Cost-effectiveness assessment of alternative surveillance measures for bovine tuberculosis using stochastic modeling simulation

To respond to the increasing bovine Tuberculosis (bTB) infection rates observed in France since 2010, national authorities are considering alternative surveillance measures for bTB in Reinforced Screening Areas, defined as municipalities within 5 or 10 km range of a bTB outbreak. The objective of this study was to determine the costeffectiveness, from the farmer expenses point of view, of three alternatives to the current mandatory surveillance scenario applied on French dairy, beef and small (<10 heads) farms located in Reinforced Screening Areas (RSA), using stochastic economic and epidemiologic modeling simulation. These alternative measures include reducing bTB minimum screening age for the annual screening program (M1), systematic bTB screening before animal movement (M2), and performing an interferon-gamma test (INF-gamma) instead of strict quarantine and post-mortem diagnosis after a bTB positive test (M3). We show that the implementation of M1 and M2 increases annual costs ranging from <euro>6 to <euro>828 per farm and from <euro>56 to <euro>647 per farm, respectively, whereas the implementation of M3 led to average annual cost reductions between <euro>3 and <euro>352 per farm. These reductions were linked to fewer animals culled for post-mortem diagnostic and to a lower probability of farm quarantine or exclusion from the international markets. We also show that using a combination of the three alternatives measures in RSA farms could lead to detect between 11 and 17 additional bTB outbreaks per year. A highly cost-effective set of measures is to keep M1 at 24 months old while implementing M2 and M3. This strategy has an average annual cost of <euro>209 per farm and allows to detect 11 additional bTB outbreaks per year compared with the baseline program. The most cost-effectiveness policy to detect more bTB outbreaks, minimizing the marginal cost of additional bTB outbreak detection, is to establish M1 at 14 months while implementing M2 and M3. This approach had average annual cost of <euro>352 per farm and allowed to detect 15 additional bTB outbreaks per year. As a conclusion, the assessment of alternative surveillance measures through simulation stochastic models could assist policymakers to implement surveillance policy that balances disease monitoring costs and health outcomes. The present work supports the adoption of M1, M2 and M3 and reducing the minimum screening age at annual testing to 14 months.