Each year an estimated quarter of a million central venous catheters (CVC) are inserted in England and 5 million in the USA [1, 2]. CVCs are important for monitoring, drug delivery, intravenous feeding and blood sampling for patients receiving critical care. They are also a common cause of hospital-acquired bloodstream infection and hence a major cause of morbidity, mortality and increased healthcare costs. The National Audit Office in the UK suggested that approximately 44% of hospital-acquired bacteraemias were associated with invasive devices, with nearly two-thirds of these related to CVCs [3]. Both adult and paediatric intensive care units (PICUs) have among the highest rates of hospital-acquired bacteraemia of all specialities, with each infection costing an estimated additional US $40,000–46,133 [4]. As many of these infections are preventable, catheter acquired bloodstream infection (CA-BSI) has become a major issue for assessing the quality of care in intensive care units on both sides of the Atlantic [5–7].

Recent quality improvement initiatives have dramatically reduced rates of CA-BSI. Bundles of care, involving multifaceted interventions to improve CVC care and monitor adherence to recommendations, have been successfully introduced across North America and the UK [5, 6, 8, 9]. CVC bundles address two main mechanisms of infection. Firstly, they reduce contamination from skin commensals during CVC insertion. Bundles typically recommend strict hand-washing, cleaning the skin with 2% chlorhexidine and allowing 30 s for drying, maximal barrier methods, and avoidance of the femoral vein. Secondly, they reduce accumulation of organisms in the CVC lumen, by reducing contamination during sampling or infusion. CVC maintenance bundles include recommendations for minimising the duration of CVC insertion, improving catheter site care, and cleaning the hub with 2% chlorhexidine when sampling and changing infusion sets [5–7]. Use of CVCs impregnated with antibiotics or heparin is another way of reducing bacterial adherence to the CVC lumen, but these are not recommended in the USA or UK unless infection rates remain high despite implementation of the CVC bundles [7, 10].

The success of CVC bundles has been measured by studies comparing CA-BSI rates before and after their introduction: there have been no randomised controlled trials of the effectiveness of CVC bundles. Pronovost et al. [11] showed a two-thirds reduction in CA-BSI rates, which, in 108 adult ICUs, was sustained for 36 months. The CVC bundle focussed on improved CVC insertion guidelines and reducing the duration of CVC insertion. Similarly large reductions have been seen in PICUs, but there is growing evidence that success depends on interventions that prevent accumulation of organisms after CVC insertion. Improvements in CVC insertion are not sufficient. In a before-after study in 29 PICUs, Miller [5] reported that maintenance bundle compliance was the only significant predictor of reduced CA-BSI rates. In another US study of a planned series of interventions, Bhutta et al. [9] saw no change after introducing maximal barrier methods for CVC insertion, but a step reduction in CA-BSI when antibiotic-impregnated CVCs were introduced. Further evidence for the importance of reducing risk factors throughout the duration of CVC in children, not just at insertion, is reported in this issue of Intensive Care Medicine by Rey et al. [12] They found that the CR-BSI rate fell from 12/1,000 CVC days to 3/1,000, after they halved the proportion of children receiving total parenteral nutrition and reduced the duration of CVC insertion by 20%.

There are several reasons why children may be more at risk than adults of accumulating organisms after CVC insertion, making bundles for maintaining CVC sterility relatively more important. First, the case mix is different with a higher number of children who are immunocompromised and have congenital abnormalities. Secondly, children have small veins, which are susceptible to thrombosis, thus increasing bacterial adhesion [13, 14]. Thirdly, although femoral venous catheterization is associated with a greater risk of infective and thrombotic complications [15, 16], the femoral vein is safer and more commonly used as a site for CVCs in children. Fourthly, CVCs may be accessed more frequently for blood sampling to avoid traumatising the child and to preserve peripheral veins. For the same reason, CVCs are more precious in children than in adults and less likely to be replaced unless blocked or clearly infected. Fifth, children are more likely than adults to receive total parenteral nutrition, which increases the risk of CA-BSI [12–14, 17].

This constellation of risk factors begs the question of whether the effectiveness of CVC maintenance bundles could be improved by including impregnated CVCs, especially for children who need a CVC for more than 2–3 days. This question is being addressed by a large randomised controlled trial involving 1,200 children in 11 of the 24 PICUs in the UK—the CATheter infections in CHildren (CATCH) trial, funded by the UK National Institute of Health Research. The primary aim is to compare the effect of impregnated with standard CVCs on any bloodstream infection, in the context of widespread implementation of CVC bundles. Both antibiotic-impregnated and heparin-bonded CVCs are being compared with standard CVCs as it remains unclear whether they are more effective than standard CVCs for all-cause BSI and which type is best. A systematic review of impregnated catheters reported significant and substantial reductions in CA-BSI with heparin-bonded and antibiotic-impregnated catheters, compared with standard CVCs [18], but no statistically significant benefit of antiseptic CVCs, coated with chlorhexidine and silver sulphadiazine (SS). However, none of the trials of antibiotic-impregnated CVCs were done in children. Two of the trials of heparin-bonded CVCs did involve children, but the study populations had unusually high infection rates, much higher than those currently seen with CVC bundles [19–21]. Head-to-head trials clearly showed that antibiotic-impregnated CVCs were more effective than chlorhexidine SS CVCs, but a head-to-head trial found that heparin-bonded CVCs were no more effective than chlorhexidine SS. There have been no head-to-head trials comparing antibiotic-impregnated with heparin-bonded CVCs.

The CATCH trial is funded by the UK National Institute for Health Research and is designed to inform policy decisions across the National Health Service about whether impregnated CVCs are effective and cost effective in the context of substantial improvements in CVC care. Antibiotic-impregnated and heparin-bonded CVCs are twice the cost of standard non-impregnated CVCs and there are concerns about serious adverse events, although heparin-induced thrombocytopenia has never been reported to the manufacturer with the use of heparin-bonded CVCs. Antibiotic resistance has been reported in vitro with antibiotic-impregnated CVCs and microbiologists are concerned about their widespread use. Moreover, current guidelines recommend their use only if CA-BSI rates remain high after implementation of CVC bundles [7, 10]. As a result, antibiotic-impregnated or heparin-bonded CVCs are used for only a small minority of children in PICU in the UK and USA (Miller and Mok, personal communication). On the other hand, we have learned that effective quality of care interventions need to be multifaceted [8]. The study by Rey et al. highlights one facet, the importance of total parenteral nutrition. We have yet to see whether the CATCH trial will add another—impregnated CVCs.