Objective: To explore the potential of long noncoding RNA (lncRNA) as a diagnostic and prognostic biomarker for patients with sepsis.
Methods: Literature was obtained from seven databases. Relevant data were examined by combining sensitivity (SEN), specificity (SPE), positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic orders ratio (DOR), and the area under the curve (AUC) of the summary receiver operating characteristic (SROC) curve with their corresponding 95% confidence intervals (CI) using a bivariate model.
Results: Seventeen works of literature were included for meta-analysis. The meta-analysis determined that combined SEN was 0.76 (95% C.I.: 0.69-0.82), combined SPE was 0.84 (95% C.I.: 0.79-0.88), combined PLR was 4.81 (95% C.I.: 3.70-6.25), combined NLR was 0.28 (95% C.I.: 0.22-0.36), combined DOR was 17.02 (95% C.I.: 11.96-24.22), and AUC was 0.88 (95% C.I.: 0.84-0.90), however, for prognostic meta-analysis, combined SEN was 0.82 (95% C.I.: 0.73-0.86), combined SPE was 0.69 (95% C.I.: 0.54-0.81), combined PLR was 2.69 (95% C.I.: 1.82-3.97), combined NLR was 0.25 (95% C.I.: 0.10-0.35), combined DOR was 10.63 (95% C.I.: 7.13-15.87), and AUC was 0.84 (95% C.I.: 0.81-0.87).
Conclusion: LncRNA possesses a significant diagnostic and prognostic ability for sepsis.
Keywords: Long noncoding RNA; diagnosis; meta-analysis; prognosis; sepsis.