Diagnostic test accuracy when screening for Haliotid herpesvirus 1 (AbHV) in apparently healthy populations of Australian abalone Haliotis spp.

The accuracy of 3 real-time PCR assays (ORF49, ORF66 and ORF77) and histopathology was evaluated for the purpose of demonstrating or certifying abalone free from Haliotid herpesvirus 1 (AbHV), the causative agent of abalone viral ganglioneuritis. Analytically, all 3 qPCRs showed equivalent limit of detection (20 copies per reaction); however, ORF49 could not detect 2 of the AbHV genotypes. A selection of 1452 archive specimens sourced from apparently healthy abalone populations was screened using all 4 tests. In the absence of a perfect reference standard, a Bayesian latent class analysis was built to estimate diagnostic sensitivity (DSe), diagnostic specificity (DSp) and likelihood ratios of a positive (LR+) and negative test result (LR-) for each individual test and for all possible combinations of test pairs interpreted either in series or in parallel. The pair ORF49/ORF66 interpreted in parallel performed the best both analytically and diagnostically to demonstrate freedom from AbHV in an established population of abalone and to certify individual abalone free from AbHV for trade or movement purposes (DSe = 96.0%, 95% posterior credibility interval [PCI]: 82.6 to 99.9; DSp = 97.7%, 95% PCI: 96.4 to 99.4; LR+ = 41.4, 95% PCI: 27.4 to 148.7; LR- = 0.041, 95% PCI: 0.001 to 0.176). Histopathology showed very poor DSe (DSe = 6.3%, 95% PCI: 2.4 to 13.1) as expected since most infected abalone in the study were likely sub-clinical with limited pathological change. Nevertheless, we recommend histopathology when clinically investigating outbreaks to find potential, new, emerging AbHV genotype(s) that may not be detectable by either ORF49 or ORF66.