Discovering Primary Medical Procedures and their Associations with Other Procedures in HCUP Data
- PDF / 922,569 Bytes
- 15 Pages / 595.224 x 790.955 pts Page_size
- 93 Downloads / 173 Views
Discovering Primary Medical Procedures and their Associations with Other Procedures in HCUP Data Mamoun T. Mardini1,2
· Zbigniew W. Ra´s3,4
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract In recent years, healthcare spending has risen and become a burden on governments especially in the US. The selection of the primary medical procedure by physicians is the first step in the patient treatment process and is considered to be one of the main causes for hospital readmissions if it is not done correctly. In this paper, we propose a system that can identify with high accuracy the primary medical procedure for a newly admitted patient. We propose three approaches to anticipate which medical procedure should be primary. Additionally, we propose the procedure graph, which shows all possible paths that a new patient may undertake during the course of treatment. Finally, we extract the possible associations between the primary procedure and other procedures in the same hospital visit. The results show the ability of our proposed system to identify which procedure should be primary and extract its associations with other procedures. Keywords Hospital readmission reduction · Clustering · Personalization · Recommender systems
1 Introduction Recently, expenditure on healthcare has risen rapidly in the United States. According to Gorman (2013), healthcare spending has been rising at twice the rate of growth of our income for the past 40 years. The projection of the growth rate in healthcare spending is 5.8 percent during the period 2014-2024, which means that the spending will rise to 5.4 trillion by 2024. At the same time, the gross domestic product (GDP) growth rate is only 4.7 percent (as of 2014) (Keehan et al. 2015). This increase in healthcare Mamoun T. Mardini
[email protected] Zbigniew W. Ra´s [email protected] 1
Department of Aging and Geriatric Research, College of Medicine, University of Florida, Gainesville, FL 32610, USA
2
Department of Health Outcomes and Biomedical Informatics, College of Medicine, University of Florida, Gainesville, FL 32610, USA
3
College of Computing and Informatics, University of North Carolina, Charlotte, NC 28223, USA
4
Polish-Japanese Academy of Information Technology, 02-008 Warsaw, Poland
spending can be attributed to several factors as listed by Price Waterhouse Coopers (PWC) Research Institute: overtesting, processing claims, ignoring doctors’ orders, ineffective use of technology, hospital readmissions, medical errors, unnecessary emergency room (ER) visits, and hospital acquired infections (Coopers 2006). According to HCUP Statistical Brief published in 2014, all-cause hospital readmissions in the United States were associated with $41.3 billion in hospital costs (Hines et al. 2014). Hospital readmissions and surgery outcomes prediction has gained a great interest in the scientific community (Lally et al. 2017; Miotto et al. 2016; Silow-Carroll et al. 2011; Touati et al. 2014; Tremblay et al. 2006)). Analyzing the reasons behind readmissions and
Data Loading...
