Assessing The Diagnostic Accuracy Of Immunodiagnostic Techniques In The Diagnosis Of Schistosomiasis Japonica: A Meta-analysis

Objective:Immunodiagnostic tests are commonly used in the diagnosis for schistosomiasis japonica in field situation. Their diagnostic effect, however, has never been evaluated in a systematic way. We set out to review the value of tests including enzyme-linked immunosorbent assay (ELISA) and indirect hemagglutination assay (IHA) in the diagnosis of schistosomiasis japonica. An integrated meta-analysis of the performance of the two tests in the diagnosis of schistosomiasis japonica was conducted.Methods:We undertook a meta-analysis of diagnostic ability that compared ELISA with IHA for the diagnosis of schistosomiasis japonica. Original articles were searched in Pubmed, Science Citation Index Expanded (SCI), Cochrane Library, Proquest and China National Knowledge Infrastructure (CNKI). Data extractions were conducted independently by 2 reviewers. Statistical analysis was performed using SAS statistical software, version 9.1. Summary receiver operating characteristic curves were used to summarize overall test performance. The bivariate model also used to analyze the diagnostic accuracy. Funnel plot was used to analyze publication bias. Excel 2003 was used to plot SROC curve and make tables. Results:Sixteen studies met inclusion criteria. In all studies combined, the diagnostic odds ratio (DOR) for IHA was 8.69 (95%CI: 7.67 to 9.84) and ELISA was 3.69 (95%CI: 3.02 to 4.52). The DOR for IHA was better than that for ELISA. IHA demonstrated TPR* =0.72 and ELISA showed TPR* =0.63. Although the sensitivity of IHA is higher than that of ELISA, the results of the Z test can not prove that a significant difference exists between the two methods in screening for schistosomiasis japonica(P>0.05). Results of the bivariate model show that the mean sensitivity of IHA is 0.81(95%CI:0.69-0.89), and the mean specificity of IHA is 0.67(95%CI:0.55-0.78); the mean sensitivity of ELISA is 0.82(95%CI:0.66-0.92), and the mean specificity of ELISA is 0.48(95%CI:0.33-0.63); the mean DOR of IHA is 9(95%CI:4-21), and the mean DOR of ELISA is 4(95%CI:3-7). IHA has higher specificity than ELISA.The sensitivity between IHA and ELISA is similar. There is no statistically significant difference in the mean value of sensitivity or specificity between IHA and ELISA. The DOR of IHA is not significantly different from ELISA.Conclusions:We conducted a meta analysis though SROC and bivariate model. We found the bivariate model can be seen as an improvement and extension of the traditional SROC approach for it can directly provides summary estimates of (logit) sensitivity and spectivity with corresponding 95% CI for diagnostic tests. Our study shows that both IHA and ELISA have high sensitivity. The diagnostic odds ratio also lies to a high level. Between the two diagnostic tests, the performance of IHA is better than that of ELISA. Nevertheless, difference of the accuracy between the two methods for diagnosis of schistosomiasis japonicum has no statistical significance showed by the result of Z test. We found that both IHA and ELISA can be used to rule in or rule out the diagnosis in field situation. Also, we draw a lesson that when selecting an appropriate test, the local availability must be considered.