Economic burden and cost-effectiveness of brucellosis control strategies in large commercial dairy herds in Bangladesh: A Bayesian decision-analytic study.

Bovine brucellosis caused by Brucella abortus is endemic in Bangladesh, but its financial cost and the economics of control have not been quantified for the commercial dairy sector. We estimated the annual economic burden of infection in 17 large commercial dairy herds (194-2850 animals; 2696 animals tested), compared six control strategies over a 10-year economic horizon, and simulated a phased programme combining adult mass vaccination with subsequent test-and-cull. Herd-specific Bayesian posterior estimates of true prevalence and force of infection (FOI, the annual hazard of infection in a susceptible animal) were carried into a stochastic Monte Carlo model (50,000 iterations) built on a Latin hypercube sampling (LHS) design, in which infected breeding females were allocated to four mutually exclusive outcome states so that loss components could be summed without double counting. A compartmental model of susceptible, vaccinated and infectious animals, with an additional class for animals whose vaccine-induced immunity had waned, projected infection over 20 years. Mean annual loss across the cohort was 36.03 million Bangladeshi taka (BDT) (95% uncertainty interval 21.65-52.06; approximately USD 307,900 at the 2024 exchange rate of 117 BDT per USD). Reduced milk yield accounted for 45.5% of the total loss and an extended calving interval for 34.6%. In a global sensitivity analysis using partial rank correlation coefficients (PRCC), abortion probability (PRCC 0.890, 95% CI 0.888-0.892) and the subclinical milk deficit (0.855) dominated uncertainty in the burden. Vaccination was the efficient instrument: mean benefit-cost ratio (BCR) was 21.00 for heifer-only vaccination and 9.08 for mass vaccination, exceeding unity in all 17 herds, but neither strategy eliminated infection within 20 years in more than two herds. Elimination in all 17 herds required serological culling. The decisive constraint was not the phase schedule but the adult vaccination protocol. All 23 combinations of phase-1 duration and phase overlap eliminated infection in every herd, at a discounted cohort cost of 69.7-168.0 million BDT. Holding that schedule fixed and replacing the reduced-dose conjunctival adult protocol with full-dose subcutaneous S19 raised cost 4.6-fold to 545.3 million BDT and turned cohort net benefit from +57.85 to -368.28 million BDT, because 22,695 uninfected animals were culled as vaccinal seroreactors instead of 818. Shortening the interval between campaign and test round from 26 to 4 weeks had the same effect. Mass vaccination is the first-line intervention for large commercial Bangladeshi dairy herds, and elimination requires a phased programme with serological culling; whether such a programme is affordable is determined by the dose, route and campaign-to-test interval of adult vaccination rather than by the length or overlap of its phases. These results apply to the large commercial sub-sector only and exclude human health costs, so the reported benefit-cost ratios are lower bounds on societal return.