Association Rule Mining to Analyze Diagnosis Pattern Based on Electronic Medical Record Data

Abstract

Association Rule Mining (ARM) is a data
mining method used to find information about
items that often appear together. In this study, the
method is applied to look for diagnoses (primary
and secondary) that often occur together. The data
in this study are electronic medical record data taken
from hospital databases which are inpatient visits
during 2019. Data analysis process began with preprocessing
data, and continued using ARM method.
Based on the data analysis by the ARM method,
resulted an information that the diagnosis code E86
has assosiation with A09.9 with the confidence level
of 0.993.