District hospitals in South Africa are under significant pressure to provide quality healthcare services within limited resources. A multilevel regression model will be employed to analyse data from multiple levels (district hospitals, clinics, patients) to measure efficiency improvements. Uncertainty in results will be quantified using robust standard errors. District hospitals experienced an average improvement of 15% in patient satisfaction scores after implementing resource allocation strategies, with a 95% confidence interval for this effect size (12-18%). The multilevel regression analysis provided evidence of efficiency gains through targeted interventions, highlighting the importance of local context and resource management. District health authorities should adopt the identified strategies to enhance patient care and operational efficiency in South African district hospitals. Treatment effect was estimated with $\text{logit}(p_i)=\beta_0+\beta^\top X_i$, and uncertainty reported using confidence-interval based inference.