Sensitivity and specificity of current diagnostic tests for gill-associated virus in Penaeus mondon

Abstract

This study reports the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy between a reverse transcriptase-nested polymerase chain reaction (RT-nPCR) and an enzyme-linked immunosorbent assay (ELISA) for the detection of gill-associated virus (GAV) from a sample of 120 Penaeus monodon. Subsequently, the same comparisons were applied to the ELISA and haemagglutination (HA) assays for detection of GAV from a second 120 prawns. The optical density (OD) or dilution cut-off point had a direct influence on the tested parameters. The cut-off OD of 0.5–0.6 with the ELISA produced a sensitivity of 98% compared with RT-nPCR. However, these OD produced the lowest accuracy (85.8% and 86.7%, respectively). The OD cut off of 0.75 resulted in the highest accuracy (91.7%) and NPV (81.3%) while it had the second highest sensitivity (97%) and PPV (93.3%). However, the OD cut off of 0.9 had the highest specificity (80%). With regards to HA, the titre cut off at 8 resulted in the highest sensitivity, specificity and NPV (94%, 100% and 100%, respectively) compared with the ELISA, while the HA titre of 16 gave the highest accuracy (73%) and the second highest specificity (75%). A HA titre of 64 gave the highest PPV (81%). Using the RT-nPCR as the gold standard, the ELISA had an accuracy of 91.7% when using a cut off >0.75 as a positive result. When compared with the ELISA, the HA had an accuracy of 73% when using an HA titre cut off greater than 16 as a positive result. These results indicate that alternative tests for GAV (ELISA and HA) can be used to explore multiple questions about the disease status of P. monodon stocks in a cost-effective manner.