Interlaboratory and between-specimen comparisons of diagnostic tests for leptospirosis in sheep and cattle

Abstract

A study was performed to investigate interlaboratory test agreement between a research and a commercial veterinary diagnostic laboratory on blood and urine samples, and to investigate test agreement between blood, urine, and kidney samples (research laboratory) for leptospirosis diagnosis. Samples were sourced from 399 sheep and 146 beef cattle from a local abattoir. Interlaboratory agreement for real-time quantitative polymerase chain reaction (qPCR) results on urine samples was almost perfect (kappa = 0.90), despite the use of different amplification targets (DNA gyrase subunit B gene vs. 16s ribosomal RNA gene), chemistries (SYTO9 vs. TaqMan probe), and pre-PCR processing. Interlaboratory agreement for microscopic agglutination test (MAT) positivity was almost perfect (kappa = 0.93) for Leptospira borgpetersenii serovar Hardjo subtype Hardjobovis (Hardjobovis) but moderate (kappa = 0.53) for Leptospira interrogans serovar Pomona (Pomona). Among animals that had different titers recorded, higher Hardjobovis and lower Pomona titers were reported by the commercial laboratory than by the research laboratory ( P < 0.005). These interlaboratory comparisons can assist researchers and diagnosticians in interpreting the sometimes discrepant test results. Within the research laboratory, the comparison of qPCR results on urine and kidney showed almost perfect agreement (kappa = 0.84), suggesting that the qPCR on these 2 specimens can be used interchangeably. The agreement between MAT positivity and urine and kidney qPCR results was fair (kappa = 0.32 and kappa = 0.33, respectively). However, the prevalence ratio of urine and kidney qPCR positivity in Hardjobovis-seropositive versus Hardjobovis-seronegative sheep indicated that Hardjobovis seropositivity found in sheep may be able to predict shedding or renal carriage.