Abstract
The increasing burden of pediatric health risks in urban Pakistan, driven by climate variability, infectious disease outbreaks, and adverse drug reactions (ADRs), necessitates advanced and integrated surveillance systems. Traditional pharmacovigilance and disease surveillance mechanisms remain fragmented, reactive, and insufficient for real-time decision-making. This study proposes an AI-Integrated Pharmacovigilance and Climate-Sensitive Disease Surveillance Framework designed to enhance early detection, prediction accuracy, and interpretability of pediatric health risks. The framework integrates heterogeneous data sources, including clinical records, pharmacovigilance reports, and environmental indicators such as temperature, humidity, rainfall, and air pollution. Machine learning and deep learning techniques were applied for predictive modeling, while Explainable Artificial Intelligence (XAI) methods were incorporated to ensure transparency and interpretability of outcomes. The model was evaluated using performance metrics and expert validation. Results indicated that the proposed framework significantly outperformed baseline models in accuracy, precision, recall, and false-risk detection, while also demonstrating strong correlation between climate variables and disease incidence. Expert assessments confirmed high levels of trust, usability, and interpretability of the system. The study concludes that integrating AI with pharmacovigilance and climate-sensitive surveillance provides a robust, scalable, and transparent solution for strengthening pediatric healthcare systems in climate-vulnerable urban environments.