[Comparison of the Multiplex Real Time Polymerase Chain Reaction Method Directly from Whole Blood and from Blood Culture Bottle with the Conventional Method for the Etiological Diagnosis in Patients with Pre-Diagnosis of Sepsis].

In this study, it was aimed to evaluate the use of multiplex real time polymerase chain reaction (RtPCR) method, directly from whole blood and blood culture bottles of intensive care unit patients with a pre-diagnosis of sepsis for the determination of the causative agent by comparing it with the conventional method. The study was carried out prospectively between November 2019 and August 2020. Ninety patients hospitalized with a pre-diagnosis of sepsis were included in the study. Samples were collected simultaneously with aseptic technique as whole blood and at least two sets of blood cultures. Blood culture bottles were monitored by the BACT/ALERT 3D (bioMerieux, France) instrument. Blood culture bottles were studied both by Sepsis Pathogens Panel Kit v1 (Anatolia, Turkey) using the multiplex Rt-PCR and with the conventional culture methods. The duration of detection and reporting of the multiplex Rt-PCR methods were compared with the conventional method. In this study, a total of 90 whole blood samples and 280 blood culture bottles were collected from the patients. It was found that 73 (81%) patients had already been administered antibiotic treatment at the time of blood sampling. Conventional blood culture was accepted as the gold standard in the diagnosis of sepsis. Rt-PCR performed from blood culture bottles was found to have a sensitivity of 89.58%; specificity of 57.14%; positive predictive value of 70.49%; negative predictive value of 82.76% and accuracy of 74.44% (p= 0.019). On the basis of the bacterial species, the sensitivity of the Rt-PCR method was determined to be between 66.7-100% and the specificity was between 86.6-100%. The Rt-PCR performed from whole blood, was found to have a sensitivity of 11.67% and a specificity of 96.67%; positive predictive value of 87.50%; negative predictive value of 35.37% and accuracy of 40% (p= 1.684). In this study, the duration of reporting of blood culture results was in average 116:50 hours, that of PCR from blood culture bottles was 80:57 hours, and that of PCR from whole blood samples was 15:38 hours (p<0.001). It was determined that the PCR from whole blood shortened the reporting time by an average of 101:12 hours (approximately four days), and that the PCR from blood culture bottles shortened the reporting time by an average of 35:53 hours (1.5 days) compared to the conventional method. In this study, the results determined by the Sepsis Pathogens Panel Kit v1 (Anatolia, Turkey) performed from blood culture bottles were superior compared to the conventional method. It has been determined that improvement is required for the usage of the sepsis kit directly from whole blood. On the other hand, considering the species detected by the blood culture method and not by the assay, the species range detected by the multiplex assay panel needs to be improved.