{ "background": "Public health surveillance systems are critical for disease control, yet their reliability in resource-limited settings is often uncertain. In South Africa, evaluating these systems has historically relied on fragmented, cross-sectional assessments lacking a formal framework for integrating heterogeneous data and quantifying uncertainty over time. ", "purpose and objectives": "This study aimed to develop and apply a novel Bayesian hierarchical model to quantitatively assess the reliability of national public health surveillance systems, providing a robust measure of data completeness and accuracy for informed decision-making. ", "methodology": "We developed a Bayesian hierarchical model to estimate surveillance system reliability, it \ (, ), where \ () and \ (\₈ₓ) are linear functions of time-varying covariates and structured temporal random effects. The model was fitted to longitudinal, district-level data on report completeness, timeliness, and verification from multiple notifiable conditions. ", "findings": "Posterior estimates revealed substantial spatial and temporal heterogeneity in reliability, with a pronounced decline in the predictive probability of high reliability (\ > 0. 8) in several provinces over the latter part of the study period. The model identified infrastructural covariates as key determinants, with the 95% credible interval for their combined effect excluding zero. ", "conclusion": "The proposed model provides a statistically rigorous tool for the continuous, integrated evaluation of surveillance system performance, moving beyond descriptive summaries to probabilistic inference. ", "recommendations": "We recommend the adoption of this modelling framework for routine surveillance evaluation to enable proactive identification of system weaknesses and targeted resource allocation. Future work should integrate climate and socio-economic data to refine predictions. ", "key words": "Bayesian inference, disease surveillance, health systems, hierarchical modelling, reliability, South Africa", "contribution statement": "This paper introduces a novel Bayesian hierarchical model for the integrated analysis of surveillance reliability, providing the first longitudinal, probabilistic assessment of South Africa's public health reporting systems
Dlamini et al. (Thu,) studied this question.