Written
Assignment 8: Data Mining
Alanna
M. Schauer
Thomas
Edison State College
Written
Assignment 8: Data Mining
Information Technology Manager Interview
The
writer interviewed the Clinical Application Manager called B.K. to gather the
information needed to answer the following questions. The author was surprised
to receive the answers given because the hospital has such foresight when it comes
to Information technology and yet this was not the case as it pertained to data
mining
AMS- "What software are we
currently using for data mining?
B.K. - "We do not have great data mining tools established
currently but we are working on it. We are rolling out some new projects that
will address the data mining issue but we are at the beginning stages. We are
focusing on the Quality Core Measures initiatives and utilizing software called
Focus by SpectraMD USA, Inc. (Are
you maximizing, 2012). This program has
dashboards that allow monitoring of the information and the ability to see trending.
This software focuses on the value based purchasing initiatives and allows you
to drill down to specific data points to see where there is missing information.
This allows the hospital to increase performance across the continuum of care
and gain financial incentives."
AMS- "Who is currently using the
Focus program?"
B.K. -"We are really targeting the quality issues and CHF
readmissions so the quality assurance department is the primary users of this new
software program. We will build this over our Siemens Decision Support System
(DSS). K.A. the VP of Quality Assurance is the person who really decides what data
targets we need to focus on. Our Congestive Heart Failure (CHF) readmission
rates are high, higher than the administration wants and we need to work fast
to reduce them."
AMS- "What are they trying to predict
with the collection of the data; the CHF values?"
B. K. -"The Focus program uses predictive and analytic statistical
techniques as a primary tool for data mining core measure information. When
they created this program, they created a panel of cardiologists for their
input so they could capture the necessary data points for a successful program.
For example, with the CHF measure, the ejection fraction is a good indicator of
cardiac muscle function but they need a clean data point. When the
echocardiogram calculates the ejection fraction, there is one value. In contrast, when the cardiologists over read
the echocardiogram they often report a percentage of forty-fifty percent and
that is not a clean data point. Cores measure reporting can cause significant financial
burdens on the organization if handled incorrectly. CHF is one of the data
mining concerns but so are myocardial infarction (MI), pneumonia, VTE, and
stroke. There is also the idea of providing quality patient care which makes
this program vital to the viability of the organization."
"We
are trying to predict or evaluate where we are the weakest so an action plan
can be instituted and our values can be superior. We are attempting to
preemptively determine where to focus our efforts for positive quality outcomes.
"
AMS- "What data are they using
in this data mining program?"
B.K. - With CHF, we are looking at the length of stay per
admission, the ejection fraction of the echocardiogram, the comorbidities that exist for the patient, like diabetes or
hypertension, the B-type Natriuretic Peptide (BNP)
blood test value, to mention a few. With the stroke patient, we are also
looking for the length of time for CAT scan or door- to CT scan time and the thirty-minute
window for tissue plasminogen activase (TPA) administration if warranted.
AMS-"What do you see as the advantages and disadvantages of the institution's approach
to data
mining?"
B.K. - "We are at an important crossroad at this institution
because we have to make a decision to adopt a hybrid approach to data mining or
wait and try to implement a global system using Siemens. Siemens DSS is moving
toward this global data mining system and this would be an advantage because we
predominantly utilize Siemens Soarian and their DSS platform is quite robust.
The new upgrade we performed in June has added many data mining features so we
are contemplating waiting to see what they come up with in the future. I am not
sure there are advantages except we are familiar with Siemens and have a good
working relationship. The disadvantages would be we are implementing this data
mining initiative somewhat late compared to local area hospitals so we are
under the gun, so to speak."
AMS- "What are the issues
with the institution's approach to data mining?
B. K. -"SpectraMD USA, Inc. is a new company and their premium program
concerns value based purchasing and CMS Cores Measures. This is their flagship data-mining
program but they intend to expand to be the premiere data mining company. They
do have a program for Physician Quality Reporting System Compliance but we use
Crimson Continuum of Care program so we do not need their program (Crimson
continuum of, 2013). We needed to get
the core measures under control so we implemented this program. We are not sure
that we want to explore their other programs at this time. We have to consider
that if we use their program for other data mining projects they may be appropriate
for some clinical areas but inappropriate for other clinical areas. This is why
we have nineteen clinical systems in our hospital because no one system fully understands
the electronic medical record documentation system of every clinical area. These
are big issues to tackle and we are investigating, researching and evaluating
the best way to move forward.
