Nursing Informatics Chapter 47

47
Downloaded by [ Faculty of Nursing, Chiangmai University 5.62.158.117] at [07/18/16]. Copyright © McGraw-Hill Global Education Holdings, LLC. Not to be redistributed or modified in any way without permission.
Computer Use in Nursing Research
Veronica D. Feeg / Theresa A. Rienzo
• OBJECTIVES
1. Describe general data and computer applications related to research in nursing,
including proposal development and project implementation in both quantita-
tive and qualitative research (computer use in research).
2. Summarize a range of computer-based, Web innovations, tablet and mobile
applications that facilitate or support the steps of the research process, includ-
ing data collection, data management and coding, data analysis, and results
reporting.
3. Compare and contrast select computer software applications that can be used
in quantitative and qualitative research data analysis related to the steps of the
research process.
4. Describe specific research studies in the literature that exemplify quantitative
and qualitative methodologies on computer applications in healthcare.
5. Describe general categories of research that focuses on computer use in health-
care, from clinical applications and informatics integration to health and wellness
(nursing research on computer use).
6. Introduce the new emphasis on “big data” and the growing interest in analysis of
existing data to answer research questions on populations.
7. Describe the drivers for promoting informatics research and computer applica-
tions today for improving safety and quality of healthcare in the future (research
on nursing informatics).
• KEY WORDS
Nursing informatics research Data management
Research process Data analysis
Research methodology Research applications
“Big Data” Computer applications
Quantitative Internet research
Qualitative Secondary analysis
Data collection Meta-analysis
Data mining
663
c47.indd 663 20-02-2015 19:04:31

664 P art 8 • R esearch A pplications
INTRODUCTION data, communication of those connected pieces of infor-
mation, and the elements in the nursing environments
Nursing research involves a plethora of tools and resources through informatics research if one is to understand how
that researchers employ throughout the research process. computers and nursing research co-exist.
From the individual or collaborative project initiation, Outside the EHR and computerized health systems,
through refinement of the idea, selection of approaches, in the rapidly changing world of Internet technologies,
development of methods, capturing the data, analyz- information management and computer-enhanced inter-
ing the results, and disseminating the findings, com- vention research on the use of computers have produced
puter applications are an indispensable resource for the a new body of science that will continue to grow. Blending
researcher. The investigators must be well prepared in a the focus of computer use in research (tools and process)
variety of computerized techniques for research activities and research on computer use (informatics research)
as they are employed in the domain of knowledge that will calls for an understanding of process and products. This
be investigated. Without the power of technology, con- chapter will provide an overview of the research process
temporary research would not reach the levels of sophis- for two separate and fundamentally different research
tication required to discover and understand health and approaches—quantitative and qualitative—and discuss
illness today. select computer applications and uses relative to these
In addition to the traditional approaches of the sci- approaches. The discussion will be supplemented by exam-
entific method, researchers today have new avenues to ples of current science and the trajectory of research on
explore in the development of knowledge. New opportu- the impact of informatics, electronic records, treatments,
nities to mine existing data for evaluation and discovery and integrated technologies using the computer as a tool.
are forming a bridge between the process of conducting The computer has been a tool for researchers in many
research and the products of discovery. New tools for aspects of the research process and has gone beyond its
automatic capture and analysis are changing the meth- historic application once limited to number crunching
ods textbooks. New online strategies and apps are being and business transactions. Field-notes binders, ring tab-
implemented as the process of researching health and lets, index cards, and paper logs have all but disappeared
the product of researched interventions in health. From in the researchers’ world. Personal computers, laptops,
mobile applications that are downloadable to smart PC-tablets and iPads, and handheld devices have become
phones for researchers or patients or both, to tools for part of the researcher’s necessary resources in mount-
exploring large data sets that have been already captured, ing a research project or study. Wireless technologies
computer use in nursing has exploded concomitantly with are ubiquitous and connect people to people as well as
computers in healthcare. researchers to devices. Cloud computing today connects
What is also of importance for computer use in nursing diverse enterprises with stable sources of software and
is the research on computerized applications and nursing data that can be shared or used by anyone, at any time
informatics, which is revolutionizing how health infor- or any place. Numerous enhancements have been added
mation is documented in the Electronic Health Record to the well-known text processing software products that
(EHR) and producing pools of searchable secondary data reduce time and effort in every research office. In addition,
waiting to be tapped. The context for nursing informat- a wide range of new technologies for database manage-
ics research has proliferated since the National Institute ment of subjects, contacts, or logistics have emerged in
of Nursing Research (NINR) published an agenda out- the research product marketplace. Nurse researchers use
lining the need for nursing informatics research in the a range of hardware and software applications that are
Nursing Informatics Research Agenda (NINR, 1993). generic to research development operations in addition
Other reports called for organizing priorities (Brennan, to the tools and devices that are specific to research data
Zielstorff, Ozbolt, & Strombom, 1998) and construct- collection, analysis, results reporting, and dissemination.
ing models to develop a context connecting nursing and New apps appear continuously, customized to the data
informatics that would provide the basis for studying the collection, management, or analysis process.
practice of nursing informatics (Effken, 2003). Today, In today’s electronic healthcare environment, numer-
hospital-wide information technology (IT) is the spine ous advances have been made with the sources of data col-
of all healthcare delivery, which is tied to reimburse- lection relative to general clinical applications in nursing,
ment, and which inevitably forms the data engine that health, and health services. System implementations for
health systems put to work to research improvement and large clinical enterprises have also provided opportuni-
outcomes-driven questions. It is therefore imperative to ties for nurses and health service researchers to identify
understand the underlying terminologies and sources of and extract information from existing computer-based
c47.indd 664 20-02-2015 19:04:31

Chapter 47 • Computer Use in Nursing Research 665
resources. In an era when the federal government is preparation, and implementation are applicable to both
calling for comparative effectiveness research (CER) to quantitative and qualitative approaches. However, no
address the rising costs of healthcare (Congressional discussion about computer use in research could be com-
Budget Office [CBO], 2007; Sox & Greenfield, 2009), the plete without acknowledging the range of research now
richness of capturing nursing data that can be managed appearing in the literature that examines the trajectory of
and mined for advanced analyses should be recognized in new technologies and computer use in patient care. With
the development of electronic health records (EHRs) and increasing emphasis on cost and quality of healthcare, the
other sources. computer-sources-of-data and computer-as-intervention
In addition, the era of Web-based applications has pro- must be part of understanding computer use in nursing
duced a wide range of innovative means of entering data research today.
and, subsequently, automating data collection in ways
that were not possible before. With the advancements in
clinical systems, acceptable terminology and vocabularies PROPOSAL DEVELOPMENT,
to support nursing assessment, interventions, and evalu- PREPARATION, AND
ation, computers are increasingly being used for clinical
and patient care research. Although research is a complex
IMPLEMENTATION
cognitive process, certain aspects of carrying out research All research begins with a good idea. The idea is typi-
can be aided by software applications. For example, exam- cally based on the nurse researcher’s identification of
ination of nursing care/patient outcomes and the effect of a problem that is amenable to study using a philosophi-
interventions would have been prohibitive in the past, but cal and theoretical orientation. The philosophical aspect
with the aid of computers and access to large data sets, sets the stage for selecting one’s approach to investigat-
many health outcomes can be analyzed quantitatively ing the problem or developing the idea. Good clinical
and qualitatively. It is noteworthy when hospital systems ideas often come from personal experiences, based on the
today can use analytics across institutions with large researcher’s foundation of knowledge that aids in drawing
samples of existing data to compare and predict best out- inferences from real clinical situations. These unfold by
comes with select interventions, for example, examining way of iterative consideration of problem and process—
pre-operative procedures that minimize post-operative leading the investigator to evolve an approach to the prob-
complications. In one hospital example, a procedure was lem, and subsequently a theoretical paradigm to address
changed with results from a rapid analysis of system-wide the problem. Because the theoretical paradigm emerges
data, reducing mortality and morbidity related to post-op from these iterative considerations, and because the theo-
complications. They went from asking the research ques- retical perspective will subsequently drive the organiza-
tion to implementing the procedural change in less than tion of the research study, it is important to distinguish
six months (Englebright, 2013)—change that in the past between these two distinct approaches. Each theoretical
would have taken 17 years in randomized trials as has paradigm directs how the problem for study will unfold.
been reported. The researcher uses a selected theoretical approach and
With a wider view of computer use in nursing research, operationalizes each step of the research process that
the objective of this chapter is fourfold: (1) to provide an will become the research design and methodology, either
overview of general computer and software applications qualitative, quantitative, or some combination of both in
related to the stages of the research process; (2) to describe mixed methods (Creswell, 2014). Each approach can be
how computers facilitate the work of the researcher in facilitated at different points along the proposal process
both quantitative and qualitative aspects; (3) to highlight with select computer applications. These will be described
research on computer use in healthcare; and (4) to give as they relate to the methodology.
