{ "background": "District hospitals are critical nodes in the South African healthcare system, yet their operational reliability is poorly quantified. Current evaluations often lack methodological rigour in accounting for the hierarchical structure of health systems and the multifactorial nature of service delivery failures.", "purpose and objectives": "This protocol details a methodological framework to evaluate the reliability of district hospital systems. The primary objective is to develop and validate a multilevel regression model that identifies facility- and system-level determinants of reliability, measured as the consistency of core service provision against mandated standards.", "methodology": "We propose a retrospective, longitudinal analysis of administrative health data. System reliability will be operationalised as a composite score derived from indicators across four domains: clinical services, infrastructure, staffing, and supply chain. The core analysis employs a three-level hierarchical linear model: $Y{ijt} = \\beta0 + \\beta X{ijt} + u{j} + v{k} + \\epsilon{ijt}$, where $i$, $j$, and $k$ index time, hospital, and district, respectively. Parameter estimates will be reported with 95% confidence intervals, and model fit will be assessed using likelihood ratio tests.", "findings": "As a research protocol, this paper does not present empirical results. The anticipated analytical output will quantify the proportion of variance in reliability scores attributable to district-level systemic factors versus facility-specific characteristics. A key expected finding is the direction and magnitude of the association between supply chain integrity and overall reliability.", "conclusion": "This protocol provides a novel, statistically robust framework for assessing health system performance. Its application will yield a validated model that moves beyond descriptive analysis to inferential understanding of reliability drivers.", "recommendations": "We recommend the adoption of this multilevel modelling approach by health departments for routine system diagnostics. Future research should apply this framework prospectively and adapt it for other tiers of care.", "key words": "health systems research, multilevel modelling, hospital reliability, health services evaluation, South Africa