The Future of Big Data in Healthcare
Over
the past few years, big data has been a significant buzzword because most
business organizations were unaware of how to tackle it. However, it is notable
that numerous sectors including healthcare have increased its application to
gather new insights that boost the
industry through the guarantee of actionable changes (Dinov, 2016).
Particularly in healthcare, there is an unprecedented amount of information on
patient care within a decade. It implies that today, data is being streamed
from almost all the aspects of an individual’s daily life.
1. Big
Data
Big
data refers to a complex and large data
sets that render traditional data processing methods and applications
inadequate or obsolete. Big data is the use of advanced methods of predictive
analytics to retrieve the value from
data, often to a specific data set size (Milicchio et al., 2016). Big data
accuracy yields confidence in decision making, which in turn results in
minimized risk and cost.
Health
monitors, mobile applications, fitness tracker, electronic medical records, and
the social media are sources of useful data and information. Notably,
significant efforts have been made in healthcare to digitize the records of
patients, especially in the developed countries. In the United States, the
government is advancing the big data agenda by providing a massive amount of healthcare data to the public
through government websites. As technology advances, new platforms and tools
emerge daily to help healthcare in leveraging the ever-expanding data sets in
new creative ways. Furthermore, as these tools evolve, healthcare professionals
will understand the current and future state of the sector from a wider
perspective (Sercar, 2013). Eventually, the move (together with a rapid
feedback) will potentially alter healthcare experience of patients and medical
professionals.
If
big data is leveraged properly, doctors will easily determine the level of
risks for chronic diseases such as cancer, diabetes, or obesity to provide
preventive healthcare measures. Additionally, big data has a potential to
change insurance reimbursement models, especially the best ways to reimburse
doctors (Selanikio, 2015). Big data can also radically alter patient
satisfaction metrics, planning, staffing, risk mitigation and identification, and
drug development.
Recent
statistics indicate that healthcare is traditionally lagging behind other
sectors because of the concerns about patients’ privacy and confidentiality.
However, if positive changes regarding big data in healthcare are observed in
the long run, it will result in better options to anticipate, treat, and
prevent illnesses. Moreover, physicians can make better decisions if data
mining tools are embedded within their workflow to facilitate the process
(McCandless, 2012).
2.
Data Mining
Data
mining is the extraction of hidden but predictive information from
significantly large databases. It is the advanced technique that helps
corporations to highlight the most useful and important data in their
information warehouses. In essence, data mining tools have a tendency to
predict future behaviors and trends hence
are handy in healthcare provision. A data mining tools answer business queries
traditionally considered as too time-consuming
and complex to resolve. In the modern times, doctors can implement data mining
techniques rapidly on existing hardware and hardware platforms to improve the
value of information resources (Provost, 2013). Besides, the system can be
integrated with the latest healthcare products found online.
A
data warehouse, on the other hand, is a platform that harbors all the corporate data in a normalized and centralized
form. The data is deployable at any time to users for decision support,
archiving, or executive reporting. In a physical sense, a data warehouse is an
important information repository required by the healthcare facilities to
thrive especially in an information age. Data mining technology allows the
management of medical facilities to focus on the most vital information in the
data gathered on patients’ behavior
(Estape et al., 2016). In fact, this technology discovers important information
within the data that reports and queries cannot reveal effectively.
3. Uses
of Big Data
It
is true that big data has advanced rapidly since 2000. People routinely collect
and generate a timeline of their activities by using technology to understand
and analyze information. The intersection of these modern trends (also referred
to as Big Data), helps businesses in all sectors to become more productive and
efficient. In specific, big data in healthcare is applied in the prediction of
epidemics, improvement of the quality of
life, and provision of a cure for
diseases (Doshi et al., 2016). Undeniably, the global population is expanding
remarkably because people are living longer. In the West, treatment delivery
models are changing at a rapid pace because data drive most of the decisions
behind these developments.
Software
developers have created apps for use in smartphones.
Owners of iPhones, for example, can install pedometers to measure how far they
walk in a day, or calorie counters to assist them in planning their diet.
