Molecular diagnostic of sepsis: episodes vs. cases
Abstract number: P1973
Regueiro B., Varela-Ledo E., Martinez-Lamas L., Perez del Molino M., Garea M., Alvarez J.
Objectives: To evaluate the interpretation of the data obtained after utilising a multiplex PCR assay that identifies relevant microorganisms in septic patients. Data analysis of episodes (samples taken at same time) vs. cases (serial samples in same patient) were compared.
Methods: Sensitivity, sensibility, positive and negative predictive values were determined in a group of 53 patients from our post surgical/ high-complexity ICU with systemic inflammatory response syndrome (SIRS) and high clinical risk factors for bloodstream infection. Agreement was measured by kappa test.
Confirmation by microbiological identification of a pathogen in culture of blood or other locations (conventional microbiology), was used as a standard for determining if patients were infected or non-infected. The evaluation was retrospective and observational. In all cases the patients were covered by aggressive empiric therapy and gave their informed consent.
PCR results were obtained using multiplex Septifast test (ROCHE), blood cultures were done with BactAlert and conventional identification was performed with VITEK 2 (Biomerieux).
Results: 366 Septifast determinations were obtained for the 53 patients included in the study. We paired blood cultures and Septifast determinations in 294 episodes, conventional microbiology and Septifast in 302 episodes, and conventional microbiology plus blood culture vs. Septifast in 294 episodes. All 53 patients were serially analysed on days 0, 1, 2, 3, 7, 14 and 28 after ICU admittance the results were analysed using Septifast as a standard for determining disease or non-disease status. These results were then matched to see if coincidence existed with the results from blood and peripheral compartments obtained consecutively over the time span of a week using conventional phenotypic methods. Results are in Figure 1.
Conclusion: Septifast test detects the special condition of DNAhemia. This condition is independent of the viability needed for microbial growth in cultures. The sensitivity (75%) and specificity (86.66%), positive (81.81%) and negative (81.25%) predictive values and kappa (0.621) all indicate that the molecular test results are substantially (0.81) related to patient infection by microorganisms (located in blood or other compartments). As these patients tend to be refractory to conventional bacteriological diagnostic approaches, molecular methods may eventually prove to be a valuable criteria for early adequate antimicrobial treatment.
|Session name:||18th European Congress of Clinical Microbiology and Infectious Diseases|
|Back to top|