attention to the explosion of research in categories delin-
eating clinical and nursing informatics research. These
Quantitative or Qualitative Methodology
will serve as a snapshot of the research on computer use
for the future with contextual influences. The important distinction to be made between the quan-
To begin, the chapter will focus on some of the con- titative and qualitative approaches is that for a quantita-
siderations related to the logistics and preparation of the tive study to be successful, the researcher is obliged to
research proposal, project planning, and budgeting, fol- fully develop each aspect of the research proposal before
lowed by the implementation of the proposal with data collecting any data, i.e., a priori, whereas, for a qualitative
capture, data management, data analysis, and information study to be successful, the researcher is obligated to allow
presentation. The general steps of proposal development, the data collected to determine the subsequent steps as
c47.indd 665 20-02-2015 19:04:31

it unfolds in the process and/or the analysis. Quantitative be integrated. Tables, charts, and images can be inserted,
research is derived from the philosophical orientations edited, and moved as the proposal takes shape, with final
of empiricism and logical positivism with multiple steps products in publishable forms. Personal computer appli-
bound together by precision in quantification (Polit & cations that allow inserting simple graphic designs give
Beck, 2011). The requirements of a hypothesis-driven the researcher a powerful means of expressing concepts
or numerically descriptive approach are logical conse- through art. Line art and scanned images using Adobe
quences of, or correspond to, a specific theory and its industry standards such as Illustrator CC (www.adobe.com)
related tenets. The hypothesis can be tested statistically or Photoshop CC (www.photoshop.com), now with cloud
to support or refute the prediction made in advance. capability, can be integrated into the document for clear
Statistics packages are the mainstay of the quantitative visual effects. These offer the researcher and grant manag-
methodologist, but are not the only connection to com- ers tools to generate proposals, reports, and manuscripts
puters for the researcher. that can be submitted electronically directly or follow-
The qualitative approaches offer different research tra- ing conversion to portable document formats (pdf) using
ditions (e.g., phenomenology, hermeneutics, ethnography, Adobe Acrobat or other available conversion products.
and grounded theory, to name a few) that share a common There are a variety of reference management software
view of reality, which consists of the meanings ascribed products available as add-ons to word processing, with
to the data such as a person’s lived experiences (Creswell, ranging prices and functionalities. For example, unique
2012). With this view, theory is not tested, but rather, per- template add-ons give Microsoft Word additional power
spectives and meaning from the subject’s point of view are to produce documents in formatted styles. Bibliographic
described and analyzed. For nursing studies, knowledge management applications emerge frequently and librar-
development is generated from the participant’s experi- ians often help sort out best ways to keep reference
ences and responses to health, illness, and treatments. materials in order. Common programs such as Reference
The requirements of the qualitative approach are a func- Manager, Procite, and Endnote are products of the
tion of the philosophical frames through which the data Thomson Corporation (2009) (www.refman.com/rmcopy
unfold and evolve into meaningful interpretations by the right.asp), the industry standard software tools for pub-
researcher. A variety of software applications assist the lishing and managing bibliographies on the Windows and
qualitative methodologist to enter, organize, frame, code, Macintosh desktops. RefWorks adds another option for
reorder, and synthesize text, audio, video, and sometimes reference management from a centrally hosted Web site
numeric data (Polit & Beck, 2011). (www.refworks.com). Searching online is one function of
these applications, and then working between the refer-
ence database and the text of the proposal document is
General Considerations in Proposal Preparation
efficient and easy, calling out citations when needed with
Several computer applications have become indispensable “cite as you write” capability into the finished document.
in the development of the research proposal and gener- Output style sheets can be selected to match publication
ally in planning for the activities that will take place when or proposal guidelines.
implementing the study. These include broad categories Research applications and calls for proposals are often
of office programs including word processing, spread- downloadable from the Internet into an interactive form
sheet, and database management applications. According where individual fields are editable and the documents
to Forrester Research, an independent research firm, can be saved in a portable format such as Adobe Acrobat,
Microsoft Office products currently capture 80% of users printed, or submitted from the Web. The Web also allows
with 64% of enterprises using Office 2007 (Montalbano, the researcher to explore numerous opportunities for
2009). Office 365 is the new release from Microsoft, designing a proposal tailored to potential foundations for
programs with cloud capability that continue to offer consideration of funding. Calls for proposals, contests, and
improved clerical tools to manage the text from numer- competitive grants may provide links from Web sites that
ous sources and assemble them in a cogent and organized give the researcher a depth of understanding of what is
package. The cloud connectivity gives researchers access expected in the proposal. There are more and more home
to all programs and data virtually from anywhere. PC, grown submission procedures today for grants and jour-
Mac, and Droid tablets and smart phones extend the reach nal manuscripts with Web-based instructions. These often
of all connectivity with a range of text processing and data convert the documents automatically to portable formats
management standard tools. (pdf ) for submission with key data fields organized and
Microsoft Office products provide capabilities and a sorted for easier review procedures. Instructions are cus-
platform into which other off-the-shelf applications can tomized for the user.
Research Study Implementation tracking and data capture devices, programs and systems
have been launched, including the exemplar of data man-
A funded research study becomes a logistical challenge agement tools from the recent U.S. Census that have cap-
for most researchers in managing the steps of the process. tured and made data available to researchers with data
Numerous demands for information management require tools (http://www.census.gov/main/www/access.html).
the researcher to maintain the fidelity of the procedures, One other important consideration related to research
manage the subject information and paper flow, and proposal implementation for the seasoned researcher or
keep the data confidential and secure. These processes the novice, doctoral dissertation investigator is the essen-
require researchers to use a database management system tial step of submitting the proposal to the Institutional
(DBMS) that is reliable. Several DBMS software applica- Review Board (IRB). Home institutions that have IRBs will
tions exist and have evolved to assist the researcher in the have specific procedures and forms for the researcher who
overall process of study implementation. These applica- can benefit from the proposal development electronically.
tions are operations oriented, used in non-research pro- In some institutions, the IRB document management has
grams and projects as well, but can assist the researcher been done through contracts with outside Internet orga-
in management of time, personnel, money, products, and nizations providing mechanisms for posting IRB mate-
ultimately dissemination, with reporting capability for rials, managing the online certifications required, and
reviews and audits. communicating with the principle investigators. One such
The ubiquitous Microsoft Office suite includes pro- example is IRBNet.org, hosting services for organizations
grams that manage data in a relational database (Microsoft to manage IRB and other administrative documents asso-
Access) and number crunch in a flat database (Microsoft ciated with the research enterprise (Research Dataware,
Excel) available to every researcher. Proprietary database LLC, 2014).
applications and new customized, more sophisticated, In summary, the general considerations of developing
integrated, and proprietary database management appli- and conducting a research study are based on philosophi-
cations from companies such as Oracle and Lotus provide cal approaches and will dictate which methodology the
the researcher with ways to operationalize the personnel, researcher will use to develop the study. Although this will
subjects, forms, interviews, dates, times, and/or track- subsequently influence the research and computer appli-
ing systems over the course of the project. Most of these cations to be used in carrying out the project, the steps of
applications require specially designed screens that are proposal preparation are less specific, and the computer
unique to the project if the research warrants compli- applications are useful in both quantitative and qualita-
cated connections such as reminders, but simple mail- tive studies. After identifying the research problem, how-
ing lists and zip codes of subjects’ addresses and contact ever, the researcher must proceed through the steps of the
information in a generic form can also be extremely useful process, where computers play an important role that is
for the researcher. Some of these applications are begin- unique to each of the methodologies.
ning to include add-ons to increase application portabil-
ity with devices such as the smartphones and tablets, and
mini-tablets. THE QUANTITATIVE APPROACH
Several other generic computer programs can aid the
Data Capture and Data Collection
researcher in daily operations and project management.
Spreadsheet applications are invaluable for budgeting and Data capture and data collection are processes that are
budget planning, from proposal development through viewed differently from the quantitative and qualitative
project completion. One multi-purpose Microsoft Office perspectives. Nurses may already be familiar with data
application is Microsoft Excel. Universally understood collection that is focused on the management of patient
and easy-to-use Excel allows the researcher to manage care. Patient monitoring, patient care documentation,
costs and calculate expenses over the course of the project and interview data are collected by nurses, although not
period, producing a self-documenting plan by categories always for research purposes. Data collection can take a
to track actual spending and money left. Templates can number of forms depending on the type of research and
be developed for repetitive tasks. Scheduling and project variables of interest. Computers are used in data collec-
planning software is also available from Microsoft includ- tion for paper-and-pencil surveys and questionnaires as
ing Microsoft Project that allows the project director to well as to capture physiologic and clinical nursing infor-
organize the work efficiently and track schedules and mation in quantitative or descriptive patient care research.
deadlines using Gantt charts over the lifetime of the proj- There are also unique automated data capturing applica-
ect. In more sophisticated research offices, customized tions that have been developed recently that facilitate
large group data capture in single contacts or allow paper
versions of questionnaires to be scanned directly into a
database ready for analysis or provided online with Web-
based survey tools.
