Mining Federated Data (MFD) — A Conceptual Framework for Exploration and Evaluation of Hospital Performance Measures

Article type: Research Article

Authors: Shettian, Kruparaj Madhupal^{; *}

Affiliations: College of Health and Public Affairs, University of Central Florida, Orlando, FL, United States

Correspondence: [*] Corresponding author. Email: mshettian@knights.ucf.edu.

Abstract: This paper proposes an original mining federated data framework (MFDF) which can be used as a conceptual framework to perform exploratory and evaluation analysis of micro and macro level performance measures of hospitals. The framework uses the data mining techniques (statistical tools/machine learning) on enterprise data warehouse (EDW) platform that federates data for hospitals from multiple sources on a continual basis. This scalable and cyclic framework is flexible to test theories and analyze the impact of independent/predictor variables on dependent/response variables by deploying various data/statistical models. The paper presents a brief exploratory analysis performed based on this framework to understand the relationships between patient perceptions of care and hospital performance scores indicate that there is a positive correlation between patient satisfaction and hospital total performance scores; albeit the relationship of patient satisfaction scores with other domain scores such as safety, timeliness, effectiveness etc. cannot be ascertained.

Keywords: hospitals, performance measures, health IT, framework, analytics

DOI: 10.3233/jid-2017-0013

Journal: Journal of Integrated Design and Process Science, vol. 21, no. 3, pp. 33-45, 2017