IMPLEMENTASI DATA MINING DALAM KLASIFIKASI HASIL DIAGNOSA PASIEN BPJS MENGGUNAKAN ALGORITMA CART

Abstract

Puskesmas is one of the BPJS patient referral health facilities that can generate and collect a lot of medical record data every day. The pile of medical record data is generated from routine operational activities and is also used for operational needs. However, sometimes the heap of data is left unattended and unused. So far, Puskesmas Umbulharjo 1 has used this data only to make policies in the form of health education to the public, by providing information on the number of patients seeking treatment, types of illness, and reports of patient discharge. However, the pattern of disease tendencies suffered by the community has not been explored further, so that it can be used as a reference in health education so that it is more targeted. Therefore, this study aims to classify the diagnosis results of BPJS patients based on the relationship between symptoms and types of disease using the CART algorithm. The data used is secondary data obtained from medical records of the results of the examination of BPJS patients at the Umbulharjo 1 Health Center as much as 200 data. Based on the results of the analysis carried out, the CART classification shows that the types of TB, pneumonia, gastritis, and hypertension are influenced by symptoms of headache, abdominal pain, nausea, and vomiting. The results of the accuracy of the classification accuracy using the CART method of 71.5%.