[Assessment of diagnostic methods for the catheter-related bloodstream infections in intensive care units].

The majority of catheter-related bloodstream infections (CR-BSI) are associated with central venous catheters (CVCs) and most of them develop in patients staying at intensive care units (ICUs). The aim of this study was to assess the performance of different methods for the diagnosis of CR-BSI in neurology and neurosurgery ICUs of our hospital. This prospective study was carried out between January 2007 and January 2008 and all of the patients were followed daily for CR-BSI after the insertion of CVCs. Blood cultures were taken simultaneously from the catheter lumen and from at least one peripheral vein when there was a suspicion of CR-BSI. Additionally, from patients whose CVCs were removed, catheter tip cultures were taken and from patients with exit site infection, cultures of the skin surrounding the catheter entrance were taken. Catheter tip cultures were done by using quantitative and semiquantitative culture methods. Blood cultures taken from the catheter lumen and peripheral vein were incubated in the BACTEC 9050 (Becton Dickinson, USA) automated blood culture system. Gram and acridine orange (AO) staining were used for the smears prepared from the catheter tips and blood cultures. To evaluate the value of culture and staining methods in the diagnosis of CR-BSI; sensitivity, specificity, positive and negative predictive values (PPV and NPV, respectively) of each method were determined. A total of 148 patients (66 male, 82 female; age range: 1-94 years, mean age: 58.7 ± 21.8 years) were included in the study, of whom 67 (45.3%) were from neurology and 81 (54.7%) were from neurosurgery ICUs. One hundred ninety-nine CVC application performed in 148 patients were evaluated. Mean duration of catheterization was 8.5 ± 5.2 days. Thirty-two episodes of CR-BSI among 199 catheterizations (16%) in 29 patients among a total of 148 patients (19.6%) were determined. The most frequently isolated microorganisms were methicillin-resistant coagulase-negative staphylococci (8/32; 25%), penicillin-resistant Enterococcus spp. (8/32; 25%) and Candida albicans (4/32; 12.5%). Sensitivity, specificity, PPV and NPVs of the quantitative and semiquantitative culture methods of the catheter tip and the differential time to positivity (positive result obtained at least two hours earlier in blood cultures drawn through the catheter than the peripheral blood cultures which were taken simultaneously) between blood cultures drawn through the catheter and those drawn from the peripheral vein were 100% for the diagnosis of CR-BSI. Sensitivity and NPV of the isolation method of the same microorganism from blood culture drawn through the catheter and drawn from the peripheral vein were 100%, specificity was 85% and PPV was 88% for the diagnosis of CR-BSI. Sensitivity, specificity, PPV and NPVs of Gram and drawn simultaneously from the peripheral vein and quantitative and semiquantitative cultures of the catheter tip in patients with removed catheter, were important factors in terms of diagnosis of CR-BSI. It was also concluded that AO staining could provide additional benefit in the diagnosis of CR-BSI since it has higher sensitivity, specificity, PPV and NPVs for peripheral blood cultures and catheter tip cultures compared to Gram staining.