Asian Jr. of Microbiol. Biotech. Env. Sc. Vol. 18, No. (4) : 2016 : 999-1001
© Global Science Publications
APPLICATION OF DATA MINING IN HEALTHCARE: A SURVEY
E. MERCY BEULAH 1, S. NIRMALA SUGIRTHA RAJINI 2 and N. RAJKUMAR3
Department of Computer Applications, Dr. M.G.R. Educational and Research Institute,
(Received 15 June, 2016; accepted 22 August, 2016)
Key words : Data Mining, Healthcare, Medical database, Healthcare services.
Abstract - Data Mining is one of the foremost motivating spaces for analysis that is mounting
progressively standard in the healthcare industry. Data mining plays an efficient role in revealing the
new emerging trends associated with this scenario. In the health industry, data processing provides
many advantages in transactional applications like Electronic Health Record (EHR), patient satisfaction
systems, lab systems, economic systems, patient identification etc. This survey highlights few
applications and future issues of Data mining in medical field. It also provides a picture of a database
which exists in health care organization.
*Corresponding author’s email : (1,3Assistant Professor, 2Associate Professor)
The growth of information technology has
generated a great deal of databases and large data
in numerous areas. The issue of health care believes
to be the prime importance for the society and is a
significant indicator of social development. In the
present era, Data Mining is becoming popular in the
healthcare field because there is a need of efficient
analytical methodology for finding unknown and
valuable information in health data (Divya Tomar
et al., 2013). The delivery of health care services thus
assumes the greater proportion, and in this context
the role played by information and communication
techniques has certainly a greater input for its
effective delivery mechanism.
The purpose of data mining is specifically
relevant and it has been successfully applied in
medical needs for its dependable precision accuracy
and expeditious beneficial results (Manaswini
The data generated by the health organizations
are very vast and complex due to which it is
difficult to analyze the data in order to make
important decisions regarding patient health. This
data contains details regarding hospitals, patients,
medical claims, treatment cost, medicine names,
lab details, phisican details etc. So, there is a need
to generate an important tool for analyzing and
extracting information from this complex
healthcare data. The analysis of health data
improves the healthcare of the patients by
enhancing the performance of patient management
tasks. The outcome of Data Mining technologies is
to provide profit to healthcare organization for
clustering the patients having similar type of
illness or health issues so that healthcare
organization provides them efficient treatments.
The patient’s length of stay in hospital can be
predicted and be planned for effective information
system management. With recent technologies
used in the medical field, enhances the medical
services in a cost effective manner. In analyzing the
various factors that are responsible for diseases, for
example, food habits, different working environ-
ment, and education level, living conditions,
availability of pure water, healthcare services,
environmental and agricultural factors data
mining techniques are used.
Data Mining In Healthcare
Data Mining came into existence in the middle of
1990’s and appeared as a powerful tool that is
suitable for fetching previously unknown pattern
1000 MERCY BEULAH ET AL.
and useful information from huge dataset. Various
studies highlighted that data mining techniques
helps the data possessor to analyze and discover
unsuspected relationships among their data which
in turn helpful in making decision. Data mining
techniques have been used intensively and
extensively by many organizations. In healthcare,
data mining is gradually increasing popularity, if
not by any case, becoming increasingly essential.
Data mining applications can greatly benefit all
parties involved in the healthcare industry
(Manaswini Pradhan, 2014). Healthcare industry
today generates large amounts of complex data
about patients, hospital resources, disease diagno-
sis, electronic patient records, medical devices etc.
The integrated medical database gives a clear
picture of the entire data that exist in any medical
The large amount of data is a key resource for
processing and analyzing of knowledge extraction
that enables support for cost-savings and decision
making. Data mining brings a set of tools and
techniques that can be applied to this processed
data to discover hidden patterns that provide
healthcare professionals an additional source of
knowledge for making decisions. Data mining is a
collection of algorithmic ways to extract
informative patterns from raw data. Data mining
is purely data-driven; this feature is significant in
health care (Nirmala et al., 2015).
Data mining can help Healthcare insurer detect
fraud and abuse, Healthcare organizations to make
customer relationship management decisions,
Physicians to identify effective treatments and best
practices, and for patients to receive better and
more affordable healthcare services.
