Public health surveillance systems are critical for disease control, yet robust methodological frameworks for their evaluation and for attributing health outcomes to their performance are lacking, particularly in resource-limited settings. This study aimed to develop and apply a novel Bayesian hierarchical model to evaluate the performance of integrated disease surveillance and response systems and to quantify their attributable risk reduction for key notifiable diseases. We conducted an intervention study analysing longitudinal surveillance data. The core model is specified as $y_{it} \sim \text{Poisson}(\lambda_{it})$, $\log(\lambda_{it}) = \alpha + \beta X_{it} + u_i + v_t + \epsilon_{it}$, where $y_{it}$ are reported cases, $X_{it}$ represents surveillance system intensity, and $u_i$, $v_t$ are structured random effects. Model parameters were estimated using Hamiltonian Monte Carlo, with inferences based on 95% credible intervals. The model estimated that a one-unit increase in surveillance system performance score was associated with a 15.2% (95% CrI: 11.8–18.4%) reduction in the risk of delayed outbreak detection. This effect was robust to adjustments for healthcare access and population density. The proposed Bayesian framework provides a rigorous, integrative tool for surveillance evaluation, demonstrating a quantifiable, significant protective effect of enhanced system performance on public health risk. Implement the model for routine, periodic surveillance system assessment. Prioritise investment in core system components linked to the strongest risk reduction, particularly timeliness and data completeness. Bayesian hierarchical model, surveillance evaluation, risk attribution, public health, intervention study, infectious disease This paper introduces a novel Bayesian model that jointly evaluates surveillance system performance and quantifies its causal effect on population health risk, a methodological advance beyond descriptive assessments.