Professional Documents
Culture Documents
BY
STEVE DENSON
CHASE BLAND
ADNAN DOHADWALA
11/28/2016
Sincerely,
Steve Denson
Chase Bland
Adnan Dohadwala
support as Lee puts it. Implementing this capital allocation strategy has
allowed Adventist HealthCare to plan capital projects around financial
performance over time, thus facilitating a massive capital investment in data
analytics over a 3-year period of time.
Capitalization and implementation of EMR/Big Data has been a highlevel priority for hospitals as according to the Congress Adoption of HIT (Feb.
2015), 97% of nearly all hospitals and 75% of physicians have certified EMR
as of 2014. One of the biggest objectives for most healthcare providers will
be establishing future proofing and sustainability in terms of Big Data. As the
utilization of AI, IoT, digitization, etc. continues to expand, real-time Big Data
generated from these technologies is expected to explode from 4.4
zettabytes in 2015 to 44 zettabytes in 2025 as predicted by IDC Directions
Conference 2016. Despite the massive potential for a 10-fold increase in the
amount of generated data over the next 10 years, there is a silver lining
within Big Data technologies that will affect the future cost, velocity, and
storage concerns. Enter community-driven software solution called Apache
Flume. Apache Flume software was developed as a solution to effectively
gather, sort, and allocate massive amounts of data. How Apache Flume will
decrease costs of managing Big Data is by: 1. Its Flexible Customizable to
work with various types of systems such as batch computation systems like
Hadoop and non-structured databases like MongoDB as well as integrating
with different data pipelines. 2. Reduced Processing Time compared to
older data collector systems, Apache Flumes performance in collecting,
transferring, and storing Big Data is currently one of the fastest in the
industry.
* Separate Section * Major hospitals and healthcare systems are
implementing IT technologies such as IMB Watson to develop predictive
modeling/analytics for the development of precision medicine and other
innovative programs. These institutions will need cost efficient efforts to
handle the continuous and increasing quantities of real-time data being
gathered, transmitted, categorized and stored in the amount of data being
generated, categorized, and stored
IBM Watson Health is one of the pioneering AIs being used in health
care. Watson uses natural language processing and semantic computing
abilities to train in clinical decision support. Watson can ingest millions of
pages of academic literature and other health care data. It then offers
decisions to providers along with confidence intervals that show how
applicable the course of action may be. (8) The outcome is a computer that
can learn at a rate humans are incapable of and a quality of output that is far
more accurate due to this consumption of massive amounts of data.
Additional benefits from AI in the health care industry include
increased patient safety, earlier and more accurate diagnoses, and fewer
missed opportunities to deliver care based on recommended protocols. (1)
Safety, governance, and regulation are among the challenges standing
in the way of achieving the potential of AI. Guidelines will need to be
implemented that require AI to be thoroughly tested and vetted before being
exposed to the public. (1)
Blockchain:
An interconnected health care industry where data and records are
shared amongst patients and physicians is an achievable goal when Big
Datas advantages are harnessed. This is easier said than done when the
amount of data is overwhelming and more is pouring in every day.
Blockchains is an emerging technology that can be utilized to help create the
of Things produced too much data to analyze efficiently. (5) The poll also said
it was difficult to capture data reliably and that a lack of analytical
capabilities was an issue when attempted to extract insights from the data.
Wearable devices will continue to provide data directly to
patients and physicians. A survey by Quest Diagnostics and Inovalon found
87% of respondents acknowledge the importance of having access to a
patients entire medical record. (7) Providers are dissatisfied with how much
data they can access when sitting in front of a patient. 64% said they do not
routinely have all the data they need about their patients at the point of
care. This is a problem that wearable devices and the IoT can rectify if the
analytical potential of this data is achieved. Ultimately having data available
quickly benefits both sides of health care. Providers can make better
decision, while patients will be more satisfied with the level of care.
Expertise:
In healthcare system today the value for big data is largely limited to
research and development part of it, because using big data requires a very
specialized set of skill. Hospital IT experts who are familiar with SQL
programming languages and traditional relational databases are not very
much comfortable and prepared for a very steep learning curve and other
complexities associated big data. (Adamson, 2014)
As a matter of fact most of the organizations needed a data
expert in order to manipulate and get data out of a big data environment.
These data experts are usually a Ph.D. level thinkers with significant
expertise and typically it is very difficult to get a hang of them , theyre not
just floating around an average health system. Only research institutions
usually have access to them because they are hard to come by and
expensive. They are also recognized as Data scientists and these are in huge
demand across industries like banking and internet companies with deep
pockets. (Adamson, 2014)
The good news for such industries is, thanks to editing done with the
tooling, people with less-specialized skills will now be able to easily work with
big data in the near future. Big data is coming to embrace SQL as the lingua
franca for querying. And when this will happens, it will become useful in a
health system setting.
Microsofts Polybase is one of the example of a query tool which will enables
users to query with both Hadoop Distributed File System (HDFS) systems and
SQL relational databases using an extended SQL syntax. There are other
tools such as Impala, which enable the user of SQL over a Hadoop database.
With the help of such types of tools a larger group of user will come close to
big data. (Fennessy, 2016)
Security:
In a healthcare system, HIPAA compliance is non-negotiable. The privacy and
the security of data of the patient is the most important of all, But, as a fact
there are not many effective and integrated ways to manage security in big
data. Although security is coming along, it has been an afterthought up to
this point. And for a good reason. If a hospital only has to give access of their
data to only a few data scientists then it really is nothing much to worry
about. But when they are opening up such fragile data access to a large,
diverse group of users, security cannot be an afterthought.
To secure the big data the healthcare organizations will have take some of
the major steps. Big data runs on a open source technology with inconsistent
security . So to avoid security threats , organizations should be very specific
about big data vendors and avoid assuming that any big data distribution
they select will be secure. (Adamson, 2014)
A well -supported and commercially distributed form of implementation of
big data is by far the best option for the organization planning to adopt big
data, rather than going for a new raw apache distribution. An example of a
company with a well supported, secure distribution is Cloudera. This
in retrieving the desire data and needs a specialist to do that job. (Adamson,
2014)
There are architecture like Late-Binding enterprise data warehouse (EDW)
that is making the transition from relational databases to unstructured big
data. It takes data from source systems (EHRs, financial systems, etc.) are
placed into source marts. It follows the principles of Big Data to keep the
data as raw as possible, relying on the natural data models of the source
systems. As much as possible, late-binding methods minimize remodeling
data in the source marts until the analytic use case requires it. The data
remains in its raw state until someone needs it. At that point, analysts
package the data into a separate data mart and apply meaning and semantic
context so that effective analysis can occur. (Barlow, 2013)
In reality today, there are many projects going on to create a massively
parallel data warehouse, which can run a traditional relational database and
a big data cluster in parallel to stores data simultaneously, which would
improve the data processing power significantly. So, the progression from
todays symmetric multiprocessing (SMP) relational databases to massively
parallel processing (MPP) databases to big data in healthcare is underway.
References:
Fennessy Josh, 2016, " 5 Reasons to Get Excited About SQL Server
2016 and Big Data", Blue Granite, retrieved from: https://www.bluegranite.com/blog/5-reasons-to-get-excited-about-sql-server-2016-and-bigdata