Submitted to: Aquaculture Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/7/2010
Publication Date: 1/5/2011
Publication URL: http://handle.nal.usda.gov/10113/55430
Citation: Bebak, J.A., Shoemaker, C.A., Arias, C., Klesius, P.H. 2011. Assay performance during validation of freezing channel catfish Ictalurus punctatus (Rafinesque)infected with a Gram-negative bacterium. Aquaculture Research. 42:169-176. Interpretive Summary: Population studies estimating prevalence, incidence, or other epidemiological measures generally require sampling of large numbers of animals to obtain accurate estimates. If the animals being sampled are catfish or other warm-water pond-cultured fish species, then sampling will most likely occur in an environment with hot average daytime temperatures, collecting fish will be a slow process and the distance from any laboratory facilities precludes bringing back large numbers to the laboratory for analysis while fresh. However, one solution to this problem would be to place fish on dry ice as they were sampled, and transport them back to the laboratory for storage at -80ºC and processing at a later date. The objective of this study was to quantify the loss of infectivity when channel catfish (Ictalurus punctatus) were injected with various concentrations of Edwardsiella ictaluri, sampled, frozen at -80ºC and then sampled 59-63 d later. As bacterial titer (cfu E. ictaluri/gm kidney tissue) decreased, the ability of bacteriological and PCR assays to detect positive fish also decreased, especially when there were less than 10,000 cfu/gm tissue. At these lower titer concentrations, fish were much more likely to be misclassified as uninfected when they were, in fact, infected. Culture of homogenized kidney tissue was least likely to misclassify the infection status at these lower titer concentrations. These results indicate that sample size estimates should be adjusted if fish will be sampled while fresh or after storage at -80ºC. This paper suggests values for diagnostic sensitivity and diagnostic specificity that can be used when making adjustments to sample size estimations.
Technical Abstract: Recovery of bacteria from infected fish during population sampling can be affected by factors including type of assay, method of specimen preservation, and concentration of bacteria present. Consequently, before use in field sampling, methods should be validated. The three objectives of this study were, first, to determine whether a channel catfish Ictalurus punctatus (Rafinesque) fingerling classified as positive for Gram-negative Edwardsiella ictaluri infection according to bacterial culture before freezing was also classified as positive after freezing, second, determine how direct culture from the kidney (DIRECT), culture of homogenate (HOMOG) and standard PCR (PCR) agree with bacterial culture in terms of classifying fish as positive or negative, and third, to estimate diagnostic sensitivity and diagnostic specificity for DIRECT, HOMOG, and PCR. In fresh and frozen fish, as bacterial concentration decreased, the ability of each assay to detect positive fish also decreased, especially when there were 10,000 colony forming units per gram (CFU·gm-1) tissue. HOMOG proved to be the most reliable at correctly classifying catfish, whether they were subclinically or clinically infected. Overall values for this study population for diagnostic sensitivity were 0.66, 0.92, and 0.43, and for diagnostic specificity were 0.86, 0.91, and 0.95, for DIRECT, HOMOG, and PCR, respectively.