Abstract
{ "background": "Community health centre (CHC) systems are critical for primary care delivery, yet robust methodological frameworks for evaluating their clinical performance are underdeveloped. Existing analyses often fail to account for the hierarchical structure of patient data nested within facilities, risking biased inference.", "purpose and objectives": "This case study aims to methodologically evaluate the application of multilevel regression for measuring clinical outcomes within a CHC system. Its objective is to demonstrate the model's utility in partitioning variance and identifying facility-level predictors of patient outcomes.", "methodology": "We conducted a secondary analysis of anonymised, longitudinal patient records from a stratified sample of CHCs. A two-level random intercepts model was specified: $y{ij} = \\beta{0} + \\beta{1}X{ij} + u{j} + e{ij}$, where $y{ij}$ is the outcome for patient $i$ in facility $j$, $X{ij}$ are patient-level covariates, $u{j}$ is the facility-specific random effect, and $e{ij}$ is the individual error. Estimation used restricted maximum likelihood with robust standard errors.", "findings": "The methodological evaluation revealed that approximately 18% of the variance in glycaemic control (HbA1c) was attributable to facility-level differences. A one-unit increase in facility-level staffing adequacy was associated with a clinically meaningful reduction in mean HbA1c (β = -0.45, 95% CI: -0.68 to -0.22), after controlling for patient demographics and comorbidities.", "conclusion": "Multilevel regression provides a statistically sound framework for analysing hierarchically structured clinical data from CHCs, offering insights into system performance that single-level models obscure.", "recommendations": "Health systems researchers should adopt multilevel modelling for routine performance evaluations. Policymakers should mandate the collection of facility-level covariates to enable such analyses and target resource allocation towards modifiable facility factors, such as staffing.", "key words": "mult