Catheter-Specific Rates of Central Line–Associated Bloodstream Infections in Critically Ill Children

Abstract

BACKGROUND: Central line-associated bloodstream infections (CLA-BSI) are the most common healthcare-associated infection among hospitalized children. Active surveillance for CLA-BSI among critically ill children at our hospital has revealed a rate of infection that is above the median rate reported by NNIS for patients in pediatric intensive care units (PICUs). In response to these data, we conducted intensive education about the use of maximal sterile barriers, introduction of chlorhexidine/alcohol for skin preparation and care, and strategies to reduce line utilization. To date, we have not detected a significant change in our rate of CLA-BSI. OBJECTIVES: To understand how central venous catheter (CVC) type influences the risk of CLA-BSI in a population of critically children. METHODS: We perform unit-based surveillance for CLA-BSI in our 38-bed PICU. CDC definitions are used to determine CLA-BSI. For 12 months, a daily tally of the type of catheter used was recorded on every patient hospitalized in the PICU with a CVC. The total and catheter-specific rates of CLA-BSI were determined. RESULTS: During the study period, the median rate for total CLA-BSI was 12.6 (9.5–16.4, 95%CI) per 1000 catheter days; the mean CVC utilization ratio was 0.38 line days per patient day. Despite greater than 4200 line days, we did not measure a significant difference in catheter-specific rates of CLA-BSI: for patients with non-tunneled CVC the rate of CLA-BSI was 12.7 (8.6–17.9) as compared to 14.3 (7.8–23.9) per 1000 catheter days for patients with tunneled CVC. Among patients with peripheral intravenous central catheters (PICC), the rate of CLA-BSI was 10.2 (4.4–20.0) per 1000 catheter days. CONCLUSIONS: These findings suggest that there may be little difference in catheter-specific CLA-BSI rates among critically ill children. However, additional study with data from multiple institutions is needed to detect small differences in the catheter-specific rate of CLA-BSI in this patient population. In addition, strategies to reduce the rate of CLA-BSI among children with long-term CVC (PICC or tunneled catheters) may be needed. Strategies we are considering include 1) reducing the frequency of CVC access; 2) modifying the protocol for CVC hub preparation prior to access; and 3) establishing a dedicated catheter-care team.