Abstract
{ "background": "Community health centres are a cornerstone of primary healthcare delivery in Ethiopia, yet systematic evaluations of the methodological rigour used to assess their operational reliability are lacking. Understanding the analytical frameworks employed is crucial for strengthening evidence-based health systems policy.", "purpose and objectives": "This systematic review aims to critically appraise and synthesise methodological frameworks from published studies that assess the reliability of community health centre systems in Ethiopia, with a specific focus on the application and reporting of multilevel regression analysis.", "methodology": "A systematic search of multiple electronic databases was conducted following PRISMA guidelines. Peer-reviewed studies employing quantitative methods to evaluate system reliability were included. Studies were screened, and data on methodological design, model specification, and statistical reporting were extracted. Quality assessment was performed using a modified tool for observational health systems research.", "findings": "Of the 42 studies meeting inclusion criteria, only 19 (45%) explicitly specified a multilevel model. A predominant theme was the inconsistent handling of clustered data, with many studies using single-level regression, potentially underestimating standard errors. The most common specified model was a two-level random intercept logistic regression, $\\logit(p{ij}) = \\beta0 + \\beta X{ij} + u{j}$, where $u{j} \\sim N(0, \\sigma^2u)$, though reporting of variance partition coefficients was frequently omitted.", "conclusion": "There is significant methodological heterogeneity and frequent underuse of appropriate multilevel modelling techniques in research assessing health centre reliability. This undermines the validity of inferences regarding system performance and the factors influencing it.", "recommendations": "Future research must prioritise the a priori specification of multilevel models where data are hierarchically structured. Journals should enforce stricter reporting standards for variance components and model diagnostics. Capacity building in advanced statistical methods for health systems researchers is urgently needed.", "key words": "health systems research, multilevel modelling, methodological review, primary healthcare, statistical reliability, health centre performance",