DATA MINING APPLICATIONS IN
Key Dimensions in Healthcare Management is
Diagnosis and Treatment, Healthcare Resource
Management, Customer Relationship Management
and Fraud and Anomaly Detection (Prasanna
Desikan et al., 2011).
Diagnosis and Treatment
Ultrasound images supports examining of tumor
response to chemotherapy by Computer-assisted
texture analysis (Hub, 2000). Using data mining
techniques with magnetic resonance spectroscopy
diagnosis the presence of brain neoplasm (Zellner,
2004). To identify and quantify senile plagues,
analysis of digital images of tissue sections are
used in evaluating the severity of Alzheimer’s
disease. Data mining could be particularly useful in
medicine when there is no dispositive support
favoring a particular healing option. Based on
patients’ profile, history, physical examination,
diagnosis and utilize previous treatment patterns,
new treatment plans can be effectively recom-
Healthcare Resource Management
Using logistic regression models, hospital profiles
based on risk-adjusted death with 30 days of non-
cardiac surgery are compared. Neural network
system is used to predict the disposition in children
presenting to the emergency room with
bronchiolitis. Effectively control the resource
allocation by classifying high risk areas and
predicting the need and usage of various resources.
If the inpatient length of stay (LOS) can be
predicted efficiently, the planning and management
of hospital resources can be greatly enhanced.
Fitness report and demographic details of patients
is additionaly helpful for utilizing the market
hospital resources effectively.
Customer Relationship Management
CRM is built on a mixed view of the customer
across the whole association (Puschmann et al.,
2011). The view of the customers is fractured
about an enterprise; the view of the enterprise is a
splintered view of the customer. Kohli et al.,?????
demonstrate a web-based Physician Profiling
System (PPS) to build up relationships with
physicians and improve hospital profitability and
Some demographic characteristics and institu-
tional characteristics consistently have a signi-
ficant effect on a patient’s satisfaction scores.
Chronic illnesses (e.g. diabetes and asthma) require
self management and a collaborative patient-
physician relationship. The principles of applying
data mining for customer relationship management
also applicable to the healthcare industry. The
detection of usage and purchase patterns and the
eventual satisfaction can be used to improve
overall customer satisfaction. Patients, pharma-
cists, physicians and clinics are the customers.
Prediction of purchasing and usage behavior can
help to provide proactive ideas to reduce the
overall cost and improve customer satisfaction.
Application of Data Mining in Healthcare: A Survey
Fraud and Anomaly Detection
The avoidance and early detection of medical
insurance fraud, data mining has been used very
successfully. The ability to detect strange behavior
based on purchase, usage and other transaction
behavior information has made data mining a key
tool in a variety of organizations to detect
fraudulent claims, wrong prescriptions and other
abnormal behavior patterns.
The various frauds in healthcare industry can
be listed as: Prescription fraud (claims for patients
who do not exist ), Upcoding(claims for a medical
procedure which is more expensive or not
performed). The various fraud detection methods
are Neural networks, genetic algorithms and
nearest neighbor methods.
Data mining applications can more benefit all
parties involved in the healthcare industry. The
integrated healthcare information system will have
Fig. 1 Integrated Medical Database
the facility for finding the patient location based
and suggest the nearest emergency center, arrange
all necessary arrangements to be ready for the help
to provide proactive initiatives to reduce the
overall cost and increase customer satisfaction.
Divya Tomar and Sonali Agarwal 2013. A survey on
Data Mining approaches for Healthcare,
International Journal of Bio Science and BioTechnology.
5 (5): 241-266.
Manaswini Pradhan 2014. Data Mining and Health
Care: Techniques of Application. ISOI Journal of
Engineering and Computer Science. 1 (1): 18-26.
Nirmala Sugirtha Rajini,S 2015. Access Control in
Healthcare Information Management Systems
Using Biometric Authentication, International
Journal of Applied Environmental Sciences. 10 (1) : 143-
Prasanna Desikan, Kuo-Wei Hsu and Jaideep
Srivastava 2011. Data Mining for Healthcare
Management, International conference on Data
Mining, Arizona, USA.