Abstract

The aim of this study is to investigate the burden of basic diseases, the epidemiology of pathogenic microbes in patients with bloodstream infections and their clinical biomarkers. Data on bloodstream infections from 2012 to 2018 were obtained from a 3A General Hospital in Lanzhou of China. 1990 patients with bloodstream infections participated in the study; their basic diseases, pathogenic microorganisms and corresponding clinical laboratory biomarkers such as procalcitonin (PCT), C-reactive protein (CRP), interleukin-6 (IL-6), neutrophil% (NEU%), blood lactate (LAC) and D-Dimer were recorded. 722 cases with non-bloodstream infections were selected randomly as control group. 8.9% of the patients had positive blood culture. The contamination rate was 0.24%. 1043 strains were isolated belonging to 75 species; they contained Gram negative bacilli, Gram-positive cocci, anaerobic and fungi accounting for 62.00, 31.64, 2.40 and 2.01%, respectively. Statistical analysis showed that hematopathy was the most frequent disease caused by bloodstream infections (number of episodes per years: 56 overall) and it accounted for 16.88%; this is followed by hepatobiliary diseases, central nervous system diseases, urinary disease, respiratory diseases and digestive tract disease which accounted for 12.31, 11.61, 10.75, 6.35 and 9.95%, respectively. Receivers operators characteristics (ROC) curve was used to evaluate the biomarker for diagnostic value of bloodstream infections, and the AUC values of NEU%, PCT, IL-6, CRP, D-Dimer and LAC were 0.974, 1.000, 1.000, 0.438, 0.932 and 0.784, respectively. This study documents a very timely picture of the strains of the bloodstream infections epidemiology in a 3A General Hospital over a 6-year period, the basic diseases with bloodstream infections in patients, and the biomarker diagnostic value for BSIs. PCT and IL-6 can be used as rapid and accurate biomarkers for bloodstream infections caused by bacteria.

Key words: Bloodstream infections, pathogenic microorganisms, basic diseases, biomarkers, receivers operators characteristics (ROC) curves.