Paper and Pencil Questionnaires. Paper and book-

let surveys do still exist today in data collection, but
new enhancements aid the researcher in time-saving
activities. Surveys and questionnaires can be scanned
or programmed into a computer application either in a
• FIGURE 47.1. MEMS (Medication Event Monitoring
microcomputer or on a Web site accessed through the
System) SmartCap Contains an LCD Screen; MEMS Reader
Internet. Computers are being used for direct data entry
Transfers Encrypted Data From the MEMS Monitor to
in studies where subjects enter their own responses via a
the Web Portal. (Published with permission of MWV
computer, and simultaneous coding of response to ques-
Healthcare/Aardex Group, www.aardexgroup.com.)
tions and time “online” can be captured or Web surveys
can be distributed widely. These online survey tools can
provide a wide range of applications, including paper or
portable versions, and range in price and functionality. increasingly popular for their cost and logistical benefits.
The use of notebook microcomputers, tablets, and The data from the Internet can be downloaded for analysis
mini-tablets have gained popularity in recent years for and several applications provide instant summary statis-
allowing the user to enter the data directly into the com- tics that can be monitored over the data collection period.
puter program at the time of the interview with a subject, Several of these programs are available for free with lim-
with innovations emerging in touch screens, screen pens, ited use; others yield advanced products that can be
and even wireless data entry with smartphones (Moss & incorporated into the research, giving mobility (e.g., smart-
Saba, 2011). Responses to questions can be entered by phones) and flexibility (e.g., scanning or online entry) to
the respondent or a surrogate directly into the computer the data capture procedures. Several of these applications
or Web site through Internet access. There are several include (1) Survey Monkey (www.surveymonkey.com);
research study examples where patients with chronic con- (2) E-Surveys Pro (www.esurveyspro.com); (3) Survey System
ditions used a computer application or the Internet as the (www.surveysystem.com); (4) Qualtrics (www.qualtrics.
intervention as well as the data capture device; patients com); and (5) SNAP Survey software (www.snapsurveys.
or caregivers responded to questions directly and the data com). Many of these products continue to enhance func-
were processed with the same system (Berry et al., 2010; tionality and delivery modes as well as integration with
Mullen, Berry, & Zierler, 2004). Other examples of unique statistical analysis programs and qualitative narrative
data capture in research include individual devices such as exportability. An example of screen flexibility and scanning
the “Smart Cap” used to measure patient compliance with output in SNAP is shown in Fig. 47.2.
medications. The Medication Event Monitoring System Several special applications have been used in nursing
®
(MEMS 6) (Fig. 47.1) automates digitized data that can be research that can facilitate large group data capture. Group
downloaded for analysis in research such as patient adher- use applications in specially designed facilities have been
ence studies (Figge, 2010; Rolley et al., 2008). developed to engage an audience in simultaneous activity,
A variety of online survey tools also provide researchers recording their impressions through electronic keypads
the power to collect data from a distance, without postage, located proximal to the users, and capturing that informa-
using the Internet. These applications can present question for display or later analysis. One type of application,
tionnaire data in graphically desirable formats, depending Expert Choice 11.5 (Expert Choice, 2014) uses the analytic
on the price and functionality of the software, to subjects hierarchy process (AHP), a mathematical technique, with
delivered via e-mail, Web sites, blogs, and even social net- handheld keypad technology to elicit group responses and
working sites such as Facebook or Twitter if desirable. automatically score, analyze, prioritize, and present infor-
Social media mechanisms such as blogs and tweets are mation back to the group graphically. AHP has been used
often providing sources of data analyses, albeit question- as the multiple criteria decision-making analysis tech-
ably scientific, that have sometimes been harnessed to nique in nursing research studies (Kimiafar, Sadoughi,
extract meaning for researchers. Web surveys, although Sheikhtaheri & Sarbaz, 2014; Kodadek & Feeg, 2002), and
often criticized for yielding poorer response rates than tra- the Expert Choice software allows for group data entry.
ditional mail (Granello & Wheaton, 2004), are becoming This kind of groupware for collaborative decision-making
• FIGURE 47.2. Sample Screens for SNAP Survey Software. (Reproduced, with permission, from Snap Surveys,
www.snapsurveys.com.)
can supplement data collection from a focus group to add products, enhanced even more with Web-based products
a quantitative component to the subjective question as such as Remark Web Survey (Gravic Inc., 2014), increase
it elicits and captures opinion via pairwise comparisons the accuracy of data entry with very low risk of errors,
(www.expertchoice.com). thereby improving the efficiency of the data capture, col-
Software packages also exist that can be integrated lection, and entry processes.
with the researcher’s scanner to optically scan a spe-
cially designed questionnaire and produce the subjects’ Physiologic Data. The collection of patient physiologic
responses in a database ready for analysis. OmniPage 18 parameters has long been used in physiologic research.
(Nuance Imaging, 2014) is a top-rated optical charac- Some of these parameters can be measured directly from
ter recognition (OCR) program that converts a scanned patient devices such as cardiac monitoring of heart rhythm,
page into plain text. Programs such as SNAP Survey soft- rate, and fluid or electrolytes and be captured in the patient
ware (www.snapsurveys.com) and Remark Office OMR care records of the hospital systems. For example, hospi-
8 (Gravic Inc., 2014) can facilitate scanning large num- tals have developed mechanisms to use information from
bers of questionnaires with speed and accuracy. These intensive care unit (ICU) data to calculate benchmarks for
mortality and resource use. Now that many measurements or tissue rejection post-heart transplantation from blood
taken from various types of imaging (e.g., neurologic, car- samples and biopsies (Berger & Berger, 2004).
diovascular, and cellular) have become digitized, they can The National Institutes of Health (NIH) is undertaking
also be entered directly from the patient into computer several initiatives to address the challenges and opportu-
programs for analysis. Each of these applications is unique nities associated with big data. As one component of the
to the measures, such as systems to capture cardiac func- NIH-wide strategy, the Common Fund in cooperation with
tioning and/or pulmonary capacity, devices that can relay all NIH Institutes and Centers is supporting the Big Data
contractions, or monitors that pick up electronic signals to Knowledge (BD2K) initiative, which aims to facilitate
remotely. Numerous measurements of intensity, ampli- broad use of biomedical big data, develop and disseminate
tude, patterns, and shapes can be characterized by com- analysis methods and software, enhance training for disci-
puter programs and used in research. For example, the plines relevant for large-scale data analysis, and establish
APACHE system in its multiple versions has been tested in centers of excellence for biomedical big data (NIH, 2012).
benchmarking hospital mortality and outcomes from cap-
tured physiologic data in several groups of patients in the Unique Nursing Care Data in Research. Scientists
ICU (Paul, Bailey, Van Lint, & Pilchers, 2012; Zimmerman, and technologists from a variety of disciplines are work-
Kramer, McNair, & Malila, 2006). Each of these measure- ing hard to identify the domain of data and information
ment systems has evolved with the unfolding of research that is transferable across situations, sites, or circum-
specific to their questions, and within each community of stances that can be captured electronically for a wide
scholars, issues about the functionality, accuracy, and reli- array of analyses to learn how the health system impacts
ability of electronic data extracted from these physiologic the patients it serves. The American Nurses Association
devices are debated. (ANA) has supported the need to standardize nursing care
Along with the proliferation of clinical diagnostic mea- terms for computer-based patient care systems. The clini-
surement systems, there has been a rapid expansion of cal and economic importance of structured recording to
unique computer applications that have emerged for the represent nursing care was recognized by the acceptance
data analysis aspects of these clinical systems, physiologic of the nursing minimum data set (NMDS) (Werley, Lang,
and record sources. Millions of gigabytes of data are stored & Westlake, 1986). The ANA has accepted seven sys-
in machines that can be tapped for multiple studies on the tems of terminology for the description of nursing prac-
existing data. Data mining is a powerful tool in the knowl- tice: the North American Nursing Diagnosis Association
edge discovery process that can now be done with a number (NANDA) taxonomy of nursing diagnosis, (Clinical Care
of commercial and open-source software packages (Berger & Classification [CCC] System) (McCormick, Lang, &
Berger, 2004). Data mining and the evolving “big data” ini- Zielstorff, 1994; Saba, 1997, 2007; Zielstorff, Lang, & Saba,
tiatives to make patient care data available introduce new 1995), Nursing Interventions Classification (Bulechek &
ways to manipulate existing information systems. McCloskey, 1997), Nursing Outcomes Classification (Daly,
With increased attention to comparative effectiveness Maas, & Johnson, 1997), patient care data set (Ozbolt,
research (CER), several government and private organiza- 1999), Omaha Home Healthcare (Martin & Norris, 1996),
tions are encouraging researchers to hone the techniques and the International Classification of Nursing Practice
to extract valid and reliable information from these large (Saba, 1997, 2007; Zielstorff et al., 1995). The Clinical
data sets (Sox & Greenfield, 2009). Data mining is a mecha- Care Classification System (sabacare.com) nursing ter-
nism of exploration and analysis of large quantities of data minology has been accepted by the U.S. Department of
in order to discover meaningful patterns and rules, applied Health and Human Services (HHS) as a named stan-
to large physiologic data sets as well as clinical sources of dard within the Healthcare Information Technology
data. The nature of the data and the research question Standards Panel (HITSP) Interoperability Specification for
determine the tool selection (i.e., data-mining algorithm or Electronic Health Records, Biosurveillance and Consumer
technique). Tools and consultants exist to help researchers Empowerment as presented to a meeting of the American
unfamiliar with data mining algorithms use data mining Health Information Community (AHIC), a federal advisory
for analysis, prediction, and reporting purposes (MSDN group on health information technology (DHHS, 2007).
