A Framework for Translational Veterinary Pathogen Risk Modeling

Thesis Abstract

Emerging infectious diseases at the animal–human interface pose significant threats to global health, food security, and veterinary public health, underscoring the need for a translational framework that bridges pathogen risk assessment across veterinary systems and human health surveillance. This study aims to develop a robust framework for translational veterinary pathogen risk modeling capable of informing decision-making in antimicrobial stewardship, biosecurity, and zoonotic surveillance. Specific objectives are (1) to synthesize theoretical constructs from One Health, risk analysis, and systems theory to define a modular translational model linking veterinary pathogen dynamics to human health risk indicators; (2) to operationalize a composite risk index using multi-omics surveillance data, veterinary diagnostic outputs, and farm-level management practices; (3) to validate the framework against historical outbreak data and prospective sentinel surveillance in diverse farming contexts; (4) to evaluate model sensitivity, calibration, and transferability across livestock species and geographies; and (5) to develop guidelines for stakeholder adoption in veterinary laboratories, regulatory agencies, and veterinary clinics. The methodological approach adopts a mixed-methods design anchored in theoretical integration and empirical validation. The population comprises large- and small-scale ruminant and porcine farming systems across three regions with varying biosafety levels. A purposive sample of 60 farms will be recruited for in-depth data collection, complemented by data from 12 national veterinary laboratories and 6 regional public health institutes. Quantitative data will be gathered from diagnostic test results (n = 10,000 records per year over three years), environmental and management variables (housing density, biosecurity scores, vaccination status), and reported clinical outcomes. Qualitative insights will be obtained from semi-structured interviews with 40 veterinarians and 20 public health officials to capture decision-making processes and perceived barriers to data integration. The primary data collection instruments include a standardized diagnostic reporting protocol, a validated farm-level biosecurity and management survey, and a semi-structured interview guide aligned with translational risk constructs. Validity and reliability will be ensured through pilot testing, inter-rater reliability checks for data coding, and triangulation across data streams. Analytical methods will integrate statistical, computational, and qualitative techniques. A Bayesian hierarchical model will be employed to estimate pathogen transmission probabilities and translate veterinary surveillance signals into human health risk estimates, incorporating prior information from literature and expert elicitation. Regression analyses (logistic and Poisson) will examine associations between farm-level factors and outbreak indicators. Time-series analyses will assess temporal alignments between veterinary detections and human health events. A network analysis will illuminate inter-farm and inter-institutional information flows that influence translational risk. The analytical framework will be augmented by machine learning approaches, including gradient boosting and random forest algorithms, to enhance predictive performance and identify key drivers of risk. Theoretical grounding will draw on One Health, Technological Frames theory, and the Protection Motivation Theory to interpret adoption of the framework and risk communication implications. The study will produce a conceptual model diagram and a computational prototype implementing the translational risk scoring mechanism. Expected findings include (i) a validated translational framework linking veterinary pathogen signals to human health risk; (ii) a calibrated composite risk index with acceptable discrimination (AUC ? 0.80) and good calibration across species and regions; (iii) identifiable high-impact drivers (biosecurity lapses, vaccination gaps, co-infection dynamics) and critical data gaps hindering risk translation; and (iv) actionable guidelines for integrating veterinary diagnostics, environmental monitoring, and public health surveillance into a unified decision-support tool. The contribution to knowledge lies in formalizing a translational pathway for veterinary pathogen data to inform human risk assessment, bridging methodological gaps between veterinary epidemiology and public health analytics, and providing a transferable framework adaptable to varied regulatory contexts. The study concludes that timely, harmonized data streams and participatory governance are essential for effective translational risk modeling. Recommendations include establishing standardized data-sharing protocols, expanding sentinel surveillance networks, investing in interoperable information systems, and training stakeholders in translational risk interpretation to bolster preparedness for emerging zoonoses.