Traitement par verrou local d'antibiotique des infections bactériennes liées aux cathéters centraux en nutrition parentérale

[Infections of central venous catheters in home parenteral nutrition: a retrospective monocentric study over 2 years]

BACKGROUND AND STUDY AIMS

Infection of the central venous catheter (CVC) is the main aggravating factor of parenteral nutrition. The aim of this study was to determine the ecology of these infections in our home parenteral nutrition center and to evaluate our care protocol.

PATIENTS AND METHODS

The present study was monocentric and retrospective, and was conducted in the parenteral nutrition service of the Marseille University Hospital between 1 January 2011 and 31 May 2012. During this period, all the children who presented fever and a positive hemoculture in a medical emergency were taken into account, and the characteristics of the infection were analyzed.

RESULTS

After 17 months, 17 children had been subject to an infection in their central catheter: 47 bacteremia were identified, which is equivalent to 5.4 infections for 1000 days of CVC. An average 2.8 hemocultures were performed during this time. The most common bacterium was Staphylococcus hominis. The children affected by a digestive stoma or by a gastrostomy were not subject to more infections and their ecology was not different. All the children were referred to the hospital and treated with an intravenous antibiotic through the CVC.

DISCUSSION

As expected during the time of the study, the infection rate was very high. This finding led us to reassess our prevention protocol. It is also very likely that the number of infections was overestimated due to the protocol for sampling in the hemocultures and their subsequent analysis. Lastly, some children relapsed easily, although no predisposing factor was found in the present study. The therapy chosen was in agreement with the best practices and the ecology recovered.

CONCLUSION

The high number of infections observed during the study encourages a prospective evaluation of current practices.

Catheter-related sepsis during home parenteral nutrition

Catheter-related sepsis (CRS) in patients receiving home parenteral nutrition (HPN) is the most frequent complication, with an annual incidence rate of 0.30-0.50 in investigative/approved centres. CRS is responsible for patient deaths in 2.5% of cases which represent 50% of HPN-related deaths. A detailed and strictly applied protocol is the basis of prevention; therefore a low rate of CRS can be used as a marker of quality of care. Skin and hub cultures are used to rule out CRS since their predictive negative value assessed against (semi) quantitative cultures of cannulae were reported to be almost 100%. In the absence of catheter removal, clinical assessment plus quantitative blood cultures with a threshold ratio of 4:1 in central and peripheral blood is a reasonable basis of established CRS. Removal of the infected external cannula is necessary and immediate in cases of tunnel infection, virulent bacteria (e.g. Candida, Staph aureus or pseudomonas species) and for all occurrences of CRS on implantable chambers since conventional treatment fails to sterilise such devices. Thus standard antibiotic treatment may be effective in the absence of tunnel sepsis in the majority of CRS cases due to non-virulent bacteria (mainly Staph coagulase negative species). In these cases of CRS, our experience suggests that the antibiotic-lock technique offers definite advantages over conventional systemic antibiotic treatment, since external catheters were sterilised without recurrent sepsis in more than 85% of cases.