Library, 2000). Many of the first commercial applications (See Appendix A for a full discussion of the CCC System.)
of data mining were in customer profiling and marketing Although none of the above has emerged as a single
analyses. Today, many special technologies can be applied, standard, national efforts (Dykes, 2013) and collab-
for example, to predict physiologic phenomena such as orative efforts with the International Classification of
genetic patterns in tumors that might respond to therapy Nursing Practice (ICNP) from the International Council
based on classification of primary tumor gene expression of Nursing (ICN) are underway to “harmonize” the data
elements in the NMDS and SNOMED CT (Systematized that warrant technical or cognitive applications. Coding
Nomenclature of Medicine Clinical Terms) (Coenen & data is a combination of cognitive decisions and mechani-
Jansen, 2013). With a structured coding system to record cal clerical recording of responses in a numerical form
patient care problems that are amenable to nursing with numerous places for error to occur. There are several
actions, the actual nursing actions implemented in the ways of reviewing and “cleaning” the data prior to analy-
care of patients, and the evaluation of the effectiveness sis. Some computer programs allow for the same data to
of these actions, researchers can analyze large n ursing be entered twice called double-data entry or two-pass
data (Bakken, 2013; Byrne & Lang, 2013; Englebright, verification. This is done preferably by different people to
Aldrich, & Taylor, 2014). Outcomes research and qual- check for errors, with the premise that if the double entry
ity indicators extracted from health information systems does not match, one entry is wrong. One also must check
(HIS) have become the data end points that can justify for missing data and take them into consideration in the
healthcare services (Nahm, Vaydia, Ho, Scharf, & Seagull, coding and analyses. New versions of advanced statistical
2007). The American Recovery and Reinvestment Act software help in these activities. Reviewing data for values
(ARRA) of 2009 (Public Law 111-5) called for increased outside of those allowable is another way of examining the
development, certification, and wide-range “meaning- data for errors. It can best be done by examining the mul-
ful use” of EHRs across healthcare. The federal govern- tiple printouts produced by the statistical software pack-
ment has incentivized development of cross-platform ages or procedures invoked in the statistical application
compatibility and collaboration. Research on outcomes and by carefully perusing for outliers or artifacts.
of care is one of the centerpieces of this massive policy Another type of data coding can be described in the
that has begun to show an impact on integrated informa- example of the process of translating data from docu-
tion technologies in healthcare that can transform prac- mentation of patient care using coding strategies. Current
tice. Nursing research on nursing practice captured from research on coding nursing data using standardized nurs-
standardized terminology will be essential to document ing terminology from standardized codes is evident in
outcomes of nursing care. several research studies in the literature (Englebright et
al., 2014). For example, using precisely coded data from
a standardized terminology can produce data that can
Data Coding
be aggregated and statistically analyzed into meaningful
In most quantitative studies, the data for the variables information. In several studies by Saba and Taylor (2007),
of interest are collected in a numerical form. These Moss and Saba (2011), and Saba and Arnold (2004),
numerical values are entered into designated fields in researchers have discussed mechanisms of aggregating
the process of coding. Coding may be inherent in soft- nursing action types, e.g., assess, perform, teach, or man-
ware programs for the physiologic data and many of age, into aggregated information on the amount of time
the electronic surveys. The coding may be generated or effort a nurse spends in a day and concomitant costs
by a computer program from measurements directly associated. Another application of coding data occurs
obtained through imaging or physiologic monitoring, or with mapping concept codes from one terminology to
entered into a computer by a patient or researcher from another. For example, coding data for a nursing diagno-
a printout or a questionnaire or survey into a database sis can be mapped between terminologies for interoper-
program. Most statistical programs contain data editors ability, e.g., mapping the CCC Nursing Diagnosis concept
that permit the entry of data by a researcher as part of of pain to a SNOMED-CT code for the same concept
the statistical application. In such a situation, fields are (CCC Acute Pain—Q63.1 mapped to SNOMED Acute
designated and numerical values can also be entered into Pain—274663001). These translations of codes support
the appropriate fields without the use of an extra pro- the ever-evolving interoperable computer systems that
gram. For mechanisms that translate and transfer source produce nursing data reports which can cross platforms
data to prepare it for analysis, generic programs such or institutions.
as Microsoft Excel serve multiple needs. In addition to
allowing simple transfers of data from source to a sta-
Data Analysis
tistical analysis package, Excel has its own powerful, but
simple, analysis capabilities and exceptionally easy-to- Data analysis in a quantitative study combines a variety
use graphic translators that can turn statistics into visual of techniques that apply statistical procedures with the
graphs and charts. researcher’s cognitive organization of research questions,
Coding data is a precise operation that needs careful results, and visual or textual information, translated into
consideration and presents the researcher with challenges tables, charts, and graphs to make the data meaningful.
c47.indd 671 20-02-2015 19:04:32

It translates the numeric and conceptual elements of the dividing scores by certain log or factor values. Combining
inquiry into meaningful representations of information. In several existing variables can also create new variables.
general, the statistical analyses are ordered by the concep- These transformations or “re-expressions” or “dummy-
tual arrangement of hypotheses, variables, measurements, coding” allow the researcher to analyze the data in appro-
and relationships, and ultimately answer the research ques- priate and interpretable scales. The researcher can then
tions. There are many ways to consider data analysis. The easily identify patterns with respect to variables as well as
presentation below is organized around the broad types of groups of study subjects of interest. Both commercial sta-
research of interest in nursing and general research goals tistical packages provide the ability to calculate these tests
or questions. The researcher may use different types of and graphically display the results in a variety of ways.
analyses depending on the goal of research. These goals IBM SPSS Statistics 22 provides the user with a broad
may require different statistical examinations: descriptive range of capabilities for the entire analytical process. SPSS
and/or exploratory analyses, hypothesis testing, estima- is a modular, tightly integrated, and full-featured software
tion of confidence intervals, model building through mul- comprised of the SPSS base and a range of add-on modules.
tivariate analysis, and structural equation model building. With SPSS, the researcher can generate decision-making
Various types of nursing research may contain a number information quickly using a variety of powerful statistics,
of these goals. For example, to test an intervention using understand and effectively present the results with high-
an experimental or quasi-experimental design, one may quality tabular and graphical output, and share the results
first perform descriptive or exploratory analyses followed with others using various reporting methods, including
by tests of the hypotheses. Quality improvement, patient secure Web publishing. SAS 9.2 provides the researcher
outcome, and survival analysis studies may likewise con- with tools that can help code data in a reliable frame-
tain a number of different types of analyses depending on work, extract data for quality assurance, exploration, or
the specific research questions. analysis, perform descriptive and inferential data analyses,
In general, the statistical analysis steps of the research maintain databases to track, and report on administrative
process rely heavily on the functions specific to a variety of activities like data collection, subject enrollment, or grant
statistical software applications. Two of the most popular payments, and deliver content for reports in the appropri-
programs in use today are the IBM SPSS Statistics 22 (for- ate format. SAS allows for creating unique programming
merly Statistical Package for Social Sciences) (IBM SPSS within the variable manipulations and is often the format
Statistics, 2014) and Statistical Analysis Services (SAS for large publicly available data sets for secondary analysis.
