Cost-effectiveness evaluation of bovine tuberculosis surveillance in wildlife in France (Sylvatub system) using scenario trees

Bovine tuberculosis (bTB) is a common disease in cattle and wildlife, with health, zoonotic and economic implications. Infected wild animals, and particularly reservoirs, could hinder eradication of bTB from cattle populations, which could have an important impact on international cattle trade. Therefore, surveillance of bTB in wildlife is of particular importance to better understand the epidemiological role of wild species and to adapt the control measures. In France, a bTB surveillance system for free-ranging wildlife, the Sylvatub system, has been implemented since 2011. It relies on three surveillance components (SSCs) (passive surveillance on hunted animals (EC-SSC), passive surveillance on dead or dying animals (SAGIR-SSC) and active surveillance (PSURV-SSC)). The effectiveness of the Sylvatub system was previously assessed, through the estimation of its sensitivity (i.e. the probability of detecting at least one case of bTB infection by each SSC, specie and risk-level area). However, to globally assess the performance of a surveillance system, the measure of its sensitivity is not sufficient, as other factors such as economic or socio-economic factors could influence the effectiveness. We report here an estimation of the costs of the surveillance activities of the Sylvatub system, and of the cost-effectiveness of each surveillance component, by specie and risk-level, based on scenario tree modelling with the same tree structure as used for the sensitivity evaluation. The cost-effectiveness of the Sylvatub surveillance is better in higher-risk departments, due in particular to the higher probability of detecting the infection (sensitivity). Moreover, EC-SSC, which has the highest unit cost, is more efficient than the surveillance enhanced by the SAGIR-SSC, due to its better sensitivity. The calculation of the cost-effectiveness ratio shows that PSURV-SSC remains the most cost-effective surveillance component of the Sylvatub system, despite its high cost in terms of coordination, sample collection and laboratory analysis.

牛结核病（Bovine tuberculosis, bTB）是牛类与野生动物的常见疫病，对健康、人畜共患传播及经济均具有重要影响。受感染的野生动物，尤其是疫病储存宿主，可能会阻碍牛群中bTB的根除工作，而这会对国际肉牛贸易产生重大影响。因此，针对野生动物的bTB监测对于深入了解野生动物物种的流行病学作用、优化防控策略具有特殊重要意义。
在法国，一套针对野外散居野生动物的bTB监测系统——Sylvatub系统——自2011年起投入使用。该系统依托三类监测组分（Surveillance Components, SSCs）开展工作，分别为针对狩猎动物的被动监测（EC-SSC）、针对病死或濒死动物的被动监测（SAGIR-SSC）与主动监测（PSURV-SSC）。
此前已有研究通过估算其灵敏度（即各监测组分、物种与风险等级区域至少检出1例bTB感染病例的概率），对Sylvatub系统的监测效能进行了评估。然而，若要全面评估监测系统的整体性能，仅依靠灵敏度指标并不足够，因为经济或社会经济等其他因素同样会影响系统效能。
本研究基于与灵敏度评估中相同的树状结构开展情景树建模，据此估算了Sylvatub系统各项监测活动的成本，以及各监测组分按物种与风险等级划分的成本效益。研究结果显示，Sylvatub监测系统在高风险省份的成本效益更优，这主要得益于其更高的感染检出概率（即灵敏度）。此外，尽管EC-SSC的单位成本最高，但其灵敏度更优，因此相较于经SAGIR-SSC强化的监测模式效率更高。成本效益比的计算结果表明，尽管PSURV-SSC在协调工作、样本采集与实验室分析方面成本较高，但其仍是Sylvatub系统中成本效益最优的监测组分。