Abstract
{ "background": "Public health surveillance systems are critical for disease control, yet their reliability in resource-limited settings is often uncertain. Existing evaluation frameworks lack robust quantitative methods to integrate heterogeneous data and account for hierarchical data structures inherent in national surveillance networks.", "purpose and objectives": "This study aimed to develop and apply a novel Bayesian hierarchical model to quantitatively assess the reliability of a national public health surveillance system, using Ethiopia as a case study. The objective was to estimate system completeness and timeliness while quantifying uncertainty.", "methodology": "We developed a Bayesian hierarchical model integrating longitudinal reporting data from multiple administrative levels. The core reliability metric, the probability of a correct and timely report, was modelled as $\\text{logit}(p{ijt}) = \\alpha + ui + vj + \\beta X{ijt}$, where $ui \\sim N(0, \\sigmau^2)$ and $vj \\sim N(0, \\sigmav^2)$ are random effects for region $i$ and disease $j$. Inference used Markov chain Monte Carlo simulation with weakly informative priors.", "findings": "The model estimated national surveillance reliability at 0.72 (95% credible interval: 0.68 to 0.76), with substantial subnational variation. Reliability showed a positive temporal trend, with a posterior probability of 0.94 that the trend was greater than zero. Completeness was a more significant constraint than timeliness.", "conclusion": "The Bayesian hierarchical model provides a robust, integrative framework for surveillance reliability assessment, quantifying both central tendency and uncertainty. The system demonstrates moderate overall reliability with clear inequities across regions.", "recommendations": "Implement the model for routine performance monitoring to identify priority regions for intervention. Allocate resources based on modelled reliability estimates and their uncertainty to improve system equity and resilience.", "key words": "Bayesian inference, disease surveillance, health systems, hierarchical model, reliability, uncertainty quantification", "contribution statement": "This paper introduces