A design science framework for research in health analytics

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Caleb Bradberry (Creator)
The University of North Carolina at Greensboro (UNCG )
Web Site: http://library.uncg.edu/
Rahul Singh

Abstract: Data analytics provide the ability to systematically identify patterns and insights from a variety of data as organizations pursue improvements in their processes, products, and services. Analytics can be classified based on their ability to: explore, explain, predict, and prescribe. When applied to the field of healthcare, analytics presents a new frontier for business intelligence. In 2013 alone, the Centers for Medicare and Medicaid Services (CMS) reported that the national health expenditure was $2.9 trillion, representing 17.4% of the total United States GDP. The Patient Protection and Affordable Care Act of 2010 (ACA) requires all hospitals to implement electronic medical record (EMR) technologies by year 2014 (Patient Protection and Affordable Care Act, 2010). Moreover, the ACA makes healthcare process and outcomes more transparent by making related data readily available for research. Enterprising organizations are employing analytics and analytical techniques to find patterns in healthcare data (I. R. Bardhan & Thouin, 2013; Hansen, Miron-Shatz, Lau, & Paton, 2014). The goal is to assess the cost and quality of care and identify opportunities for improvement for organizations as well as the healthcare system as a whole. Yet, there remains a need for research to systematically understand, explain, and predict the sources and impacts of the widely observed variance in the cost and quality of care available. This is a driving motivation for research in healthcare. This dissertation conducts a design theoretic examination of the application of advanced data analytics in healthcare. Heart Failure is the number one cause of death and the biggest contributor healthcare costs in the United States. An exploratory examination of the application of predictive analytics is conducted in order to understand the cost and quality of care provided to heart failure patients. The specific research question is addressed: How can we improve and expand upon our understanding of the variances in the cost of care and the quality of care for heart failure? Using state level data from the State Health Plan of North Carolina, a standard readmission model was assessed as a baseline measure for prediction, and advanced analytics were compared to this baseline. This dissertation demonstrates that advanced analytics can improve readmission predictions as well as expand understanding of the profile of a patient readmitted for heart failure. Implications are assessed for academics and practitioners.

Additional Information

Language: English
Date: 2016
Analytics, Healthcare, Information systems
Health services administration $x Decision making $x Mathematical models
Medical informatics $z North Carolina
Heart failure $x Treatment $z North Carolina

Email this document to