Diagnostic Accuracy of Nucleic Acid Amplification Based Assays for Clostridium perfringens Associated Diseases: a Systematic Review and Meta-analysis.

Timely and accurate methods for detecting Clostridium perfringens associated diseases (CPAD) are crucial to improve patient care. A number of studies have evaluated the accuracy of nucleic acid amplification tests (NAAT) in detecting CPAD, but decisive results about its effectiveness have not been reported. We conducted a meta-analysis to evaluate the diagnostic performance of NAAT for detecting C. perfringens in clinical diarrheal samples. Five databases including PubMed, Embase, Scopus, Web of Science, and the Cochrane library were systematically probed up to December 06, 2019. From 2632 citations, we identified five eligible studies comprising 817 samples. Three studies (n = 695) compared NAAT with a microbiological culture while the other three studies (n = 322) compared it with an immunoassay. NAAT revealed higher diagnostic accuracy against immunoassay [sensitivity: 0.53 (0.35–0.7), specificity: 0.97 (0.95–0.99), positive likelihood ratio (PLR): 23.2 (3.49–153.98), negative likelihood ratio (NLR): 0.25 (0–245.28), diagnostic odds ratio (DOR): 74.11 (2.11–2593.7)] as compared to microbiological culture [sensitivity: 0.31 (0.22–0.41), specificity: 0.95 (0.93–0.97), PLR: 11.56 (3.87–34.6), NLR: 0.57 (0.27–1.21), DOR: 18.1 (4.83–67.8)]. NAAT pooled specificity was consistently ≥95% against applied reference standards. A meta-regression and sub-group analysis of sample condition, gene target, study design, and reference standards could not explain the heterogeneity (P >0.05) in the diagnostic efficiency. The analysis has demonstrated that the diagnostic accuracy of NAAT is relatively insufficient to replace traditional reference standards as a single diagnostic test. NAAT can be applied in combination with microbiological culture because of the advantage of time to result and in scenarios where traditional tests are not feasible. Further investigations in this direction with larger sample sizes are still warranted to support our findings.