Diagnostic test accuracy for detecting Schistosoma japonicum and S. mekongi in humans: A systematic review and meta-analysis

BACKGROUND: Most of national schistosomiasis elimination programmes in Asia are relying on stool examination, particularly Kato Katz stool examination technique for regular transmission monitoring. However, the Kato-Katz technique has shown low sensitivity for the detection of light-intensity infections, and therefore highly sensitive diagnostic tools are urgently required to monitor prevalence of infection in low transmission settings. The objective of this systematic review was to evaluate and synthesize the performance of diagnostic tests for detecting Schistosoma japonicum and S. mekongi infection in people living in endemic areas. METHODOLOGY/PRINCIPAL FINDINGS: We comprehensively searched these nine electronic databases and other resources until July 2019, with no language or publication limits: PubMed, EMBASE, MEDLINE, Web of Science, BIOSIS Citation Index, HTA, CINAHL PLUS, The Cochrane Library, and PsycINFO. We included original studies that assessed diagnostic performance using antibody, antigen, and molecular tests with stool examination test as a reference standard. Two reviewers independently extracted a standard set of data and assessed study quality. We estimated the pooled estimates of sensitivity and specificity for each index test. We used diagnostic odds ratio to determine the overall accuracy and hierarchical summary receiver operating characteristics (HSROC) curve to assess the index tests performance. Fifteen studies (S. japonicum [n = 13] and S. mekongi [n = 2]) testing 15,303 participants were included in the review. Five studies reported performance of enzyme-linked immunosorbent assay (ELISA), seven studies reported indirect hemagglutination assay (IHA), and four studies reported polymerase chain reaction (PCR) for detecting S. japonicum. The pooled sensitivity and specificity were 0.93 (95% CI: 0.84–0.98) and 0.40 (95% CI: 0.29–0.53) for ELISA, 0.97 (95% CI: 0.90–0.99) and 0.66 (95% CI: 0.58–0.73) for IHA, and 0.89 (95% CI: 0.71–0.96) and 0.49 (95% CI: 0.29–0.69) for PCR respectively. A global summary indicated the best performance for IHA, closely followed by ELISA. We were unable to perform meta-analysis for S. mekongi due to insufficient number of studies. CONCLUSIONS/SIGNIFICANCE: IHA showed the highest detection accuracy for S. japonicum. Further studies are needed to determine the suitable diagnostic methods to verify the absence of transmission of S. mekongi and also to compare detection accuracy against more sensitive reference standards such as PCR.