Use of Text Mining to Predict Patient Compliance

Overview

The purpose of this study is to examine standards of care in the Dental School of the University of Louisville. The central theme is to consider issue of compliance on behalf of the patients and how to define it in an unbiased way. The paper will examine the relationship of visit intervals, treatment needs, and patient compliance. With the use of SAS 9.1.3 software, data mining techniques such as clustering, kernel density, linear models and mixed models estimation will be used to define and analyze compliance. Statistical methods such as text mining were used to examine the severity of patient conditions. Confidence interval estimation and bootstrapping were explored to assist with the allocation of patients to compliance levels.