Comparison of real-time PCR and blood culture for the diagnosis of bloodstream infections in onco-haematologic patients: microbiological and clinical assessment
Abstract number: 1733_906
Varani S., Stanzani M., Nardi L., Paolucci M., Vianelli N., Baccarani M., Landini M., Dal Monte P., Cavrini F., Giannini B., Melchionda F., Pession A., Sambri V.
Objectives: Sepsis is a syndrome that results from a dysregulated host response to infection which can rapidly lead to organ dysfunction and death. The aim of this study was the microbiological and clinical assessment of a new real-time PCR test (SeptiFast: SF, Roche Diagnostics) for the diagnosis of blood stream infections (BSI) in a population of neutropenic patients. The results of SF were compared with blood cultures (BC) obtained on the same specimens and the clinical significance of the two methods was assessed.
Methods: This single-centre study was conducted from the beginning of May 2006 until the end of July 2006. Eighty-six samples were obtained simultaneously for culture and SF from peripheral blood of 51 onco-haematologic patients with clinical suspicion of bacterial or fungal BSI. All patients included in the study exhibited leukopenia (WBC < 200/mL). The SF was performed following manufacturer's instructions. BC was performed by using an automated system (BacTAlert, bioMérieux). Standard laboratory procedures were employed for identification and antimicrobial susceptibility testing of the bacterial isolates from BC.
Results: Fourteen and 57 samples were identified, respectively, as positive and negative by both the methods. 8 specimens were found positive by BC, but negative by SF and 7 samples were detected as negative by BC, but positive when tested by SF. The agreement between the two tests was 16.3% and 66.3% for the positive and the negative specimens, respectively. The overall agreement between the two methods was 82.6%. In 3 out of 6 cases that resulted positive for coagulase-negative staphylococci (CoNS) at BC, SF showed the presence of contamination by CoNS. A clinical surveillance of these cases suggested that the detection of CoNS by culture was a false positive result probably due to contamination in the collection procedure. A comparative identification of bacteria from BC and SF was performed and we found that SF identified Gram-negative bacteria more efficiently than BC. The average time to identification of different bacteria was 85.1±7.9 hours (mean±standard error mean) for BC versus 27.7±4.3 for SF (p < 0.000001).
Conclusion: The capability to identify ``contaminating'' low bacterial load, together with the fast availability of clinical relevant results suggest that SF should be performed together with standard BC in neutropenic febrile patients to improve the sensitivity and specificity of the laboratory diagnosis of BSI.
|Session name:||European Society of Clinical Microbiology and Infectious Diseases|
|Location:||ICC, Munich, Germany|
|Back to top|