Version 9.2), however a variety of other packages and pro- Stata 13 and SYSTAT 13 are also fully integrated statistical
grams exist, such as STATA or Minitab 17 (Minitab, Inc., packages with full database management capabilities and
2014) unique to particular scientific disciplines. Which a range of sophisticated statistical tests particularly use-
package one selects depends on the user’s personal prefer- ful for epidemiologists and physical scientists. All of these
ence, particular strengths, and limits of the applications statistical packages have evolved to provide an integrated
including number of variables, options for analyses, and collection of tools that assist in aspects of research study
ease of use. These packages have given the user the power management—from planning to dissemination—in addi-
to manipulate large data sets with relative ease and test out tion to the reputable statistical analyses and data manipu-
statistical combinations that have exponentially improved lation capabilities that they have provided for many years.
the analyses possible in a fraction of time that it once took. As part of exploratory analysis, simple, binary, and
The different types of analyses required by the goals multiple regression analyses can be used to examine the
of the research will be addressed further. This description relationships between selected variables and a dependent
will be followed by examples of types of nursing research measure of interest. Certain models can be developed to
that incorporate some of these types of analyses. determine which collection of variables provides the best
prediction of the dependent measure. Printouts of correla-
Descriptive and Exploratory Analysis. The researcher tion matrices, extensive internal tests of data assumptions
may first explore the data means, modes, distribution on the sample, and regression analysis tables provide the
pattern, and standard deviations, and examine graphic researcher with condensed, readable statistical informa-
representations such as scatter plots or bar graphs. Tests tion about the relationships in question.
of association or significant differences may be explored
through chi-squares, correlations, and various univariate, Hypothesis Testing or Confirmatory Analyses. Hypothesis
bivariate, and trivariate analyses, and an examination of testing or confirmatory analyses are based on an inter-
quartiles. During this analysis process, the researcher may est in relationships and describing what would occur if
recode or transform data by mathematically multiplying or a hypothesis were true. The analysis of data allows us to
c47.indd 672 20-02-2015 19:04:32

compare the actual outcomes with the hypothesized out- whether observed (such as survey data) or latent (such as
comes. Inherent in hypothesis testing is the probability (P satisfaction or loyalty), can be used to predict any other
value) of an event occurring given a certain relationship. variable.
These are conditional relationships based on the variables
selected for study, and the typical mathematical tables Meta-Analysis. Meta-analysis is a technique that allows
and software for determining P values are accurate only researchers to combine data across studies to achieve
insofar as the assumptions of the test are met (Polit & more focused estimates of population parameters and
Beck, 2011). Certain statistical concepts such as statistical examine effects of a phenomenon or intervention across
power, type II error, selecting alpha values to balance type multiple studies. It uses the effect size as a common metric
II errors, and sampling distribution are decisions that the of study effectiveness and deals with the statistical prob-
researcher must make regardless of the type of computer lems inherent in using individual significance tests in a
software. These concepts are covered in greater detail in number of different studies. It weights study outcomes in
research methodology courses and are outside the scope of proportion to their sample size and focuses on the size of
the present discussion. An example, Power and Precision the outcomes rather than on whether they are significant.
(Version 2), is a computer program for statistical power Although the computations can be done with the aid
analysis and confidence intervals (Biostat, 2014). of a reliable commercial statistical package such as Meta-
Analysis (Borenstein, Hedges, Higgins, & Rothstein,
Model Building. An application used for a confirmatory 2009), the researcher needs to consider the following spe-
hypothesis testing approach to multivariate analysis is cific issues in performing the meta-analysis (Polit & Beck,
structural equation modeling (SEM) (Byrne, 1984). Byrne 2011): (1) justify which studies are comparable and which
describes this procedure as consisting of two aspects: are not, (2) rely on knowledge of the substantive area to
(1) the causal processes under study are represented identify relevant study characteristics, (3) evaluate and
by a series of structural (i.e., regression) equations and account for differences in study quality, and (4) assess the
(2) these structural relations can be modeled pictorially generalizability of the results from fields with little empiri-
to enable a clearer conceptualization of the theory under cal data. Each of these issues must be addressed with a
study. The model can be tested statistically in a simultane- critical review prior to performing the meta-analysis.
ous analysis of the entire system of variables to determine Meta-analysis offers a way to examine results of a
the extent to which it is consistent with the data. If good- number of quantitative research that meet meta-analysis
ness of fit is adequate, the model argues for the plausibility researchers’ criteria. Meta-analysis overcomes problems
of postulated relationships among variables (Byrne, 1984). encountered in studies using different sample sizes and
Most researchers may wish to consult a statistician to dis- instruments. The software application Meta-Analysis pro-
cuss the underlying assumptions of the data and plans for vides the user with a variety of tools to examine these stud-
testing the model. Traditionally, different types of mod- ies. It can create a database of studies, import the abstracts
eling programs, such as LISREL (Joreskog & Sorbom, or the full text of the original papers, or enter the research-
1978) (see www.ssicentral.com) or EQS (Byrne, 1984) (see er’s own notes. The meta-analysis is displayed using a sche-
Multivariate Software, www.mvsoft.com), are commer- matic that may be modified extensively, as the user can
cially available. The researcher will identify latent (unob- specify which variables to display and in what sequence.
servable) variables of interest (e.g., emotions) and link The studies can be sorted by any variable including effect
them to those that are observable (direct measurement) size, the year of publication, the weight assigned to the
and plan with the statistician to specify and examine the study, the sample size, or any user-defined variables to facil-
impact of one latent construct on another in the modeling itate the critical review done by the researcher (Fig. 47.3).
of causal direction.
IBM SPSS 22 offers Amos 22, a powerful SEM and path Graphical Data Display and Analysis. There are occa-
analysis add-on to create more realistic models than if sions when data need to be displayed graphically as part
using standard multivariate methods or regression alone. of the analysis and interpretation of the information or for
Amos is a program for visual SEM and path analysis. User- more fundamental communication of the results of com-
friendly features, such as drawing tools, configurable tool- putations and analyses. Visualization software is becom-
bars, and drag-and-drop capabilities, help the researcher ing even more useful as the science of visualization in
build structural equation models. After fitting the model, combination with new considerations of large data from
the Amos path diagram shows the strength of the relation- the “Fourth Paradigm” unfolds (Hey, Tansley, & Tolle,
ship between variables. Amos builds models that realisti- 2009). These ideas begin with the premise that mean-
cally reflect complex relationships because any variable, ingful interpretation of data intensive discoveries needs
• FIGURE 47.3. Comprehensive Meta-Analysis (CMA) User Interface. (www.meta-analysis.com/pages/features/
spreadsheet.html. Published with permission of Biostat, Inc.)
visualizations that facilitate understanding and unfold- The information sets include maps and globes with three-
ing of new patterns. Nurses are currently discovering new dimensional capabilities to describe networks, topologies,
ways to present information in meaningful ways through terrains, surveys, and attributes.
these visualization techniques (Delaney, Westra, Monsen, In summary, the major emphasis of this section has
Gillis, & Docherty, 2013). provided a brief discussion about the range of traditions,
Most statistical packages including SPSS, SAS, and statistical considerations, and computer applications that
STATA, and even spreadsheets such as Excel provide the aid the researcher in quantitative data analysis. As com-
user with tools for simple to complex graphical transla- puters have continued to integrate data management
tions of numeric information, thus allowing the researcher functions with traditional statistical computational power,
to display, store, and communicate aggregated data in the researchers have been able to develop more extensive
meaningful ways. Special tools for spatial representations and sophisticated projects with data collected. Gone are
exist, such as mapping and geographic displays, so that the days of the calculator or punch cards, as the comput-
the researcher can visualize and interpret patterns inher- ing power now sits on the researchers’ desktops or lap-
ent in the data. Geographic information system (GIS) tops, with speed and functionality at their fingertips.
technology is evolving beyond the traditional GIS com-
munity and becoming an integral part of the information
infrastructure of visualization tools for researchers. For THE QUALITATIVE APPROACH
example, GIS can assist an epidemiologist with mapping
Data Capture and Data Collection
data collected on disease outbreaks or help a health ser-
vices researcher graphically communicate areas of nurs- The qualitative approach focuses on activities in the steps
ing shortages. GIS technology illustrates relationships, of the research process that differ greatly from the quan-
connections, and patterns that are not necessarily obvious titative methods in fundamental sources of data, collec-
in any one data set, enabling the researcher to see over- tion techniques, coding, analysis, and interpretation.
all relevant factors. ArcGIS 10 system is a GIS for man- Thus, the computer becomes a different kind of tool for
agement, analysis, and display of geographic knowledge, the researcher in most aspects of the research beginning
which is represented using a series of information sets. with the capture and recording of narrative or textual data.
