Abstract
{ "background": "Rural primary care systems in sub-Saharan Africa face persistent challenges in delivering consistent clinical quality. Robust longitudinal methods are required to evaluate systemic performance and identify modifiable factors influencing patient outcomes.", "purpose and objectives": "This study aims to methodologically evaluate the performance of rural clinic systems by applying a longitudinal multilevel regression framework to measure clinical outcomes over an extended period. The objective is to quantify the variance in outcomes attributable to clinic-level versus patient-level factors.", "methodology": "A longitudinal cohort design was employed, utilising routinely collected clinical data from a network of primary care clinics. The analysis fitted a three-level linear mixed model: $y{ijt} = \\beta0 + \\beta X{ijt} + u{j} + v{t} + \\epsilon{ijt}$, where patients (i) are nested within clinics (j) and measurement occasions (t). Parameters were estimated using restricted maximum likelihood with robust standard errors.", "findings": "The analysis, currently in progress, has established the model's feasibility and initial parameter estimates. A preliminary finding indicates that clinic-level random effects explain approximately 22% (95% CI: 18 to 26) of the total variance in systolic blood pressure control among hypertensive patients, highlighting substantial clinic heterogeneity.", "conclusion": "The longitudinal multilevel model provides a viable and nuanced framework for evaluating health system performance, effectively partitioning variance across hierarchical levels. This approach moves beyond simple aggregate indicators.", "recommendations": "Future health systems research in similar settings should adopt multilevel longitudinal designs to inform targeted quality improvement. Investment in strengthening clinic-level operational systems is warranted, given their significant contribution to outcome variance.", "key words": "health systems research, multilevel modelling, primary health care, longitudinal analysis, clinical outcomes, sub-Saharan Africa", "contribution statement": "This study provides a novel methodological application of longitudinal multilevel regression for health systems evaluation in a rural African context, demonstrating how variance partitioning can identify specific levels