Estimation of sensitivity and specificity of five serological tests for the diagnosis of porcine brucellosis.

While serological tests are essential in surveillance and control programs of animal diseases, to date none of the common serological tests approved in the EU (complement fixation test or Rose-Bengal test) has been shown to be reliable in routine individual diagnosis of porcine brucellosis, and some more recent tests like ELISA have not been fully evaluated yet. In the absence of a gold standard, this study allowed the estimation of sensitivities and specificities of these tests with a Bayesian approach using Markov Chain Monte Carlo algorithms. The pig population that was tested included 6422 animals from Metropolitan France. Serum samples were collected from a large population of pigs, representative of European swine population and tested with five brucellosis serological tests: Rose-Bengal test (RBT), fluorescence polarization assay (FPA), indirect ELISA (I-ELISA) and two competitive ELISAs (C-ELISA). The sensitivity and the specificity of each test were estimated. When doubtful results were excluded, the most sensitive and specific test was C-ELISA(2) (Se C-ELISA(2)=0.964, [0.907; 0.994], 95% credibility interval (CrI); Sp C-ELISA(2)=0.996, [0.982; 1.0], 95% CrI). When doubtful results were considered as negative, C-ELISA(2) was still the most sensitive and specific test (Se C-ELISA(2)=0.960, [0.896; 0.994], 95% CrI and Sp C-ELISA(2)=0.994, [0.977; 0.999], 95% CrI). The same conclusions were reached when doubtful results were considered as positive (Se C-ELISA(2)=0.963, [0.904, 0.994], 95% CrI and Sp C-ELISA(2)=0.996, [0.986; 1.0], 95% CrI).

[1]  B. Goichot,et al.  La statistique bayésienne : une approche des statistiques adaptée à la clinique , 2009 .

[2]  M Greiner,et al.  Epidemiologic issues in the validation of veterinary diagnostic tests. , 2000, Preventive veterinary medicine.

[3]  I A Gardner,et al.  Estimation of diagnostic-test sensitivity and specificity through Bayesian modeling. , 2005, Preventive veterinary medicine.

[4]  W O Johnson,et al.  Estimation of sensitivity and specificity of diagnostic tests and disease prevalence when the true disease state is unknown. , 2000, Preventive veterinary medicine.

[5]  David B. Dunson,et al.  Bayesian Data Analysis , 2010 .

[6]  H Stryhn,et al.  Conditional dependence between tests affects the diagnosis and surveillance of animal diseases. , 2000, Preventive veterinary medicine.

[7]  Alvin D. Wiggins,et al.  Cut-off points for aggreate herd testing in the presence of disease clustering and correlation of test errors , 1994 .

[8]  P M Vacek,et al.  The effect of conditional dependence on the evaluation of diagnostic tests. , 1985, Biometrics.

[9]  Joseph L. Gastwirth,et al.  Bayesian Inference for Medical Screening Tests: Approximations Useful for the Analysis of Acquired Immune Deficiency Syndrome , 1991 .

[10]  M. Greiner,et al.  Scientific Opinion of the Panel on Animal Health and Welfare , 2009 .

[11]  C. Chevalet,et al.  Genetic diversity within and between European pig breeds using microsatellite markers. , 2006, Animal genetics.

[12]  E. Skjerve,et al.  Effectiveness of Rose Bengal test and fluorescence polarization assay in the diagnosis of Brucella spp. infections in free range cattle reared in endemic areas in Zambia , 2009, Tropical Animal Health and Production.

[13]  G. Briones,et al.  Evaluation of Primary Binding Assays for Presumptive Serodiagnosis of Swine Brucellosis in Argentina , 2000, Clinical Diagnostic Laboratory Immunology.

[14]  Emmanuel Lesaffre,et al.  Estimating Disease Prevalence in a Bayesian Framework Using Probabilistic Constraints , 2006, Epidemiology.

[15]  K. Nielsen.  Diagnosis of brucellosis by serology. , 2002, Veterinary microbiology.

[16]  Bradley P. Carlin,et al.  Bayesian measures of model complexity and fit , 2002 .

[17]  L. Joseph,et al.  Bayesian Approaches to Modeling the Conditional Dependence Between Multiple Diagnostic Tests , 2001, Biometrics.

[18]  B. Garin‐Bastuji,et al.  Efficacy of Several Serological Tests and Antigens for Diagnosis of Bovine Brucellosis in the Presence of False-Positive Serological Results Due to Yersinia enterocolitica O:9 , 2005, Clinical Diagnostic Laboratory Immunology.

[19]  Wesley O Johnson,et al.  On the interpretation of test sensitivity in the two-test two-population problem: assumptions matter. , 2009, Preventive veterinary medicine.

[20]  K. Nielsen,et al.  Enzyme Immunoassays for Serological Diagnosis of Bovine Brucellosis: A Trial in Latin America , 1998, Clinical Diagnostic Laboratory Immunology.

[21]  K. Nielsen,et al.  Evaluation of an indirect enzyme-linked immunoassay for presumptive serodiagnosis of Brucella ovis in sheep , 2003 .

[22]  K. Nielsen,et al.  Validation of a second generation competitive enzyme immunoassay (CELISA) for the diagnosis of brucellosis in various species of domestic animals. , 2008, Veterinary immunology and immunopathology.

[23]  P. Bossuyt,et al.  Evaluation of diagnostic tests when there is no gold standard. A review of methods. , 2007, Health technology assessment.

[24]  Andrew Thomas,et al.  WinBUGS - A Bayesian modelling framework: Concepts, structure, and extensibility , 2000, Stat. Comput..

[25]  K. Nielsen,et al.  Serological diagnosis of bovine brucellosis: a review of test performance and cost comparison. , 2004, Revue scientifique et technique.