c47.indd 674 20-02-2015 19:04:33

In terms of qualitative research requiring narrative transcribed by a typist. Coding qualitative text data was
content analysis, the computer can be used to record a time-consuming task, often involving thousands of
the observations, narrative statements of subjects, and pages of typewritten notes and the use of scissors and tape
memos of the researcher in initial word processing appli- for the development of coding and categories. With the
cations for future coding. Software applications that aid advent of computer packages, the mechanical aspects of
researchers in transcription tasks include text scanners, the coding and sorting have been reduced. The researcher
such as OmniPage 18 (Nuance Imaging, 2014). Other must decide on which text may be of interest and can use
devices include vocal recorders or speech recognition a word processing program to search for words, phrases,
software such as consumer priced Dragon Naturally or other markers within a text file. However, this process
Speaking 12 (Nuance Dragon Solutions, 2014), where the is cumbersome and time consuming, with limited ways to
researcher can input the information into text documents aggregate text into meaningful combinations for identify-
by speaking into a microphone without typing. New digi- ing themes from the narrative.
tal recorders are also on the market that use sophisticated Some specific software packages developed for quali-
and higher cost voice recognition software. From these tative data analysis (QDA) interface directly with the most
technologies, researchers or transcriptionists can eas- popular word processing software packages. The early
ily manipulate the recording and type the data verbatim. application program Ethnograph (Seidel & Clark, 1984;
Even iPhones and smartphones have high-quality record- Seidel, Friese, & Lenard, 1994) was one of the first pack-
ing applications that aid the qualitative research capture ages developed specifically for the purpose of managing
narrative statements. These narrative statements, like the some of the mechanical tasks of qualitative data analysis.
quantitative surveys, can be either programmed for use in Ethnograph 6.0 gave users a project management interface
other applications or subjects’ responses can be entered with functions to code, edit, and search data. The older
directly into the computer. non-numerical unstructured data indexing, searching,
and theorizing (NUD-IST) software was another qualita-
Qualitative Data Collection. Audiotaping is often used tive package commonly used, now absorbed in other QDA
for interviews in qualitative studies, whereby the content products. This program assisted the researcher to estab-
is transcribed into a word processing program for analy- lish an index of data codes and seek relationships among
sis. The narrative statements are stored for subsequent the coding categories. The ease with which researchers
coding and sorting according to one’s theoretical frame- can code and recode large amounts of data with the aid
work. Through analysis, categories from the data emerge of computerized programs encourages the researcher to
as interpreted by the researcher. It is important to point experiment with different ways of thinking about data and
out that for both quantitative and qualitative data, the re-categorizing them. Retrieval of categories or elements
computer application program is only a mechanical, cleri- of data is facilitated by computer storage. Newer technol-
cal tool to aid the researcher in manipulating the data. ogies have evolved from Ethnograph and NUD-IST with
Using the Internet for indirect and direct data collection improved user interfaces, including the latest versions
in qualitative studies can also provide a vehicle for data of NVivo 10, MAXQDA 11 (MAXQDA: The Art of Data
analysis that yields a quantitative component as well as the Analysis, 2012), and ATLAS.ti 7 (ATLAS.ti 7, 2014).
qualitative analysis. Computers are not only able to record Qualitative research, like quantitative research, is not
the subject’s responses to the questions but can record the a single entity, but a set of related yet individual tradi-
number of minutes the subject was online and the number tions, aims, and methods. Some individual traditions
of times they logged in. Many new online technologies are within qualitative research are ethnography, grounded
providing functionality for qualitative studies: for exam- theory, phenomenology, and hermeneutics. The distin-
ple, Audacity (audacity.sourceforge.net), an open source guishing feature of qualitative research is that the goal is
free audio recording package, can edit captured voice and to understand the qualities or essence of phenomena and/
export audio data to be analyzed; conversely, online sur- or focus on the meaning of these events to the participants
vey packages such as SurveyMonkey (surveymonkey.com) or respondents in the study. The forms of data are usually
can now export participants’ free text data into qualitative the words of the respondents or informants rather than
software packages. numbers. Computerization is especially helpful to the
researcher in handling large amounts of data. However, it
must be stressed that the computer applications aid the
Data Coding and Data Analysis
analysis as a management tool rather than an analytical
Historically, qualitative researchers have relied on nar- one. Synthesis of the data is still the interpretive work of
rative notes, often first recorded as audio and later the researcher.
Computer Application Programs. A number of gen- building, it also can manage documents, audio and video
eral-purpose or specific software packages can be used in files as categories, attributes, or nodes in visual displays
qualitative analysis: one package is a free text retrieval pro- that show the structure and properties of the document
gram such as that available in a word processing program; (Fig. 47.4). The latest version of NVivo 10 also allows
another is any number of standard database management researchers to import exported data from other applica-
or indexing programs; third is a program specifically tions such as the online survey tool, Survey Monkey, as
developed for the purpose of qualitative analysis. well as bibliographic management programs, Facebook
and Twitter. New analysis tools provide the research ways
General Purpose Software. Word processing pro- to cluster, map, and visualize text and images in meaning-
grams offer a number of features useful to the qualitative ful ways to aid the interpretive process in the qualitative
researcher in the early stages of analysis. The ability to analysis.
search for certain key words allows the researcher to tag
the categories of interest. In addition, such features as cut Conceptual Network Systems. A system known as con-
and paste; linking texts; insertion of pictures, tables, and cept diagrams, semantic nets, or conceptual networks is
charts; and the inclusion of video and audio data enhance one in which information is represented in a graphic man-
the application. Add-on applications specific to integrat- ner. The objects in one’s conceptual system (e.g., age and
ing multiple elements help the researcher organize a range experiences) are coded and represented by a box diagram
of data and materials for analysis. These files can then be (node). The objects are linked (by arcs) to other objects
incorporated into more sophisticated programs. A com- to show relationships. Like rule-based systems, semantic
prehensive program is ATLAS.ti (Version 7.0), a powerful nets have been widely used in artificial intelligence work.
workbench for the qualitative analysis of large bodies of In order to view the relationships of an object in the sys-
textual, graphical, audio, and video data. It offers a variety tem, the researcher examines the node in the graph and
of tools for accomplishing the tasks associated with any follows the arcs to and from it. Semantic network appli-
systematic approach to “soft” data, such as material that cations may be useful in model building and providing
cannot be analyzed by formal and statistical approaches in a pictorial overview. Decision Explorer (Banxia, 2014)
meaningful ways (ATLAS.ti 7, 2014). offers the user a powerful set of mapping tools to aid in
the decision-making process for audience response activi-
Special Purpose Software. Several software products ties. Ideas can be mapped and the resulting cognitive map
have evolved and improved for the specific purpose can be further analyzed (Fig. 47.5). The software has many
of analyzing qualitative data. Ethnograph from Qualis practical uses, such as gathering and structuring interview
Research (2008) is one such older program, which is data and as an aid in the strategy formulation process. The
used after the data have been entered with a word pro- software is primarily described as being a recording and
cessing program and converted to an ASCII file. Each file facilitation tool for the elicitation of ideas, as well as a tool
can be designated by its context and identifying features to structure and communicate qualitative data. It allows
using markers provided by the computer program. The the user to gather and analyze qualitative data and thus
researcher can have the program to produce a file that make sense of many pieces of qualitative data in order
numbers each line of the narrative data. From this line to achieve a coherent picture of a given issue or problem
file, the researcher can begin to assign each line or para- (Banxia, 2014).
graph a category. The researcher keeps track of the cat-
egory definitions and is alert to dimensions that emerge. Data Analysis for Qualitative Data. Qualitative data
Recoding can be done to provide for inductive thinking analyses often occur on an ongoing basis with data collec-
and iterative comparisons. Through the use of a search tion in a reflexive and iterative fashion. There is no clear
command, the computer program can be made to search demarcation of when data collection should end and anal-
for data segments by categories throughout the typed ysis should begin. The process of obtaining observations,
document. interviews, and other data over a period of time results
NVivo 10 and XSight from QSR provide a new gen- in a vast body of narrative that may include hundreds or
eration of software tools with multiple advantages for thousands of pages of field notes and researcher memos.
researchers. Because qualitative research takes many Although computer applications can aid considerably in
forms, these two applications can be selected based on the organizing and sorting this mass of data, the theoretical
user’s specific methodologic goals, the nature and scale of and analytical aspects of decision-making about concepts
the study, and the computer equipment. While NVivo 10 and themes must be made by the researcher. Researchers
supports fluid, rich data, detailed text analysis, and theory can only use the tools to help in creating composites
c47.indd 676 20-02-2015 19:04:33

• FIGURE 47.4. Screenshot of NVivo9 Main Window. (www.qsrinternational.com. Published with permission from QSR
International.)
