Clinical application value of metagenomic next-generation sequencing in children with fever of unknown origin.

PURPOSE: Infectious diseases constitute the predominant cause of fever of unknown origin (FUO). Conventional microbiological testing is limited by prolonged turnaround times, susceptibility to host/environmental interference, low detection sensitivity, and limited capacity to identify rare pathogens. Metagenomic next-generation sequencing (mNGS) enables parallel broad-spectrum screening for microbial agents. This study aimed to investigate the clinical utility of mNGS in children presenting with FUO, to generate descriptive observational data on pathogen detection and temporally associated anti-infective regimen adjustments. METHODS: This retrospective single-center analysis enrolled 41 hospitalized children diagnosed with FUO who underwent mNGS testing at the Department of Infectious Diseases, Affiliated Children's Hospital of Shandong University, from June 2022 to July 2025. Initially, all patients underwent comprehensive routine systemic evaluations. For cases where fever persisted despite conventional testing and an infectious etiology was highly suspected, or where there was a poor therapeutic response to empirical anti-infective treatments, mNGS was subsequently performed. All specimens submitted for testing were sterile body fluids. Each sample was divided into two aliquots: one was subjected to conventional microbiological testing (including culture, smear microscopy, and PCR), while the other was cryopreserved for mNGS analysis. The performance of pathogen detection was compared between mNGS and conventional testing modalities using paired specimen data. RESULTS: In this study, we analyzed 41 pediatric cases, which included three types of specimens: blood, cerebrospinal fluid (CSF), and tissue fluid (comprising deep pus, postoperative drainage fluid, subdural effusion, and aspirated fluid from the mass). mNGS identified 30 microbial isolates from 20 patients, which included bacteria, viruses, fungi, and mycoplasmas; of these, 17 isolates were ultimately confirmed as causative pathogens. No statistically significant differences in positivity rates were observed between mNGS and conventional assays, as indicated by paired 2 × 2 contingency tables (all P > 0.05).The present study also recorded changes to antimicrobial regimens that occurred after pathogen identification by mNGS testing, including adjuvant antiviral therapy for 4 patients, antimicrobial escalation for 6 patients, antimicrobial de-escalation for 2 patients, and comprehensive regimen modifications for an additional 5 patients. CONCLUSION: We analyzed a targeted pediatric FUO subgroup, and the overall pathogen detection positivity rate showed no statistical difference between mNGS and routine microbial testing. Accordingly, mNGS cannot currently replace standard workflows or routinely screen all FUO children. The two testing methods exhibited complementary pathogen detection spectra. mNGS may act as an auxiliary tool for complicated infectious cases with negative conventional test results. This study generates descriptive observational data on pathogen identification and temporally associated anti-infective regimen adjustments in a selected cohort of FUO children. Further prospective studies with larger sample sizes are required to validate these findings.