{ "background": "District hospitals are critical nodes in Senegal's healthcare system, yet their operational efficiency remains a persistent challenge. While quasi-experimental designs (QEDs) are increasingly employed to evaluate interventions aimed at improving efficiency, the methodological rigour and comparability of these studies require systematic assessment.", "purpose and objectives": "This systematic review aims to critically appraise the application of quasi-experimental methodologies in studies measuring efficiency gains within Senegalese district hospital systems, identifying common designs, analytical strengths, and limitations.", "methodology": "A systematic search of multiple electronic databases was conducted following PRISMA guidelines. Peer-reviewed studies employing QEDs (e.g., difference-in-differences, regression discontinuity, instrumental variables) to assess hospital efficiency outcomes were included. Studies were critically evaluated for design quality, confounding control, and statistical validity. A meta-analysis was precluded due to heterogeneity; findings were synthesised narratively.", "findings": "Of the 14 studies meeting inclusion criteria, difference-in-differences was the predominant design (n=9). A key theme was the frequent omission of necessary parallel trends testing and inadequate discussion of selection bias. One specific analysis, using a two-stage data envelopment analysis combined with a propensity score matching estimator $\\hat{\\tau}_{PSM} = E[Y(1)|D=1, p(X)] - E[Y(0)|D=0, p(X)]$, reported efficiency gains of 18% (95% CI: 12 to 24) from a management training intervention, though with concerns over unobserved confounders.", "conclusion": "The application of QEDs in this context is growing but methodologically inconsistent. Many studies fail to fully meet the key identifying assumptions of their chosen design, casting uncertainty on the robustness of reported efficiency gains.", "recommendations": "Future research should prioritise pre-registration of analysis plans, more robust sensitivity analyses (e.g., for hidden bias), and the use of stronger identification strategies,