Performance of Sequencing Assays in Diagnosis of Prosthetic Joint Infection: A Systematic Review and Meta-Analysis

Abstract

BackgroundA prompt, accurate diagnosis of prosthetic joint infection (PJI) allows early treatment, and with identification of the causative organism, sensitive antibiotics could be applied. However, routine methods cannot identify the causative organism under certain circumstances. Gene sequencing assays have unique superiority in promptness and broad coverage of pathogens, but evidence of its accuracy is quite limited. MethodsOf 247 citations identified for screening, 12 studies with 1965 patients in total were included. The diagnostic value of sequencing assays in PJI was systematically reviewed. Subgroup analysis was conducted to explore the source of heterogeneity. ResultsPooled sensitivity was 0.81 (95% confidence interval [CI], 0.73-0.87); pooled specificity was 0.94 (95% CI, 0.91-0.97); positive likelihood ratio was 14.2 (95% CI, 8.7-23.4); negative likelihood ratio was 0.20 (95% CI, 0.14-0.29); and the area under the curve was 0.94 (95% CI, 0.18-1.00). The results of subgroup analysis revealed that antibiotics reduced the sensitivity of sequencing-based diagnosis compared with withholding antibiotics before sampling (0.71 vs 0.94). In another subgroup analysis, sequencing by synthesis (Illumina sequencing) had better specificity than other next-generation sequencing methods (0.963 vs 0.829) and specificity similar to time-consuming and laborious Sanger sequencing (0.963 vs 0.967). ConclusionSequencing assays had favorable diagnostic accuracy of PJI. When sequencing assays were applied to diagnosing PJI, an antibiotic-free interval before sampling may enhance the ability to detect the causative organism and, among next-generation sequencing methods, sequencing by synthesis seemed to have advantages over other methods in specificity.