Semester of Graduation

Fall 2018

Degree

Master of Science in Industrial Engineering (MSIE)

Department

Mechanical and Industrial Engineering

Document Type

Thesis

Abstract

Asthma is associated with frequent use of primary health services and places a burden on the United States economy. Identifying key factors associated with increased cost of asthma is an essential step to improve practices of asthma management.

The aim of this study was to identify factors associated with over utilization of primary health services and increased cost via claims data and to explore the effectiveness of case management program in reducing overall asthma related cost.

Claims data analysis for Medicaid insured asthma patients in Louisiana was conducted. Asthma patients were identified using their ICD-9 and ICD-10 codes, forward variable selection was used to identify significant factors in the regression model with total cost as the dependent variable, multivariate regression was used to identify patients’ factors associated with frequent utilization of primary health services, and finally, a T-test was used to compare the difference in cost over time for case managed and non-case managed patients.

Cost of four claims categories was significant to the total cost variable: primary physician visits, pharmacy prescriptions, emergency room visits and urgent care clinics visits. Median income and enrollment in case management were significant in predicting number of emergency room visits. Patients who had higher income were more likely to utilize urgent care clinics. As a side finding, this study built a prediction model for total cost, the linear regression model accuracy was compared to neural networks and the proposed threshold point in which neural network outperforms the regression model is around 6,000 data points.

Patients with a history of utilization of certain health services are more likely to need case management for better health outcomes and controlled cost. future work is to perform analysis on a larger scale and include more patients related factors to identify a more holistic definition of high-risk patients.

Date

10-10-2018

Committee Chair

Nahmens, Isabelina

Available for download on Friday, October 11, 2019

Share

COinS