The diagnosis of spinal infections (SI) (vertebral osteomyelitis, epidural abscess, spondylodiscitis) remains a challenge in low-income countries due to limited access to advanced microbiological techniques, prolonged culture times, and low diagnostic yields, particularly for fastidious and antibiotic-exposed bacteria. Metagenomic next-generation sequencing (mNGS) offers a culture-independent approach with high sensitivity, but its clinical utility in resource-constrained settings is limited by cost and infrastructure challenges. We conducted a systematic review and meta-analysis of studies assessing the diagnostic performance of shotgun or targeted mNGS for bacterial detection in SIs. A comprehensive literature search of Embase, Medline, and Web of Science was performed till January 15, 2025. Studies including tubercular or fungal infections were excluded. The reference standard comprised composite clinical criteria and culture-based methods. Risk of bias was evaluated using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool. A total of 15 studies with 944 patients were included. In comparison to the culture mNGS demonstrated a pooled sensitivity and specificity of 91% (95% CI: 87.1%–94%) and 35.0% (95% CI: 31.1%–38.9%) respectively. In comparison with the combined reference standard, the pooled sensitivity and specificity of mNGS was 86.0% (95% CI: 81.6%–90%) and 71.4% (95% CI: 63.1%–78.8%) respectively. The area under the summary receiver operating characteristic curve was 0.90, indicating moderate-to-high diagnostic performance. mNGS improves pathogen detection in SI. To enhance accessibility, future strategies should include the development of targeted NGS panels for common spinal pathogens, the establishment of centralized sequencing hubs for cost-sharing, integration with affordable molecular diagnostics, and public-private partnerships to subsidize sequencing costs. A multi-tiered diagnostic approach combining conventional microbiology, rapid molecular tests, and selective mNGS may provide a feasible strategy to enhance SI diagnostics globally.
Choudhary et al. (Fri,) studied this question.