Discovery Park Undergraduate Research Internship Program

"Predicting Modeling for the Risk of Medically Unnecessary Readmission to Acute Care"

About the Project

Project Time & Type:
Spring 2013 - DURI
Research area(s):
Statistics, Healthcare Engineering
Project Description:
The total US healthcare expenditure topped $2.5 trillion in 2009, more than 17.6% of the GDP, and is projected to increase further in the coming decade. A significant contributor to the skyrocketing spending is medically unnecessary readmission to acute care, which is avoidable with careful design of acute care admission and discharge protocols. For a careful design, a prerequisite is to understand the relationship between the readmission risk and the length of stay in the inpatient facility. In this project, we intend to design a predictive model for the readmission risk with the length of stay being the independent variable and patient characteristics being the classifiers. The approaches that may be explored include decision tree and neural net.
Expected Student Contributions:
Data analysis using SAS/R.
Related Website(s):
www.ics.purdue.edu/~nkong
Desired Qualifications:
GPA > 3.3; junior/senior students in Stats are preferred.
Estimated Weekly Hours:
6
Department awards independent research credits for this project?
Yes, 2 credit hours

Professor in Charge

Name:
Kong, Nan
Deptartment/College:
Biomedical Engineering

Student Supervisor

Name:
Nan Kong
Title:
Assistant Professor

Cooperating Faculty

Name:
Lingsong Zhang
Deptartment/College:
Statistics