{ "background": "Public health surveillance systems are critical for early disease detection and response, yet their economic evaluation remains methodologically challenging. Current cost-effectiveness analyses often rely on deterministic models that inadequately capture the inherent uncertainty and hierarchical structure of surveillance data, particularly in resource-limited settings.", "purpose and objectives": "This protocol details a novel Bayesian hierarchical modelling framework designed to evaluate the cost-effectiveness of public health surveillance systems. The primary objective is to provide a robust, probabilistic method for integrating heterogeneous cost and outcome data to estimate system performance and value for money.", "methodology": "We propose a Bayesian cost-effectiveness model structured as $\\text{log}(\\text{CE}{ij}) = \\beta0 + ui + vj + \\epsilon{ij}$, where $\\text{CE}{ij}$ is the cost-effectiveness ratio for surveillance component $j$ in region $i$, $\\beta0$ is the overall mean, $ui$ and $vj$ are random effects for region and component, and $\\epsilon{ij}$ is the residual error. Prior distributions will be informed by expert elicitation and historical data. The model will be fitted using Markov chain Monte Carlo methods to data on costs, detected cases, and response timeliness from multiple surveillance tiers.", "findings": "As this is a protocol, no empirical findings are presented. However, the proposed model is designed to yield probabilistic estimates, such as the posterior probability that a given surveillance component achieves a cost per disability-adjusted life year averted below a willingness-to-pay threshold, with credible intervals quantifying uncertainty.", "conclusion": "This methodological framework is expected to advance the rigour of economic evaluations in public health surveillance by formally accounting for data uncertainty and hierarchical dependencies, thereby supporting more informed resource allocation decisions.", "recommendations": "Future applications should incorporate real-time data feeds to enable dynamic, ongoing cost-effectiveness assessments. The framework should be validated and adapted for other disease programmes and country contexts.", "key words": "Bayesian inference