Knowledge Discovery and Data Mining

Knowledge Discovery and Data Mining Melchor Abejon With the increasing use of electronic health records (EHR), data collection in many Health Care Organizations (HCOs) has improved and accumulated at a remarkable pace. EHRs have enabled HCOs to generate and collect a vast amount of data and information from their daily encounter of patients. And this is when Health Informatics (HI) comes into play- to develop and employ computational theories, tools and techniques that can assist in extracting useful information and knowledge from these volumes of data, and use this knowledge to uncover useful patterns and to create models that can enhance decision-making and processes in the healthcare and HI industry. This process is called the Knowledge Discovery and Data Mining (KDDM). The purpose of this paper is to give an overview on why KDDM is a necessity in the healthcare and HI industry, and also to discuss how the aforementioned technique continues to improve the healthcare and HI industry. Benefits of KDDM in the Healthcare Industry The abundance of data and information in healthcare had made KDDM a necessity. According to Taranu (2015), with the growing rate of data accumulation in HCOs, there is a need for expert analysis of these vast medical data. The ability to use these data in order to extract useful information is a key factor for many health institutions to establish quality healthcare. Listed below are some of the benefits of KDDM in the healthcare industry: Increase in the accuracy of diagnoses. The use of predictive algorithms can help healthcare providers such as the physicians to make diagnosis more accurate and decide for the appropriate treatment for their patients. Prediction of patient population risk. KDDM enables the creation of programs and risk models that can be employed to recognize and detect high-risk patients and chronic diseases. Because of this, healthcare providers are able to design the right clinical intervention for their patients. Reduction in the rate of hospital admissions and readmissions. The creation of algorithm and predictive analytics can enable the identification of patients who are at high risk for hospital admissions, thus enabling providers to design more efficient clinical interventions to better treat their patients. Enhancement of clinical decision support. KDDM enables the comparison of symptoms, causes, treatments and analysis of effective clinical intervention for a group of patients. Prevention of diseases and promotion of general public health. The application of predictive analytics particularly in genomics can aid physicians to recognize their patients who are highly at risk of certain diseases. Also, through predictive analytics, pharmaceutical companies are able to develop drugs that suit the needs for specific groups of people. Better patient-related decisions and patient satisfaction. The identification of usage patterns, preferences and the current and future needs of patients can give information that will assist staff in their interaction with their patients. Patients will also be happy because they will be receiving a treatment that will really work for them. Detection of medical insurance fraud and abuse. The detection of unusual claims patterns can assist insurance companies in the detection of medical insurance fraud and abuse. Benefits of KDDM in the Health Informatics Industry According to Shukla, Patel and Sen (2014), HI can be subdivided into four main subfields which are the (a) clinical care, (b) administration of health services, (c) medical research, and (d) education and training; and that each subfield and can be extended and improved with the application of KDDM. Clinical care. One of the applications of HI in healthcare is in the aspect of clinical decision-making through the implementation of the Clinical Decision Support System (CDSS). This computer program is designed to assist healthcare providers in making clinical decisions through (a) information retrieval, (b) alert systems, (c) reminders, (d) suggestion systems, and (e) prediction models. The application of KDDM techniques on the database will render health providers analytical tools and as well as predictive tools that go beyond what is evident from the surface of the data. For example, predictive models can assist physicians to decide whether a certain patient would be treated as inpatient or as an outpatient. Administration of health services. Being an administrator of a healthcare entity can be a very tough job as this position requires daily critical decision making. The quality of information that these critical decisions are based on is an essential factor for this job. The KDDM technique can help in the creation of systems that can predict disease outbreaks, and can give representation of the benefits and the costs of the different preventive measures that are effective against a disease outbreak. Medical research. The application of KDDM is very successful in the medical research. Data mining methods can be applied on the vast medical data to extract useful patterns, predictive scoring systems and cause and effect relationships. Nelson and Staggers (2014, p.56) states KDDM can also be used to patch weaknesses in clinical data that pose a barrier to research. Education and training. E-learning is one of the rapidly growing method of learning in the healthcare and even in the HI industry. The application of KDDM in e-learning can efficiently monitor the progress in the learning process and as well as enhance the learning experience of students, administrators, and educators by recommending different learning methods, resources, and study materials. For educators, it can provide objective feedback about the course and students learning patterns. For administrators, they can learn about the users behavior, so that servers can be optimized and network traffic can be distributed. Through KDDM, the effectiveness of educational programs can be efficiently assessed. Conclusion With the growing accumulation of data in healthcare, no wonder KDDM will continue to be an indispensable tool than can be utilized to extract knowledge and insightful patterns which are essential in the development of systems and models to improve the safety and quality of healthcare. And as the use of health information systems continue to grow, KDDM will continue to mend the weaknesses and imperfections in clinical data to make these data more usable for the benefit of the healthcare and HI industry. References Crockett, D., Johnson, R., Eliason, B. (2014). What is data mining in healthcare? Retrieved January 15, 2017, from https://www.healthcatalyst.com/wp-content/uploads/2014/06/What-is-data-mining-in-healthcare.pdf Fayyad, U., Shapiro, P., and Smyth, P. (1996). From data mining to knowledge discovering in databases. Retrieved January 15, 2017, from http://www.aaai.org/ojs/index.php/aimagazine/article/viewFile/1230/1131 Miner, L.A. (2014). Seven ways predictive analytics can improve healthcare. Retrieved January 15, 2017, from https://www.elsevier.com/connect/seven-ways-predictive-analytics-can-improve-healthcare Nelson, R. Staggers, N. (2014). Health Informatics: An Interprofessional Approach. St. Louis, MO.: Elsevier Mosby Shukla, D.P., Patel, S.P., Sen, A.K. (2014). A literature review in health informatics using data mining techniques. Retrieved January 15, 2017, from http://ijournals.in/ijshre/wp-content/uploads/2014/02/IJSHRE-2220.pdf Taranut, I.(2015). Data mining in healthcare: Decision making and precision. Retrieved January 15, 2017, from http://eds.b.ebscohost.com.lib.kaplan.edu/eds/pdfviewer/pdfviewer?sid=0be35910-eda5-4fda-a7c2-1a72f0962ec8%40sessionmgr102vid=2hid=114 The Modeling Agency (2015). How data mining is helping healthcare. Retrieved January 15, 2017, from https://the-modeling-agency.com/how-data-mining-is-helping-healthcare/