Performance evaluation of detecting adenovirus by using rapid diagnostic kits among Japanese people

This study evaluated the diagnostic performance of detecting adenovirus by rapid diagnostic kits among Japanese people. A meta-analysis was conducted to pool the sensitivity, specificity, positive and negative likelihood ratios, and diagnostic odds ratio of immunochromatography methods and latex agglutination tests; enzyme-linked immunosorbent assay tests had already been evaluated in another meta-analysis, and no other diagnostic kits have been appropriately studied. Immunochromatography methods were shown to have sufficiently high diagnostic power, regardless of whether conjunctiva or throat swabs, or stool specimens were sampled, based upon the area under the curve (0.961-0.991); this was significantly higher than enzyme-linked immunosorbent assay tests when sampling conjunctiva swabs (P = 0.00). Additionally, the relative diagnostic odds ratio showed a higher diagnostic power of immunochromatography methods than latex agglutination tests (P = 0.01) by sampling stool specimens. Furthermore, immunochromatography methods had a nearly confirmative level of diagnostic power when sampling throat swabs and stool specimens (positive likelihood ratio = 28.0 and 43.6, respectively, and negative likelihood ratio = 0.111 and 0.121, respectively), with enzyme-linked immunosorbent assay tests not sharing this characteristic when sampling either conjunctiva or throat swabs, or stool specimens. However, even regarding immunochromatography methods, the level of pooled sensitivity (0.667) when sampling conjunctiva swabs was not acceptable. It is apparent that, therefore, among the existing types of rapid detection methods, immunochromatography methods are overall preferable for diagnosing adenoviral infections, even though they still cannot completely rule out infections by negative results, especially when sampling conjunctiva swabs.