Abstract
Spinal infections (SI) are on the rise due to an aging population and the prevalence of more invasive procedures. This study aims to systematically review the microbiological spectrum of SI to enhance diagnostic accuracy and inform effective antibiotic treatment strategies. The last search was conducted on May 9th, 2024, from databases including EMBASE, PubMed, and Web of Science. The outcome variable is infection rate, and the detection method used should be blood culture, tissue culture, or molecular biology method. Two researchers independently extracted research data and evaluated its quality using the JBI Critical Appraisal Tools. Out of 14,639 identified records, 156 studies (encompassing 13,539 patients) were included. Staphylococcus aureus was identified as the most prevalent pathogen, with pooled infection rates of 17.6% (95%CI: 12.8-22.9%; I(2)=93%) in blood culture, 16.8% (95%CI: 14.0-19.8%; I(2)=96%) in tissue culture, and 12.0% (95%CI: 9.3-15.0%; I(2)=35%) in molecular methods. The bacterial spectrum also featured Staphylococcus epidermidis, Escherichia coli, and Mycobacterium tuberculosis (MTB). Molecular methods, particularly metagenomic next-generation sequencing (mNGS), demonstrated markedly superior sensitivity for MTB detection, with a pooled rate of 9.7% (95%CI: 4.6-16.3%; I(2)=90%) compared to 1.3% (95%CI: 0.6-2.1%; I(2)=86%) by tissue culture. The odds ratio for MTB detection with mNGS versus conventional culture was 4.24 (95%CI: 1.68-10.73). This review confirms that a core group of pathogens, including Staphylococcus aureus, Staphylococcus epidermidis, MTB, and Escherichia coli. Our findings underscore that tissue culture is fundamental for common pyogenic bacteria, while metagenomic next-generation sequencing is indispensable for detecting fastidious organisms like MTB. Trial registration: The protocol was registered with PROSPERO (No. CRD42023427429). Registered on May 28, 2023.