Arguably, millions of people use mobile technology to improve their lifestyles
and health. In addition, there is a steady stream of wearable devices designed by
start-up tech firms like Samsung Gear Fit, Fitbit, and Jawbone that allow the
users to track their health progress and upload the data for comparison to that
of other device users. It is predictable that in the near future, patients can
share this data with doctors for use as part of the physician’s diagnostic
toolbox when visited by ailing individuals. In cases where the device user is
not sick, their ability to access enormous, ever-growing information database
on general public’s state of health ensures that the problems can be spotted
early before symptoms can show. Eventually, the healthcare professionals can
prepare and provide educational or medicinal remedies in advance (Kolb &
Kolb 2013).
A
new and essential use of big data is emerging, where data professionals and
medical practitioners partner to look into the future and identify problems
before they can happen. A classic example of this partnership is seen in
Pittsburg Heath Data Alliance (PHDA) that aims to collect data from insurance
and medical records, genetic data, social media, and wearable sensors
(Milicchio et al., 2016). Thereafter, the group uses the information to draw a
clearer comprehensive picture of a patient to provide a tailored package of
healthcare.
Furthermore,
IBM and Apple have collaborated to design a big data healthcare platform that enables Apple Watch and iPhone users to share
data using Wayson Health IBM cloud healthcare service. This project aims at uncovering the latest medical
insights from biomedical data and real-time
activity of millions of technology users.
Further,
telemedicine is gaining a wide acceptance, especially in urban areas. Patients
are able to receive medical treatment and assistance remotely, usually in their
own homes with the aid of internet connection and a computer. There are
hundreds of websites with useful data for self-diagnosis
at the moment. Besides, ailing individuals can receive online medical services
through one-on-one sessions with qualified healthcare professionals.
Essentially, interactions such as these leave a data trail that can be analyzed
to improve future services. Then, the valuable information can be integrated
into the general public health trends to boost the way people access
healthcare.
Medicines
treat various diseases affecting the patients. What many people do not know is
that big data is used to design the portions and pills provided to the sick.
Significant amounts of data and information on applicant enable medical
researchers to select the best subjects. Commonly, data sharing arrangements
between global pharmaceuticals lead to a breakthrough
in healthcare. For example, with the help of big data and data sharing, it was
discovered that desipramine, often used as an anti-depressant, can cure acute
lung cancer and other chronic illnesses.
Medical
data is personal, hence the need for extreme security and exclusive
accessibility to a personal physician, and the patient. However, the internet
and online platforms crawl with cyber thieves that target confidential medical
records for malicious use. Many of them earn more from stolen health data than
from hacked credit card information.
4. Statistical
Ways to Analyze Big Data
The
data sets in big data are mostly too big and complex to process solely on
database management tools. Medical records, large-scale
e-commerce data, and military
surveillance are examples big datasets that require more than one terabyte of
storage. One of the statistical ways to analyze big data is the use of
association rule learning. It is a statistical method used to discover
intriguing correlations between large databases and variables. Major
supermarket chains in the United States applied this method to discover
interesting links among the products using the business data obtained from the
POS supermarket systems. Particularly, association rule learning is beneficial
in the analysis of biological data to discover new relationships, the
extraction of information on internet web visitors from the log in information, and monitoring of internet
logs to track malicious activities such as phishing or hacking.
Classification tree analysis is a statistical method
applied to the identification of specific
categories for new observations. The historical
data set is required to ensure consistency and accuracy of the statistical
outcomes. In the modern times, statistical tree method is beneficial in the
categorization of organisms into new groups, automatic assignment of documents
to new classifications, and the development of the profiles of patients or
students using online platforms.
Genetic algorithms are inspired by the work of evolution
through natural selection, inheritance,
and mutation. Therefore, this mechanism is used in evolving constructive
solutions to issues that demand statistical optimization. In this way, the
technique allows for scheduling of physicians for emergency medical rooms.
Regression
analysis, at a basic level, involves manipulation of independent variables to
observe their influence on dependent variables. Essentially, this statistical
method describes in detail how the dependent variable data value alters with
the variation of the independent
variable. Statisticians recommend the use of this technique for continuous
quantitative big data such as patient’s age or weight.
Social
network analysis was adopted first in the telecommunications industry. Later,
sociologists used the statistical technique to analyze interpersonal relations.
