Hospital-wide surveillance of catheter-related bloodstream infection: from the expected to the unexpected.

Catheter-related bloodstream infections (CRBSIs) are among the most frequent healthcare-associated infections and cause considerable morbidity, mortality, and resource use. CRBSI surveillance serves quality improvement, but is often restricted to intensive care units (ICUs). We conducted a four-month prospective cohort study of all non-cuffed central venous catheters (CVCs) to design an efficient CRBSI surveillance and prevention programme. CVCs were assessed on a daily basis for ward exposure time, care parameters, and the occurrence of laboratory-confirmed CRBSI. Overall, 248 patients with 426 CVCs accounted for 3567 CVC-days (median: 5) and 15 CRBSI episodes. CVCs were inserted by anaesthetists, ICU physicians and internists in 45%, 47%, and 8% of cases, respectively. CVC utilisation rates for intensive care, internal medicine, non-abdominal surgery and abdominal surgery were 29.8, 3.8, 1.7 and 4.9 per 100 patient-days, respectively. Fourteen percent of patients changed wards while having a CVC in place, so spending CVC-days at risk within multiple departments. CRBSI incidence densities for ICU, internal medicine, surgery and abdominal surgery were 5.6, 1.9, 2.4 and 7.7 per 1000 CVC-days at risk, respectively. In a univariate Cox proportional hazards model, the high CRBSI rate in abdominal surgery was associated with longer CVC duration, frequent use of parenteral nutrition and CVC insertion by anaesthetists. CRBSI numbers were insufficient to perform a multivariate analysis. Our surveillance revealed similar CRBSI rates in both ICU and non-ICU departments, and when frequent ward transfers occurred. Hospital-wide CRBSI surveillance is advisable when a large proportion of CVC-days occur outside the ICU.