De-identified Clinical Administrative Dataset for Metabolic Disease Risk Analysis from RSI Banjarnegara, Indonesia

This dataset contains de-identified clinical administrative records of patients diagnosed with metabolic-related diseases, including diabetes mellitus, hypertension, and stroke, collected from Rumah Sakit Islam (RSI) Banjarnegara, Central Java, Indonesia. The dataset reflects patient management and service utilization characteristics rather than detailed physiological or laboratory measurements. All personal identifiers have been fully removed to ensure patient anonymity and compliance with ethical research standards. The dataset includes demographic attributes, service unit codes, case status, and diagnostic indicators, making it suitable for exploratory analysis, risk stratification, and machine learning–based predictive modeling in healthcare research. This dataset was primarily used for developing and evaluating artificial intelligence models (ANN, KNN, and hybrid ANN–KNN) for metabolic disease risk prediction under resource-constrained settings. Due to the scarcity of publicly available Indonesian clinical datasets, this dataset is expected to support future research on lightweight AI, health data analytics, and decision support systems in developing-country contexts. For further information please contact our PM (Mr. Ir. Rosyid) by [e-mail]