PENERAPAN METODE K-MEANS UNTUK MENGELOMPOKKAN REKAM MEDIS PASIEN BERDASARKAN DIAGNOSA PENYAKIT GUNA MENENTUKAN DIAGNOSA TERTINGGI PADA SUATU PERIODE (Study Kasus : Klinik Dokter Kita)

Abstract

The development of information technology in the digital era has driven major transformation in the health sector, especially in data management. One technique that plays an important role is data mining, which allows the discovery of hidden patterns and complex relationships in large amounts of data. This technique is very relevant for patient data analysis and disease diagnosis, especially in grouping patients based on disease type in order to understand distribution patterns and carry out more appropriate interventions. The K-Means method is an algorithm that is often used in the data grouping process, which allows identifying groups of patients with similar characteristics and helps determine the dominant disease in a certain period. At Our Doctor's Clinic, patient data continues to increase, but management is still manual so it is not optimal for in-depth analysis. The current grouping process based on disease diagnosis is general and makes it difficult for management to identify diseases that frequently appear in a certain period, which has an impact on difficulties in making strategic decisions. Therefore, This research applies the K-Means clustering method and RapidMiner Studio software version 10.2. with the aim of automating patient grouping based on diagnosis. The data used is patient data for 2023 which consists of 966 patients with a total of 1609 controls with 20 types of disease.  The results of the research show that there are three groups of disease diagnoses that occur frequently (highly dominant) and therefore require careful attention, namely the diagnosis of Grastitis, the diagnosis of ISPA and the diagnosis of Myalgia. Through this analysis, it is hoped that clinics can identify dominant diseases, understand distribution patterns, and increase the effectiveness of drug procurement planning and resource allocation. The results of this clustering are also expected to provide a basis for predicting future disease trends, allowing clinics to take preventive measures more proactively. Thus, it is hoped that the K-Means method can improve the quality of health services at Our Doctor's Clinic, make data-based decision making easier, and provide faster and more precise treatment for patients.