Pengelompokkan Data Rekam Medis Pasien Berdasarkan Jenis Dengan Algoritma K-Means (Studi Kasus Puskesmas Parongpong)

Abstrak

Medical records, also known in the health sector as ICD (International Classification of Diseases), are records of patients’ medical histories during treatment in hospitals or clinics. Parongpong Community Health Center is one of the healthcare facilities in West Bandung Regency that receives thousands of patient visits each year with various diseases. However, the available information system, such as SIMPUS, has not been able to provide analytical information in the form of disease groupings that could be utilized for health service planning. This study aims to cluster patient medical record data based on disease types using the K-Means algorithm with the CRISP-DM approach as the stages of data mining analysis. The dataset consisted of 16,366 patient records from the General Clinic of Parongpong Community Health Center in 2023. The data included five main attributes: medical record number, gender, age, region of origin, and disease name. The clustering process was carried out using RapidMiner, and the best results were achieved when the data were grouped into three clusters based on a Davies Bouldin Index value of -0.826. The findings show that Cluster 0 was dominated by influenza (Acute Nasopharyngitis), commonly affecting toddlers; Cluster 1 was dominated by dyspepsia and hypertension, mostly found in adults and the elderly; while Cluster 2 was dominated by acute respiratory infections (ARI), diarrhea, and scabies, which frequently affect children and adolescents. These groupings provide a clear overview of disease distribution by age, gender, and region, and can serve as a basis for decision-making to improve preventive and curative health services at the community health center.