{ "background": "District hospitals are critical nodes in Rwanda's healthcare system, yet robust methodological frameworks for evaluating their clinical performance are lacking. Existing assessments often fail to account for the hierarchical structure of patient data and the inherent variability between institutions, limiting the validity of comparative analyses.", "purpose and objectives": "This review aims to critically evaluate the application of Bayesian hierarchical modelling as a methodological framework for assessing clinical outcomes across district hospital systems. It seeks to synthesise methodological strengths and limitations, and to propose a standardised analytical approach for health systems research.", "methodology": "A systematic search and narrative synthesis of literature on health systems evaluation and Bayesian statistical applications in low-resource settings was conducted. The core methodological proposition is a model where the log-odds of a positive clinical outcome for patient $i$ in hospital $j$, $y{ij} \\sim \\text{Bernoulli}(p{ij})$, is given by $\\text{logit}(p{ij}) = \\alpha + \\beta X{ij} + uj$, with hospital-specific random effects $uj \\sim N(0, \\sigma^2_u)$. Posterior credible intervals quantify uncertainty in institutional performance estimates.", "findings": "The review finds that this approach effectively partitions variance, distinguishing systemic hospital-level factors from case-mix variation. A key methodological finding is that models incorporating random effects yield more conservative and reliable performance rankings, with 95% credible intervals for hospital-level intercepts often spanning null effect, highlighting the uncertainty in labelling institutions as outliers.", "conclusion": "Bayesian hierarchical modelling provides a statistically rigorous framework for clinical outcomes assessment in district hospitals, offering advantages in handling multi-level data and quantifying uncertainty. It represents a superior alternative to crude, direct standardisation methods.", "recommendations": "Future health systems research in Rwanda and similar contexts should adopt hierarchical modelling to inform resource allocation and quality improvement. Investment in building local analytical capacity for these advanced statistical techniques is essential.", "key words": "health systems research, Bayesian statistics,