Introduction: Despite eradication and control measures applied across Europe, bovine tuberculosis (bTB) remains a constant threat. In Belgium, after several years of bTB disease freedom status, routine movement testing, as currently practiced, revealed itself inadequate to detect some sporadic breakdown herds. The aim of this study was to strike the balance between cost and effectiveness of different surveillance system components to identify sustainable alternatives for early detection and substantiation of freedom of bTB while maintaining acceptance of these amongst the different animal health stakeholders. Methods: Stochastic iteration model was built to simulate, first, the expected current surveillance system performance in terms of sensitivity and specificity of detection. These results were then descriptively compared to observed field results. Secondly, the cost and effectiveness of simulated alternative surveillance components were quantified. To measure impact of key assumptions (i.e. regarding diagnostic tests and true prevalence), sensitivity analysis was performed. Results: Discrepancies between the predicted and observed performance of bTB surveillance in Belgium were observed. Secondly, simulated alternatives revealed that targeted IFN-γ as well serological testing with Antibody ELISA towards risk herds would enable increasing the overall cost and effectiveness of the Belgian bTB surveillance system. Sensitivity analysis showed that results remained constant despite modification of some key assumptions. Discussion: Performance of current bTB surveillance system performance in Belgium was questionable. This exercise highlighted that not only sensitivity, but specificity is a key driver for surveillance performance. The quantitative and participative conceptual framework revealed itself a useful tool to allow evidence-based decision making regarding future tuberculosis surveillance in Belgium, as required by the international standards.

Vertical and horizontal transmissions of Salmonella spp. along the food chain (FC) have been reported to play an important role in the contamination of foodstuffs and of humans. This study aims first to clarify to what extend some selected major Salmonella serotypes are represented along the chicken FC and potentially transmitted from feed to humans. Then to discuss if the sampling results of the Belgian National Salmonella monitoring program may be considered as proxy to predict possible dominant serotypes and trends in humans. By using results of this programme, this study tried to highlight statistically significant differences in yearly prevalences between broilers exit status and other matrices (e.g. feed, other chicken categories, foodstuffs). Then to show up statistically significant changes from year to year prevalences in broilers exit status and to look if these changes can also be observed in the other matrices. Results have showed significant differences in yearly prevalences. They also pinpointed that the prevalence for the same serotypes could significantly vary from year to year in broilers exit status but this variation was inconsistently observed in the other matrices. These observations do not support the assumption of significant vertical and horizontal bacterial transmission along the chicken FC. Results in the national program cannot be considered as proxy. Determining the contribution of contaminated animal feed to human illness is challenging as the interplay of many factors makes it difficult. Coupling the actual programme with network analysis, source attribution and genome sequencing would be a way to explore deeper the spread of Salmonella serotypes along the FC. Belgium has put efforts to decrease Salmonella Enteritidis and Salmonella Typhimurium prevalence along the FC, but attention should be also paid to other pathogens that could fill the vacant niche left as Salmonella infantis which prevalence has been increasing in some matrices.