A Quasi-Experimental Evaluation of Systems Optimisation for Improved Diagnostic Yield in Tanzanian District Hospitals

{ "background": "Diagnostic yield in sub-Saharan African district hospitals is often suboptimal, constrained by fragmented and under-resourced clinical systems. Systems optimisation interventions are frequently implemented, but robust methodological evaluations of their impact on diagnostic outcomes are lacking.", "purpose and objectives": "This study aimed to evaluate the impact of a multi-component systems optimisation package on diagnostic yield in a real-world hospital setting, using a quasi-experimental design to estimate causal effects.", "methodology": "A controlled interrupted time-series analysis was conducted across four matched district hospitals. Two hospitals received the intervention, which integrated laboratory workflow re-engineering, clinical decision support algorithms, and targeted staff training. The primary outcome was the monthly proportion of conclusive diagnoses from admitted patients. The effect was estimated using a generalised linear mixed model: \(\πit) = \β0 + \β1 Tt + \β2 Xit + \β3 (Tt \× Xit) + ui + \εit, where \πit is the probability of a conclusive diagnosis in hospital i at time t, $Tt$ is the post-intervention period, and Xit is the intervention indicator. Robust standard errors were clustered at the hospital level.", "findings": "The intervention was associated with a significant increase in diagnostic yield. The adjusted odds ratio for a conclusive diagnosis was 2.45 (95% CI: 1.78 to 3.37). The absolute improvement in the intervention hospitals was 18.7 percentage points (from a baseline of 54.3% to 73.0%) post-implementation, a change not observed in control sites.", "conclusion": "A structured systems optimisation package can substantially improve diagnostic conclusiveness in resource-constrained hospital settings. The findings demonstrate that synergistic adjustments to laboratory, clinical, and human resource subsystems can effectively mitigate diagnostic delays and uncertainties.", "recommendations": "Health policymakers