DISCUSSION The incidence of CR‐BSI in the study cohort was comparable with or lower than that found in other recent multi‐centre studies,2‐6 but higher than the infection rate attained in some intervention studies.25,28 However, the different infection criteria used (catheter‐ associated bloodstream infection or CR‐BSI or variations thereupon) also affect these rates. The effects of antibiotics, TPN and insertion vein in the surveillance cohort have frequently been shown to be associated with CR‐BSI.29 L’Hériteau2 et al also found a higher risk of CR‐BSI associated with insertion in the jugular vein, and a lower risk when the CVC was used for administration of antibiotics. Selective digestive tract decontamination with four days of intravenous cefotaxime, as practiced in some Dutch hospitals in order to reduce mortality, was also associated with a reduced rate of ICU‐ acquired bacteraemia.30 There could be several reasons for the negative association between antibiotics and CR‐BSI, apart from their antimicrobial effect. CR‐BSI may be underdiagnosed in patients on antibiotics because clinicians may apply a higher threshold before taking blood cultures. This is less likely to account for the association in the present data, because it was possible to diagnose an infection without blood cultures. It is not possible to assess the extent to which false‐negative blood cultures for patients on antibiotic therapy played a role. Similar to the present results, Wylie et al and Tacconelli et al found an increased risk for CR‐BSI associated with TPN.31,32 However, L’Hériteau et al did not find an increased risk for CR‐BSI associated with CVC insertion through the femoral vein although haemodialysis catheters were excluded in that study.2 In the present cohort, CVC insertion in the femoral vein was strongly associated with dialysis (71% of the CVCs used for dialysis were inserted in a femoral vein), and therefore, although not significant in the multi‐variate analysis, an independent deleterious effect of dialysis cannot be excluded. In a randomized trial with dialysis catheters alone, no difference was found in the infection rates for femoral and jugular access.33 An increased APACHE II score was not significantly associated with increased risk of CR‐BSI in the present study, but patients in the ICU with a CVC inserted had a higher risk of CR‐BSI than patients in other departments, and prolonged ICU stay prior to CVC insertion increased the risk. APACHE II scores are determined during the first 24 h of ICU admission, whereas the length of stay in the ICU prior to CVC insertion is associated with recovery or deterioration of the patient while at the ICU. However, when adjusted for APACHE II score, the effect of a prolonged ICU stay was reduced, as higher APACHE II scores tended to be associated with higher risk of infection (data not shown). 3 51 Catheter application, vein and pre-insertion ICU stay affect CRBSI risk
RkJQdWJsaXNoZXIy MTk4NDMw