable these results are to other veterinary healthcare settings.