AMS-"Thinking of what this
institution is trying to do with data mining, what criteria
would you want
to use in evaluating data mining software.
B. K. - "We look
for many things when evaluating data mining software. The number one item I
look for is functionality because
that determines the adaptability of the program and how it can meet the
challenges that may occur with data mining. Performance is another category because
there is going to be many different interfaces and data sources and this
demonstrates the efficiency of the product. Another category is the ease of use
or user friendly capability of the data mining software. It has to be easy to learn
or there can be errors or incorrect documentation. Are there different levels
in this program that would affect learning is another question to consider. This
software also has to have other functions that will facilitate report production
and data manipulation. The data needs to be capable of filtering, randomizing,
and potentially deleting outliers. The data needs to be clean for the finished product
to be accurate and truly reflect the subject in which we are looking to mine
data."
"We
looked at these features with Focus but we were ready to purchase because of
time constraints and their program had acceptable criteria. That is why we are
taking more time with our decision to adopt hybrid or global system for data
mining."
AMS-" What role do you see the nurse playing regarding data mining, including a
discussion of the uses of data
mining to improve nursing practice?"
B. K. - "Nurses
are key players in documentation in electronic medical records and information
technology as it pertains to patient care. We have the IT meeting monthly to
have nurses input with the various projects that we are implementing. Nurses
have a different approach to IT systems and documentation and their thoughts and
opinions are valued. "Program developers look for nurses to be
involved in the development of nursing documentation," (Lyden, 2008, pg.9).
We have restructured Case Management to
have Care Coordinators and Clinical Data Informatics specialists that review
charts, round on the floors and input the data into Focus."
"In the future as data mining
expands here there will be more jobs that will be necessary to collect, clean
and input the data. With the capability of Soarian currently to run reports on
your individual patients or your assignment, you can structure your nursing
practice and look for trends. The
increased use of technology can improve nursing practice especially when we
convert to hand held devices. We have one more upgrade to convert Soarian
software to the hand held device capability and then we will begin introducing
the devices in certain clinical areas. All of these initiatives will change and
improve nursing care toward their patients."
Conclusion
The interviewer was helpful and
provided great insight into the world of data mining at the writer's institution.
The nurse informaticist uses science, nursing, evidence based theory, and
nursing care to provide quality patient care (Shuler, 2011) . From
a patient outcomes perspective there are informatics tools that help "shared
decision-making and risk communication from the perspective of patient safety:
(1) interactive education to improve risk comprehension; (2) multiple,
individualized formats for conveying risk; (3) individualized risk
calculations; (4) application of decision analysis methods to calculate options
with the highest expected value; (5) automated updates of evidence to support
shared decision making; and (6) utilization of different preference elicitation
techniques and formats," Data mining is also another tool for improving
patient outcomes," (Bakken, 2006, pg 233).
The increased use of data mining can
improve patient safety with shared decision-making,
increased communication, identifying
potential risks and interactive education (Shuler, 2011). There
will be greater roles in data input, analyzing and collecting data so patient outcomes
can be improved.
References
Are you
maximizing data to avoid financial penalties?. (2012). Retrieved from
https://www.spectramd.com/userfiles/File/FOCUS Actionable Analytics.pdf
Bakken, S.
(2006). Informatics for patient safety: a nursing research perspective. Annual
Review of Nursing Research, 24. Retrieved from http://resources.njstatelib.org/login?url=http://search.ebscohost.com/login.aspx?direct=true&db=mdc&AN=17078416&site=ehost-live
End of citation-->
Crimson
continuum of care. (2013). Retrieved from http://www.advisory.com/Technology/Crimson
Lyden, C. (2008). From Paper to Computer
Documentation: One easy step? Online
Journal
of Nursing Informatics (OJNI), 12 (3). Available at
http:ojni.org/12_3/Lyden.htm
Shuler, G. K.
(2011). Role of nursing informatics for leadership. Nursing Advance,
Retrieved from
Leadership.aspx
