A novel Bayesian Latent Class Model (BLCM) evaluates multiple continuous and binary tests: A case study for Brucella...

Area of Science:

• Veterinary Epidemiology
• Diagnostic Test Evaluation
• Bayesian Statistics

Background:

• Bovine brucellosis, caused by Brucella abortus, poses significant risks to animal and human health.
• Accurate diagnosis is essential for effective control and prevention strategies.
• The absence of a gold standard test complicates the evaluation of diagnostic test performance and cut-off value determination.

Purpose of the Study:

• To develop and apply a novel Bayesian Latent Class Model for calibrating optimal cut-off values for diagnostic tests.
• To evaluate the diagnostic performance of four serological tests for bovine brucellosis.
• To estimate true disease prevalence and identify risk factors in Chinese dairy farms.

Main Methods:

• A Bayesian Latent Class Model was developed, integrating binary and continuous test results with fixed (parity) and random (farm) effects.
• Six hundred fifty-one serum samples from six dairy farms in Henan Province, China, were tested using Rose Bengal Test, Serum Agglutination Test, Fluorescence Polarization Assay, and Competitive Enzyme-Linked Immunosorbent Assay.
• Optimal cut-off values were determined by maximizing the Youden Index.

Main Results:

• Optimal cut-off values were identified as 94.2 mP for Fluorescence Polarization Assay and 0.403 PI for Competitive Enzyme-Linked Immunosorbent Assay.
• Test sensitivities ranged from 69.7% to 89.9%, and specificities ranged from 97.1% to 99.6%.
• True prevalences were 4.7% and 30.3% in the two study regions, with parity influencing serological status (ORs 1.2–2.2).

Conclusions:

• The developed Bayesian model offers a robust framework for validating diagnostic tests in the absence of a gold standard.
• The findings provide optimized cut-off values and performance estimates for key serological tests used in bovine brucellosis surveillance.
• This research supports the design of targeted detection strategies and effective brucellosis control measures in Chinese dairy populations.