Quantification of the early detection surveillance: combining technique to estimate the virus spread and the sensitivity of the surveillance system

Early disease detection and estimation of the sensitivity surveillance system are used for various purposes, including the early detection of non‐communicable diseases, of unusual increases of disease frequency, the first occurrence of a disease in a previously free population, or providing evidence of disease eradication. Quantifying the sensitivity of early detection surveillance allows important aspects of the performance of different systems, approaches and authorities to be evaluated, compared and improved. Quantifying the sensitivity of early detection surveillance allows important aspects of the performance of different systems, approaches and authorities to be evaluated, compared and improved. This work, using a contagious animal disease example (i.e., African swine fever (ASF)), tested a new approach to provide evidence of disease absence quantifying the overall sensitivity of the surveillance system, in terms of population coverage, temporal coverage and detection sensitivity. A practical interpretation of the strategy was implemented based on the failure probabilities of wrongly declaring the freedom of an area even if the disease agent is still present but undetected. ASF is a devastating disease, resulting in the high mortality of domestic and wild pigs, that as from 2007 is spreading around the world. ASF has been endemic in Sardinia for more than 40 years, and the last ASFV detection dates to 2019. This work aimed at: i) providing an overview of the main actions and measures put in place to fight the disease in the wild boar, ii) compare the disease trends between 2011-2016 and 2017-2019, describing the strengths and weaknesses of the surveillance plans, iii) estimate the efficacy of these measures (sensitivity of the ASF surveillance system) from the last virus detection (2019-2023). Analyses were performed using a number of different design prevalence combinations, to examine the impact of the assumptions under the null hypothesis. The results suggest that the surveillance system was able to detect virus circulation at a design prevalence below 1% and 2%. High values of negative predictive values (>95%) demonstrated that we could be confident affirming that the Sardinian domestic pig farms and wild boar population are free from ASFV from 2019 and 2021, respectively, given that the surveillance system has given negative outcomes. The results of analysis indicated that (not surprisingly) the estimated system sensitivity (or equivalently, confidence in the surveillance system) was very sensitive to the design prevalence assumptions under the null hypothesis. Indeed, the role of passive surveillance as target surveillance was confirmed in term of efficacy in detecting the virus, and amount of sample need. Otherwise, its sustainability during several years is currently under discussion. Further targeted research, such as long-term experimental studies and molecular epidemiology, is necessary to improve our knowledge on the epidemiology of ASF and to control the disease more effectively, including the use of more formal methods to generate estimates from expert opinion, and the application of a proposed methodology to account for the lack of independence between surveillance systems.