Currently, the method has gained a wide acceptance, especially in the
commercial and healthcare sector, where it is used to study the relationships
between people and how it affects their well-being
and health. While nodes represent persons within a given network, ties are
representative of individual relationships.
5.
Critique of Big Data and Data Mining
Even though big data effectively detects subtle
correlations often missed by the analysis of smaller data sets, it cannot
explain the reasons for the correlations or their
meaningfulness. Secondly, big data are useful in scientific inquiries but are
unsuccessful as a replacement of traditional techniques. For instance,
molecular biologists are interested in the inference
of 3D protein structure of underlying DNA sequences, which begs the need for
the use of big data as one of the numerous available scientific tools. However,
big data and data mining cannot be independently relied on to crunch complex
genome data to solve the problem, irrespective of how effective these
statistical methods are. In fact, there is always a need for basic knowledge and understanding of
biochemistry and physics.
Data mining is associated with the unauthorized
collections of information on the behavior
of people, hence infringement on personal privacy and ethics. Mined data can be
applied in different contexts and instances that beg the questions of legality.
In the past, governments through federal agencies such as NSA and CIA have gathered
data for law enforcement or national security purposes, thus raising concerns
and fears among the citizens. Furthermore, data aggregation (introduced by data
mining) poses a threat to a subject’s privacy because the data miner has an
access to new but previously anonymous sets of compiled data.
6. Ideas
for Future Applications of Big Data in Healthcare
In
future, patients that battle multiple health problems will access care more
frequently. However, the utilization of big data will lead to a dramatic slump
in the cost of healthcare because the doctors will have the patient’s
information in the EHR. The data systems, once fully operational and connected,
will notify the providers of problematic trends and data errors for swift
intervention. Eventually, savings on late interventions will be attained and
the funds can be redirected towards boosting the innovation and system
functionality. As information on patients is gathered over time, the system
operators have an opportunity to learn and improve the provision of healthcare
in future.
References
Dinov, I. D. (2016). Volume and value of big
healthcare data. Journal of Medical Statistics and Informatics J Med Stat
Inform, 4(1), 3. doi:10.7243/2053-7662-4-3
Doshi, J. A., Hendrick, F. B., Graff, J. S.,
& Stuart, B. C. (2016). Data, Data Everywhere, But Access Remains a Big
Issue for Researchers: A Review of Access Policies for Publicly-Funded
Patient-level Health Care Data in the United States. EGEMs (Generating Evidence
& Methods to Improve Patient Outcomes), 4(2). doi:10.13063/2327-9214.1204
Estape, E. A., Mays, M. H., & Sternke, E. A.
(2016). Translation in Data Mining to Advance Personalized Medicine for Health
Equity. IIM Intelligent Information Management, 08(01), 9-16.
doi:10.4236/iim.2016.81002
Kolb, J., & Kolb, J. (2013). The big data
revolution: The world is changing, are you ready? Chicago, IL: Applied Data
Labs.
McCandless, David. The beauty of data
visualization. (n.d.). Retrieved May 28, 2016, from http://www.ted.com/talks/david_mccandless_the_beauty_of_data_visualization?language=en
Milicchio, F., Rose, R., Bian, J., Min, J.,
& Prosperi, M. (2016). Visual programming for next-generation sequencing
data analytics. BioData Mining, 9(1). doi:10.1186/s13040-016-0095-3
Provost, F. (2013). Data Science for Business:
What You Need to Know about Data Mining and Data-Analytic Thinking.
Selanikio, J. (2015). The surprising seeds of a
big-data revolution in healthcare. (n.d.). Retrieved May 28, 2016, fromhttp://www.ted.com/talks/joel_selanikio_the_surprising_seeds_of_a_big_data_revolution_in_healthcarehttps://www.ted.com/talks/joel_selanikio_the_surprising_seeds_of_a_big_data_revolution_in_healthcare.
Sercar, T. (2013). Viktor Mayer-Schnberger and Kenneth Cukier, Big Data: A
Revolution That Will Transform How We Live, Work, and Think. OZ Organizacija
Znanja, 18(1-4), 47-49. doi:10.3359/oz1314047
No comments:
Post a Comment