No gold standard assay exhibiting error-free classification of results has been identified for detection of Renibacterium salmoninarum, the causative agent of salmonid bacterial kidney disease. Validation of diagnostic assays for R. salmoninarum has been hindered by its unique characteristics and biology, and difficulties in locating suitable populations of reference test animals. Infection status of fish in test populations is often unknown, and it is commonly assumed that the assay yielding the most positive results has the highest diagnostic accuracy, without consideration of misclassification of results. In this research, quantification of R. salmoninarum in samples by bacteriological culture provided a standardized measure of viable bacteria to evaluate analytical performance characteristics (sensitivity, specificity and repeatability) of non-culture assays in three matrices (phosphate-buffered saline, ovarian fluid and kidney tissue). Non-culture assays included polyclonal enzyme-linked immunosorbent assay (ELISA), direct smear fluorescent antibody technique (FAT), membrane-filtration FAT, nested polymerase chain reaction (nested PCR) and three real-time quantitative PCR assays. Injection challenge of specific pathogen-free Chinook salmon, Oncorhynchus tshawytscha (Walbaum), with R. salmoninarum was used to estimate diagnostic sensitivity and specificity. Results did not identify a single assay demonstrating the highest analytical and diagnostic performance characteristics, but revealed strengths and weaknesses of each test.
