BackgroundBlood cultures are commonly employed to identify bacterial pathogens causing sepsis. PCR assays to diagnose septicemia require extraction of bacterial DNA from blood samples and thus, delay the initiation of appropriate antimicrobial treatment. The presence of abundant human DNA may hamper the sensitivity of PCR in the detection of bacteria.MethodsWe used serial dilutions of E. Coli spiked pseudo-blood-sepsis samples to develop a simple method that combines the use of a polar detergent solvent and adjustment of the basic pH to remove human DNA. A 16S rRNA gene-based screening algorithm was established to differentiate Gram-positive and Gram-negative groups of bacteria and the family of Enterobacteriaceae. A stringent validation with appropriate controls was implemented. The method of human DNA removal was then applied on 194 sepsis blood samples and 44 cerebrospinal fluid (CSF) samples by real-time PCR.ResultsThis uncomplicated and straightforward approach allows to remove up to 98 % of human DNA from peripheral blood of septic patients. The inhibitory effect of human DNA is efficiently prevented and the detection limit of real-time PCR is increased to 10 E. Coli CFUs/ml. This sensitivity is 10 times higher compared to conventional real-time PCR assays. The classical blood culture detected 58/194 (30 %) of sepsis and 9/44 (21 %) of CSF samples. Out of the 194 blood samples tested, the conventional real-time PCR targeting 13 common sepsis causing pathogens correctly detected the bacterial DNA in 16/194 (8 %) only and 14/44 (32 %) in cerebrospinal fluid samples. Our newly established approach was able to provide correct diagnoses in 78 (40 %) of the 194 blood samples and in 14 (32 %) of the CSF samples. The combination of both blood cultures and our technique raised the rate of sepsis diagnoses to 112/194 (58 %). Of the total group tested positive, 46 (24 %) cases showed overlap with the classical methodology.ConclusionWe report a simple optimized in-house protocol for removal of human DNA from blood sepsis samples as a pre-analytical tool to prepare DNA for subsequent PCR assays. With the detection increase of our in-house DNA removal approach, subsequent PCR assays can reach detection limits of 10 E. coli CFUs/ml and significantly improve the diagnostic rate in blood sepsis cases.Electronic supplementary materialThe online version of this article (doi:10.1186/s12879-016-1568-1) contains supplementary material, which is available to authorized users.

Surgical site infection (SSI) is common in Vietnamese post-operative patients. It contributes to increased morbidity, mortality, hospitalization time and health care expenditure. Bacterial culture is considered the gold standard procedure to identify SSI pathogens and antibiotic resistant properties; however, it can detect microbes that can readily grow and is time-consuming. We propose optimized multiplex PCR assays to diagnose the most relevant microbes and associated genes encoding for acquired extended spectrum betalactamases (ESBL) or carbapenemases from Vietnamese patients with SSI in a hospital setting in Hanoi.MethodsNinety-one patients (n = 91) were collected in order to identify microbial pathogens and associated genes encoding for acquired extended spectrum betalactamases (ESBL) or carbapenemases by both conventional bacterial culture and in-house multiplex PCR assays.Result and conclusionThe novel in-house multiplex PCR assays are comparable to the bacterial culture approach in screening for common pathogens causing SSI and for relevant genotypes conferring betalactam/carbapenem resistance for bacteria. This is the first report of Turkey-specific ESBL gene (PER-1) and two Oxacilinase families (Oxa23 and Oxa 58) in Vietnam.

BackgroundClinical progression of HBV-related liver diseases is largely associated with the activity of HBV-specific T cells. Soluble fibrinogen-like protein 2 (sFGL2), mainly secreted by T cells, is an important effector molecule of the immune system.MethodssFGL2 levels were determined by ELISA assays in sera of 296 HBV patients clinically classified into the subgroups of acute hepatitis B (AHB), chronic hepatitis B (CHB), liver cirrhosis (LC), hepatocellular carcinoma (HCC) and patients with LC plus HCC. As control group, 158 healthy individuals were included. FGL2 mRNA was quantified by qRT-PCR in 32 pairs of tumor and adjacent non-tumor liver tissues.ResultssFGL2 levels were elevated in HBV patients compared to healthy controls (P <  0.0001). In the patient group, sFGL2 levels were increased in AHB compared to CHB patients (P = 0.017). sFGL2 levels were higher in LC patients compared to those without LC (P = 0.006) and were increased according to the development of cirrhosis as staged by Child-Pugh scores (P = 0.024). Similarly, HCC patients had increased sFGL2 levels compared to CHB patients (P = 0.033) and FGL2 mRNA was up-regulated in tumor tissues compared to adjacent non-tumor tissues (P = 0.043). In addition, sFGL2 levels were positively correlated with HBV-DNA loads and AST (Spearman’s rho = 0.21, 0.25 and P = 0.006, 0.023, respectively), but reversely correlated with platelet counts and albumin levels (Spearman’s rho = − 0.27, − 0.24 and P = 0.014, 0.033, respectively).ConclusionssFGL2 levels are induced by HBV infection and correlated with the progression and clinical outcome of HBV-related liver diseases. Thus, sFGL2 may serve as a potential indicator for HBV-related liver diseases.