described by methodologists in coding levels and categor- (1978) indicated that the researcher may engage in theo-
ical clusters (Polit & Beck, 2011). retical sampling, which is a deliberate search for episodes
As an example, some of the tasks the computer can in incidents that enlarge the variances of properties and
facilitate in data analysis using grounded theory (one place boundaries around categories. Using software, these
approach to qualitative research) are as follows. Once a cognitive processes are applied by the researcher in data
researcher has determined which parts of the interviews analysis of narrative interviews, field notes, and supple-
and observations can be tagged as categories, certain mentary data.
properties or dimensions can be determined and coded
up through levels. The researcher may engage in “constant Uses and Caution. Software programs exist for qualita-
comparison,” comparing the meanings of all incidents that tive research that save researcher time doing file man-
have been similarly categorized. This process should con- agement, reducing the manual labor of cutting, pasting,
tinue until the researcher determines that the categories sorting, and manual filing. They may also encourage the
are internally consistent, fit with the data, and are satu- researcher to examine the data from different perspectives,
rated. Saturation is achieved when the researcher can find recoding and reorganizing the data in different frame-
no more properties for a category and new data are redun- works. However, one must be mindful that qualitative
dant with the old (Creswell, 2012). analysis is a cognitive process, not a mechanical one. The
Classic qualitative methods experts Strauss and essence of qualitative research is the meaning and inter-
Corbin (1990) suggested that in the later stages of pretation of the data within context. The ability of software
research, the researcher may engage in axial coding. In enhancements to generate quasi-frequency distributions
this stage, the researcher elaborates and explains key cat- and cross-tabulations tend to further increase the investi-
egories, considering the conditions under which the event gator’s confidence in believing such findings and relation-
occurs, the processes that take place, and possible con- ships, when in fact these may be an artifact of the way in
sequences. Another well-known methodologist, Glaser which the data are manipulated. While computer programs
• FIGURE 47.5. Example of Cognitive Mapping Analysis. (Published, with permission, of BANXIA® Software Ltd., www.
banxia.com/dexplore/pictures.html.)
facilitate coding, organization of data, and preparation of times. Nearly all organizations have their own Web sites.
the data for interpretation, they cannot replace the think- Some examples are the Alzheimer’s Disease Education and
ing and decision-making that is at the heart of qualitative Referral Center (www.alzheimers.org), American Heart
analysis. As in all research, the burden of analysis and inter- Association (www.americanheart.org), American Medical
pretation rests with the researchers (Polit & Beck, 2011). Informatics Association (www.amia.org), and RAND
Corporation (www.rand.org). The Cochrane Collection
has numerous centers all over the world through the
Dissemination of Results
Cochrane Collaborative (www.cochrane.org). As with all
While dissemination of results continues to occur by tra- publications, online as well as hardcopy, the information
ditional means such as presentations at professional meet- accessed must be evaluated by the users regarding appro-
ings and publication in journals and monographs, online priateness for the purpose for which it was retrieved.
reporting is becoming increasingly common. Some Web Reports to most government and some non-govern-
sites frequented by nurses are peer-reviewed journals ment agencies require the researcher to submit a converted
such as Online Journal of Nursing Informatics (www. document online. Grant proposals submitted to the federal
cisnet.com) and selected nursing articles on various Web government currently require online submission with con-
sites such as that of the ANA (www.nursing-world.org). version to PDF. National Institutes of Health (NIH) appli-
Nursing forums sponsored by various professional nursing cants are directed to a page with downloadable programs
organizations (e.g., American Journal of Nursing, Sigma to convert the documents before submitting them (www.
Theta Tau, and National League for Nursing) often allow grants.gov/help/download_software.jsp). In fact, there
participants to chat online with presenters or authors is a trend for all manuscripts to be submitted online for
of certain articles on designated dates during scheduled print, online, or both. Online journals continue to grow.
In addition, there has been a rise in the number of open that decision support was feasible with the technology
access journals that give researchers more options for dis- support (Berry et al., 2010). Decision regret was signifi-
semination. Online journals have been discussed in the cantly influenced by personal characteristics and post-
nursing and academic community with mixed reception; treatment symptoms, although the P3P was not itself
while it allows the researcher-consumer of articles the abil- significant on the outcomes measured in the study (Berry,
ity to search wider publicly available science, other stake- Wang, Halpenny, & Hong, 2012). In another Web-based
holders in the publishing and academic worlds have been intervention, caregivers were randomly assigned to one of
concerned on the ramifications of this disruptive innova- two types of online support groups and compared to non-
tion (Broome, 2014). Regardless of the method of submis- active participants on their depressive symptoms, care-
sion and medium for publication, the published article may giver burden, and quality of life. In this study, both types
be incorporated into one or several online bibliographic of online support groups reduced depressive symptoms
retrieval systems. The researcher is not finished until the and improved quality of life over non-active participants
work is disseminated. (Klemm, Hayes, Diefenbeck, & Milcarek, 2014).
This chapter has summarized the processes of quanti-
tative and qualitative research and described select com-
Technology, Electronic Data, and Electronic
puterized tools that can assist the researcher in proposal
Documentation Research
preparation, data collection, data coding, data analysis,
and dissemination for both types of research. The fol- There are several different studies that highlight using
lowing section highlights examples for three categories electronic data and electronic health records in data min-
of research on computer use and nursing informatics in ing or care documentation.
(1) electronic data such as data mining large electronic
data sets and electronic nursing documentation; (2) Web- Secondary Analysis of Large Data Sets. Large public
based interventions; and (3) specialized computer appli- data sets are becoming more available to nurse research-
cations in clinical practice. The examples include both ers to explore health-related questions. The sites provide
quantitative and qualitative studies in which the nurse tutorials and assistance, making them more accessible
researchers inevitably used a variety of software tools in for secondary analyses. For example, the Agency for
the proposal development, data collection, measurement Healthcare Research and Quality (AHRQ) provides health
of variables, analysis, and dissemination activities. services investigators with the tools and data sources
for a variety of health-related systems. One data source
is the Medical Expenditures Panel (MEPS) database, a
multi-year set of large-scale surveys of families and indi-
EXAMPLES OF RESEARCH STUDIES viduals, their medical providers, and employers across the
Computers are inextricably tied to the process of conduct- United States (meps.ahrq.gov). MEPS is the most com-
ing research, but there are also good examples of research plete source of data on the cost and use of healthcare and
on computer use in the nursing literature. Several of the health insurance coverage (AHRQ, 2011). Another collec-
following examples also describe computerized processes tion used by a variety of nurse researchers is the HCUP
for conducting quantitative and qualitative research data from AHRQ. HCUP databases bring together the
approaches. These examples provide focus on nursing data collection efforts of State data organizations, hospital
research related to computer use and informatics as well associations, private data organizations, and the Federal
as using computers in the process of doing the research. government to create a national information resource of
encounter-level healthcare data. It includes the largest
collection of longitudinal hospital care data that enable
Clinical Interventions With Computers
research on a broad range of health policy issues, includ-
Over the past 10 years, Internet applications have been ing cost and quality of health services, medical practice
introduced into practice and tested in a variety of clini- patterns, access to healthcare programs, and outcomes of
cal trials aimed to improve conditions for patients. For treatments at the national, State, and local market levels
example, a team of researchers developed the Personal (hcup-us.ahrq.gov) (AHRQ, 2014).
