Abstract
{ "background": "District hospitals are critical nodes in Nigeria's healthcare system, yet robust evidence on the effectiveness of systemic interventions remains sparse. Quasi-experimental designs (QEDs) offer a pragmatic approach for evaluating clinical outcomes in such complex, real-world settings where randomised controlled trials are often infeasible.", "purpose and objectives": "This systematic review aims to methodologically evaluate the application of quasi-experimental designs in studies measuring clinical outcomes within Nigerian district hospital systems, assessing their rigour, appropriateness, and reporting standards.", "methodology": "A systematic search of multiple electronic databases was conducted following a pre-registered protocol. Peer-reviewed studies employing QEDs (e.g., interrupted time series, difference-in-differences, regression discontinuity) to evaluate hospital-level interventions and report clinical outcomes were included. Study quality was appraised using the ROBINS-I tool. A meta-regression model, $Yi = \\beta0 + \\beta1X{1i} + \\beta2X{2i} + ui$, was specified to explore associations between design features and reported effect sizes, where $Yi$ is the standardised effect, $X{1i}$ is a rigour score, and $X{2i}$ indicates the use of propensity score matching.", "findings": "Of the screened records, 27 studies met the inclusion criteria. The methodological rigour was highly variable; a predominant theme was the inadequate handling of confounding and secular trends. Only a minority (approximately 26%) of studies employed robust methods like instrumental variables to address endogeneity. The meta-regression indicated a positive association between methodological rigour and more conservative effect estimates (95% CI: -0.41 to -0.12).", "conclusion": "The application of quasi-experimental designs in this context is increasing but often methodologically suboptimal, limiting the reliability of causal inferences about hospital system interventions on clinical outcomes.", "recommendations": "Future research should prioritise the use of more