Evaluation of sensitivity and specificity of RBT, c-ELISA and fluorescence polarisation assay for diagnosis of brucellosis in cattle using latent class analysis

G. Matope, J. B. Muma, Nils Toft, E. Gori, A. Lund, K. Nielsen, E. Skjerve

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

The sensitivity (Se) and specificity (Sp) of the Rose Bengal test (RBT), competitive ELISA (c-ELISA), serum (sFPA) and blood (bFPA) fluorescence polarisation assay for brucellosis were evaluated using latent class analysis using sera and whole blood collected from infected cattle reared in smallholder dairy farms of Zimbabwe. The latent class model allowed estimation of Se and Sp in the absence of a gold standard test. The c-ELISA had the highest Se (99.0%; 95% credible posterior interval (CPI): 94.8; 100%), while the RBT and sFPA had the highest Sp (99.0%; 95% CPI: 98.0; 99.6%). The bFPA had the lowest Se (71.3%; 95% CPI: 56.2, 83.5%), while its Sp (96.3%; CPI: 93.9; 98.0%) was marginally higher than that of the c-ELISA (95.4% CPI: 93.7; 96.8%). Therefore based on these data, test regimen using the RBT and c-ELISA could be suitable for diagnosis of brucellosis in smallholder dairies in Zimbabwe. Based on cost and ease of performance, the sFPA may be adopted as a confirmatory test, but its performance may be optimised by altering cut-off points to suit the Zimbabwean conditions. Thus, latent class models provide an alternative method for evaluating Se and Sp of diagnostic tests, which could be used to optimise test performance in different cattle populations.
Original languageEnglish
JournalVeterinary Immunology and Immunopathology
Volume141
Issue number1-2
Pages (from-to)58-63
Number of pages6
ISSN0165-2427
DOIs
Publication statusPublished - 2011
Externally publishedYes

Keywords

  • RBT
  • c-ELISA
  • FPA
  • Brucellosis
  • Latent class analysis

Fingerprint

Dive into the research topics of 'Evaluation of sensitivity and specificity of RBT, c-ELISA and fluorescence polarisation assay for diagnosis of brucellosis in cattle using latent class analysis'. Together they form a unique fingerprint.

Cite this