Necessity of establishing a background microbial library in different medical institutions for the diagnosis of periprosthetic infection using metagenomic next-generation sequencing: Data-Based Analysis

Abstract Background Due to the extremely high sensitivity of metagenomic next-generation sequencing (mNGS) testing, even a small amount of nucleic acid fragments exposed during sampling or testing may lead to false positives, which is one of the biggest challenges in interpreting mNGS testing reports. In this study, for the first time, we experimentally detected and established Periprosthetic joint infection (PJI)-related interfering nucleic acid background microbial libraries (BML) in different medical institutions to clarify Necessity of establishing a BML in different medical institutions for the diagnosis of periprosthetic infection using mNGS. MethodsSamples were taken from 3 different acetabular reamer for hip arthroplasty in 7 different hospitals. The whole process was strictly aseptic, mNGS was performed according to standard operating procedures. The sterility of instruments was confirmed by culture method. The sequencing results of specimens from different hospitals were compared to analyze the difference of background bacteria. Bioinformatics analysis and visualization were presented through R language. ResultsA total of 26 samples were processed by mNGS, including 24 instrument swab samples, 1 blank swab control, and 1 blank water control. 254,314,707 reads were sequenced in all samples. The results showed that 1.13% of Clean Reads can be matched to pathogenic microorganism genomes, of which bacterial sequences account for 87.48%, fungal sequences account for 11.18%, parasite sequences account for 1.26%, and virus sequences account for 0.06%. The results of PCA (Principal Component Analysis) demonstrated that the distribution of bacteria on the surface of instruments was significantly different between medical institutions. Through the Venn diagram, it was found that 465 species of bacteria in all region hospitals, Liaocheng People's Hospital had a maximum of 340 species of bacteria, followed by Guanxian County People's Hospital with 169 species. The clustering heat map illustrated that the distribution of bacterial groups in three different instrument samples in the same hospital was basically the same, and the bacterial genera varied significantly among hospitals. The residual microbial nucleic acid fragments are mainly bacterial DNA and represent differences in different medical institutions.ConclusionsIt is necessary to establish independent BML in different medical institutions to improve the accuracy of mNGS on the diagnosis of PJI.