Evaluation of the diagnostic effectiveness of next generation sequencing in sepsis etiology: a systematic review and meta-analysis

Introduction: Systematic evaluation of the diagnostic value of next generation sequencing (NGS) in sepsis etiology. Methodology: We conducted a systematic search on four databases (Web of Science, Cochrane, PubMed, and Embase) and compiled diagnostic experiments using NGS to evaluate sepsis etiology. Two researchers conducted research and obtained data independently. Results: Nine documents were included comprising 747 patients, 988 blood samples, 175 bronchoalveolar lavage fluid (BALF) samples, 16 cerebrospinal fluid samples, and one urine sample. The combined sensitivity of each study was 0.89 (95% CI: 0.82-0.95). The combined specificity was 0.40 (95% CI: 0.25-0.55). The combined positive likelihood ratio was 1.51 (95% CI: 1.18-1.98). The combined negative likelihood ratio was 0.28 (95% CI: 0.11-0.48). The diagnostic odds ratio (DOR) was 6.38 (95% CI: 2.53-15.32) and the area under the curve (AUC) was 0.84, (95% CI: 0.62-0.94). Conclusions: Based on the data we collected, we found that compared with the blood culture technology, NGS has the advantages of high sensitivity and wide detection range, but its specificity was low. Further study is needed to confirm the value of NGS in the etiological diagnosis of patients with sepsis.