Abstract
<jats:p>Brucellosis, primarily caused by Brucella abortus, is a significant zoonotic disease that adversely affects both public health and the livestock economy, particularly in regions reliant on animal husbandry. This study aimed to evaluate the spatial distribution and estimate the Relative Risk (RR) of B. abortus infections in Malaysia from 2018 to 2024, utilizing two statistical approaches: the Standardized Mortality Ratio (SMR) and the Bayesian Poisson-Gamma model. Disease incidence data were obtained from the World Organization for Animal Health (WOAH), and analyses were conducted using R Programming and ArcGIS software to generate risk estimates and produce spatial disease maps across 14 Malaysian states. The SMR provided initial risk estimates but exhibited limitations in regions with zero reported cases, often underrepresenting potential disease burden. In contrast, the Poisson-Gamma model yielded more nuanced and robust risk estimations, identifying additional high-risk areas such as Kuala Lumpur and Sarawak, where the SMR reported no observed risk. This discrepancy highlights the Bayesian model’s strength in addressing data sparsity and underreporting. Both models consistently identified Perlis, Perak, and Johor as very high-risk states. The study concludes that the Poisson-Gamma model offers superior performance in detecting spatial risk patterns of B. abortus, particularly in areas with incomplete surveillance data. However, its limitations include reduced flexibility in adjusting for covariates and spatial dependencies. Furthermore, these findings underscore the importance of adopting advanced spatial modeling techniques in disease surveillance to inform targeted interventions, optimize resource allocation, and support evidence-based policy development in managing brucellosis and similar zoonotic diseases.</jats:p>