Abstract
District hospitals in sub-Saharan Africa face systemic inefficiencies that constrain service delivery and agricultural yield outcomes linked to community nutrition. Current evaluations often lack rigorous causal designs to isolate the effect of specific systemic interventions on measurable outputs. This protocol details a randomised field trial to evaluate the impact of a structured systems optimisation bundle on operational yield—defined as the composite of patient throughput and agricultural-nutritional programme efficiency—within the Kenyan district hospital context. We will conduct a cluster-randomised controlled trial across 24 facilities, stratified by region and baseline capacity. Intervention facilities will implement a co-designed bundle comprising logistics management, clinical pathway streamlining, and integrated agricultural extension linkage. The primary outcome is the yield improvement score (YIS). Analysis will use a mixed-effects model: $Y{it} = \beta0 + \beta1 Ti + \gamma X{it} + ui + \epsilon{it}$, where $Ti$ is treatment assignment, with inference based on cluster-robust standard errors. As a protocol, no empirical results are presented. The anticipated primary analysis will test the hypothesis that the intervention increases the mean yield improvement score by a minimum of 15 percentage points relative to control facilities. This protocol provides a methodological framework for robustly quantifying how targeted systems engineering impacts multifactorial yield in resource-constrained hospital settings, with potential applicability across similar food system interfaces. Future research should adopt similar experimental designs to disaggregate the effects of individual system components. Policy mechanisms should integrate routine yield metrics into hospital performance assessments. health systems research, operational research, randomised controlled trial, health-agriculture linkage, service delivery, sub-Saharan Africa This protocol introduces a novel composite yield metric and a causal experimental design to evaluate hospital system interventions, directly generating evidence for optimising the health-agriculture nexus.