Analysis of Diabetes Mellitus Patient Medical Records through Data Mining Implementation at Hospital Aloe Saboe Gorontalo

Abstract

This study aims to analyze the medical record data of Diabetes Mellitus patients through the application of data mining techniques to uncover patterns and information that can support decision-making in hospitals. The data used are secondary data derived from patient medical records at Hospital Aloe Saboe Gorontalo, encompassing variables such as age, sex, blood glucose level, disease history, and type of treatment. This research employs a quantitative approach, with the analysis following the Knowledge Discovery in Databases (KDD) process, which includes data selection, data cleaning, data transformation, the data mining process, and evaluation of the results. The data mining techniques applied consist of a classification method using Decision Tree and a clustering method using K-Means. The results show that the clustering technique was able to group patients according to similar characteristics, such as condition severity and risk of complications, while the classification method was able to predict patients' risk levels with a fairly good level of accuracy. Variables such as age, blood glucose level, and disease history were found to have a significant influence on patients' condition. These findings indicate a consistent pattern between risk factors and the severity of Diabetes Mellitus. In conclusion, applying data mining techniques to Diabetes Mellitus patients' medical records can generate useful information and support improved quality of healthcare services. The methods used proved effective in processing large volumes of data and in supporting more accurate and efficient decision-making