USE OF DIAGNOSTIC DATA TO RAPIDLY DETECT AND RAISE STAKEHOLDER AWARENESS OF EMERGING BIOLOGICAL ANIMAL HEALTH AND PRODUCTION THREATS

<b>Project Methods</b><br> Research 1: Early detection of emerging animal health threats through the usage of genetic sequences: Average PRRSV Ct values from samples submitted for PRRSV ORF5 sequencing will be scanned daily by an EWMA model to monitor and inform potential changes from expected baselines. A computational bioinformatics and cyberinfrastructure approach will provide a pipeline implementing classic phylogenetic and minimum spanning tree (MST) networks to show genetic proximity between sequences. Pairwise evolutionary distances will be computed and presented as a complete graph where the vertices represent the sequences. An MST is computed, i.e., a tree spanning the graph's vertices with the minimum overall weight of the edges involved. Based on a carefully selected cut-off value, short-distance edges will be added to the MST, resulting in the MST network demonstrating and revealing strong interactions between clades and lineages.Research 2: Develop and implement a monitoring co-agent detection at the overall and state-level: A multivariate spatial-temporal framework for rapid detection of changes in the network's structure will be implemented, allowing to investigate patterns of various disease/agents by sample type, age category as well as forecast significant changes at various locations and times of interest. Each state will be considered a network node. Daily updates will add support for online prediction, model selection, and changepoint detection. Implicit model penalization through a BOCPDMS will compute the MSE as well as the NLL for model selection. Spatial distribution of agent/co-agent detection will be learned and weighted from historical data allowing for local dependency across time points and weighted to reduce the detection delay and increase robustness to potentially irrelevant past data. A threshold will be set to control false alarms. A SBM to characterize edge densities between and within different data sources (agents) will model networks at a time of interest. Community detection algorithms will identify clusters of geographic areas where there is increased detection.Research 3: Assessment of SDRS data regional representativeness for the significance of the number of (#) cases tested across the USA. Swine population data based on an external and appropriate source, such USDA for hog inventory and US Census of Agriculture for # farms by state, will be used. Baseline for the # submissions by state will be stablished based on historical data to compare with the current year's submission. A state-specific normalization approach will determine the expected # submissions based on the # swine farms or hog inventory. A chi-square goodness of fit test will assess whether the expected and observed are significantly different over time to shed insight into regions with low submissions. Unlabeled site state submissions will be accounted. False discovery rate will be controllled using a Benjamini-Hochberg procedure.Outreach and extension 1: content sharing through meetings with the swine industry stakeholders: The project PIs will participate and help organize workshops and seminars targeting conferences heavily attended by veterinarians, producers, academics, and allied industries associated with the US swine industry to disseminate knowledge generated from the SDRS and capture feedback on the type, format, and quality of information from the industry back to the research, education, and extension teams. This will enhance the ability to continuously improve the project channels of communication to support stakeholders with animal health information.Outreach and extension 2: Real-time delivery of information to comunicate early detection of emerging or re-emerging threats in a timely manner. The frequency of reporting will move from monthly to on-demand (as frequently as daily) based on the change in patterns of detection. Task Scheduler batch files will be programmed to run surveillance scripts and update online dashboards daily, scanning the database for significant changes in the overall pattern of PCR detection, co-agent PCR detection at the overall and state-level, and PRRSV ORF5 strains detection. Relevant signals will be shared with the industry via the user-friendly newly developed dashboards updated daily and via the monthly newsletteror, SDRS online communication channels, or thought customized and extra newsletter if emerging threats are identified. At least three extra webinars (1/year) will be necessary to share time-critical information. A survey directed to different swine stakeholder groups will be performed within 3 months of the project and yearly after that to document changes in the perception of knowledge on early detection and regional spread of infectious diseases affecting swine industry.Outreach and extension 3: Producer targeted education through regional meetings and webinars. The Iowa Pork Industry Center (IPIC), will organize and host regional meetings and workshops to deliver and train stakeholders on the information generated under this proposal. Regional meeting rounds (n=4) will take place across Iowa. Based on the feedback from the regional meetings, a second phase will take place and target the delivery of information to pork producers across other USA. A paper or online survey will be distributed at the end of each in-person meeting or webinar to capture producer feedback on the SDRS content and improvement opportunities. This will raise producers awareness of how strategically address measures to control and prevent the spread of emerging or re-emerging threatening agents/diseases.Educational 1: Veterinary student education, training, and engagement with the swine industry: Curriculum Development and Distribution: a 8 hours of professional and graduate package of educational materials containing modules organized to cover the importance and purpose of the SDRS surveillance, case definitions, the requested diagnostic data from each of participant VDL, terminology such LOINC codes and SNOMMED CT, importance, significance, result interpretation, comparison with production system-specific diagnostic data analysis, and aggregated (multiple VDLs) veterinary diagnostic data usage will be developed. Further, content developed by the SDRS program will be provided to supplement introductory and specific syndromic surveillance explained using practical SDRS concepts and made readily and freely available to be used by courses taught at USA universities. A pre and post-class survey will be used to measure the success of delivering this information. Individual interviews will be conducted to assess how the students would implement these concepts in real-world scenarios. Each year, the program will advertise and solicit at participant institutions veterinary student and graduate student applications to serve on the SDRS Advisory Board, enabling both traditional and non-traditional students to gain valuable experience in extension, outreach, and research activities outlined in this grant by serving on the advisory board.Educational 2: Post Graduate Education Training and Development: Scholarships will also be applied to graduate students by combining classroom experiences, research and extension training, and professional development. Graduate students can get engaged in the SDRS project and have hands-on syndromic surveillance development and application using practical examples from SDRS. These students will have the opportunity to develop and validate early detection concepts and models and share their advancements with stakeholders through the outreach and extension channels. Graduate students will be required to participate in quarterly Advisory Board meetings and present their scholarly activities in research and extension to the entire investigative and collaborative team. A pre and post-class survey will be used to measure students' knowledge of the subject before and after the course.