Public health surveillance systems are crucial for monitoring infectious diseases in Kenya. However, their reliability is often questioned due to potential biases and inefficiencies. A randomized field trial was conducted across three regions in Kenya. A binary logistic regression model with robust standard errors was used to analyse data collected from healthcare facilities. Systematic underreporting of cases was detected in one region, leading to a 15% underestimation of actual case numbers compared to reported figures. The randomized field trial demonstrated that public health surveillance systems can be improved by addressing identified weaknesses. Health authorities should prioritise training and standardization of reporting protocols across all regions to enhance system reliability. Treatment effect was estimated with $\text{logit}(p_i)=\beta_0+\beta^\top X_i$, and uncertainty reported using confidence-interval based inference.