Patient Profile-Prostate (P3P), a Web-decision support The electronic health records (EHRs) today are fre-
system for men newly diagnosed with prostate cancer quently providing source data for studies. These hospital
that assesses patients’ preferences prior to clinic visit and systems have been used in the literature. For example,
gives providers’ and patients’ information to aid decision- EHRs were used in a study done by Almasalha and col-
making among choices of treatment. The studies showed leagues (2012). They examined 596 episodes of care that
included pain as a problem on a patient’s care plan using patient’s bedside. Measures were developed to assess the
statistical and data mining tools to identify hidden pat- quality of the care plans and the participants’ evaluation
terns in the information on end-of-life (EOL) hospitalized of using the system. Data were analyzed revealing a sta-
patients. Findings suggested new understanding about tistically significant difference (p = .05) for the Microsoft
patient care, for example, EOL patients with hospital stays Access care plans completed by students and a statistically
less than three days were less likely to meet the pain relief significant difference in the students’ reports on using the
goals than those with longer hospital stays. Westra and system (p = .025). In a follow-up field study, the system
colleagues (2011) studied urinary and bowel incontinence was implemented on 49 students over three semesters
for home health patients using electronic health record who documented care for each patient they were assigned
data to predict improvements. In these cases, the EHR to in their clinical rotations. Students were assessed on
served as the data sources. their care plan documentation and they reported usabil-
Figar and colleagues (2011) used a private health ity and satisfaction with the computer-based documen-
information system (HIS) clinical documentation system tation system. Results continued to show that the coded
to study whether captured data could predict influenza language and PC care planning method was efficient and
outbreaks. They used a local interface terminology server effective (Mannino & Feeg, 2011).
which provides support through data auto-coding of clini-
cal records and analyzed specific data sets to compare the Web-Based Tools and Interventions. A significant body
burden of influenza in epidemiological weeks identified of research has been conducted on using the Internet as a
among 150,000 Health Maintenance Organization mem- tool for conducting research, as well as studies on Web-
bers in Argentina. The HIS detected the outbreak two designed interventions for clinical problems. For example,
weeks before the health department gave a national alert Yen and Bakken (2009) tested the usability of a Web-based
and was useful in assessing morbidity and mortality dur- tool for managing open shifts on nursing units. Using
ing the 2009 influenza epidemic H1N1 outbreak. observational and interview approaches, they evaluated
a Web communication tool (BidShift) designed to allow
Computerized Documentation of Nursing Care Plans. managers to announce open work shifts to solicit staff to
Moss and Saba (2011) studied the utility of costing out request their own work shifts. They used specialized soft-
nursing care with the CCC terminology on five most com- ware to capture screens and vocal utterances as partici-
monly executed interventions. Using an observation study pants were asked to think aloud as they completed three
of nurses performing routine care on an acute-care unit, subtasks associated with the open-shift management pro-
investigators collected data with a specialized data collec- cess. After task completion, they were asked about the
tion program entered directly into the PC database. A total process and their responses were recorded. Their data
of 251 interventions were observed, coded, and analyzed. were managed and coded using Morae, specialized soft-
From the analysis of time spent on each entered interven- ware developed for usability testing (www.techsmith.com/
tion, researchers could describe the four action types by morae.asp). This example of qualitative research reported
average cost and percent of activity. The study demon- participants’ patterns of use and themes related to their
strated the feasibility of valuating the nursing care given perceptions of usability of the communication tool.
to patients based on the standardized CCC terminology. In another qualitative study on electronic encoun-
In a randomized trial on electronic documentation of ters using Web-based videoconferencing, Nystrom and
nursing care plans, Feeg, Saba, and Feeg (2008) tested the Ohrling (2008) created a series of e-meetings for new
quality of nursing student care planning on a bedside per- fathers of children under one year old to “meet” in paren-
sonal computer (PC) using a standardized nursing termi- tal support groups. The technology allowed both one-on-
nology in a specially designed Microsoft Access database one and group encounters. The fathers were interviewed
program (Clinical Care Classification System) compared using a narrative approach and content analysis was
with an open-text format type-in application with the applied to the interview data. The researchers identified
same terminology. Students were randomly assigned to three categories from the transcripts: (1) being unfamiliar
one of the two versions of electronic nursing documenta- and insecure talking about fatherhood, (2) sharing experi-
tion formats and interviewed two simulated patients who ences and being confirmed, and (3) being supported and
served for all of the participants. The simulated patients limited by the electronic encounters.
were interviewed about their symptoms: one presented Andersen and Ruland (2009) studied an Internet-based
with congestive heart failure, the other with pneumonia. online patient–nurse communication (OPNC) service to
Students were instructed to document the care immedi- support patients with prostate and breast cancers. Using
ately following the interview on a laptop stationed at the qualitative content analysis, they examined 276 messages
in a tailored Internet support intervention over 15 months. using simple cell phones and sophisticated smartphones.
Two main themes emerged: (1) concerns about physical The MyMediHealth (MMH) is a medication management
symptoms and treatment side effects and (2) worries and system that includes a medication scheduler, a medication
questions about treatment and follow-up. They concluded administration reminder engine and sends text messages
that the OPNC service can meet patients’ needs for advice to patient phones (Stenner, Johnson, & Denny, 2012).
and information, thus improving the quality of care. In a review done by Schroeder (2009), a variety of
In a qualitative study by Lichenstein, McDonough, studies used computerized telephone technologies as an
and Matura (2013), 98 participants who self-identified assessment tool specifically for the collection of daily,
as caregivers for a person with pulmonary hypertension self-reports of HIV risk behaviors. The review presented
(PH) engaged in an online discussion board posted by the advantages of applications of interactive voice response
Pulmonary Hypertension Association over an 18-month technology (IVR) to HIV risk behavior research, including
period. Clinical variables collected were medications and feasibility studies, assessment mode comparisons between
oxygen use, and years since diagnosis. Thematic analysis IVR and alternative self-reporting methods, and unique
yielded four themes: fear and frustration, questions and findings derived from event-level data analyses illuminat-
concerns, someone to listen to, and moving on with life. ing risk factors for unprotected intercourse on within-
Results showed that caregivers of people with pulmonary person level. The author concluded that these specialized
hypertension may be ill equipped to care for their loved interactive, computerized voice systems are highly prom-
one because of lack of knowledge or psychological distress. ising tools for various research and healthcare applica-
In a review of Web-based cognitive behavioral interventions that should be considered more frequently for use in
tions for chronic pain, researchers conducted a systematic HIV-risk populations.
review and meta-analysis to quantify the intervention effi- Some studies that incorporate remote technologies
cacy for treatment of patients with chronic pain. Using 11 in special applications have been described in the litera-
studies from MEDLINE and other data sources, the inves- ture. Mahoney (2011) reported on several mixed methods
tigators found that Web-based interventions for chronic studies that describe innovative monitoring intervention
pain resulted in small pain reductions in the intervention research with older adults and their informal and/or for-
groups compared with waiting-list control groups (Macea, mal caregivers. The studies were conducted in “real world”
Gajos, Calil, & Fregni, 2010). homes and focused on improved ways to monitor elders,
taking into account the goal of independence and privacy,
Specialized Computer Applications in Clinical Care. to alert caregivers on elders’ activities and safety issues. In
Computer-based administration of assessment is a reli- another collection of case studies, Rantz and colleagues
able means of collecting patient assessment data as (2010) describe instances of technology-enhanced moni-
demonstrated by the report by Wilkie and colleagues toring capabilities, using sensors, alarms, and environ-
(2009) who evaluated the feasibility and acceptability of a mentally embedded devices, that wirelessly communicate
pentablet-based software program that assesses patients’ with programmed systems for the purpose of detecting
cancer-related symptoms, the PAINReportIt . In the potential problems.
study, 131 patients were able to use the computerized These and other innovations that connect computer
tool and reported high acceptability scores. technology with nursing practice have emerged with
Vawdrey and colleagues (2011) pilot tested a tablet research to support their functionality, ease of adoption,
computer application for patients to participate in their and efficacy. The range of innovation has been astound-
hospital care. A prototype application was developed ing as the technologies have increasingly become less
for a tablet computer using EHR queries and updates in expensive, smaller, wireless, and now interconnected
real time. Patients were invited to participate in the study with cloud computing. Nursing research and computers
after consultation between the patient’s providers and today are inseparable in both areas of using computers
the investigator, an attending cardiologist. They were for the research and studying the impact of computers on
given an iPad device and encouraged to use the applica- patient care.
tion. Structured interviews were used to test the patient
engagement and the tablet usefulness of having access to
patient’s own medication and hospital history. SUMMARY
Mobile technologies have also proliferated in health-
related applications today. For example, systems that This chapter has reviewed two research paradigms and
support medication management of patients with SMS philosophical orientations—qualitative and quantitative
texting and Web-based interface programs have emerged methods—that specify different underlying approaches
c47.indd 681 20-02-2015 19:04:34

to research and the use of computers in various stages REFERENCES
of the research process. It also highlighted examples of
research on computer use with these quantitative and Agency for Healthcare Research and Quality. (2011). Medical
qualitative approaches. A variety of computer applica- Expenditure Panel Survey. Rockville, MD: Agency for
tions are available via commercialized packages that serve Healthcare Research and Quality. Retrieved from
www.meps.ahrq.gov/mepsweb/
the nurse researcher in conducting research. With the
Agency for Healthcare Research and Quality (AHRQ). (2014,
advancements in applications development and deliv- January). HCUP Home. Healthcare Cost and Utilization
ered wirelessly and the wide use of smaller devices such Project (HCUP). Agency for Healthcare Research and
as smartphone technologies, the potential for innovation Quality, Rockville, MD. Retrieved from www.hcup-us.
and research on health abounds. Prompted by large con- ahrq.gov/home.jsp
sortia that foster technology innovation such as Health Almasalha, F., Xu, D., Keenan, G., Khokhar, A., Yao, Y.,
Datapalooza (healthdatapalooza.org), an annual confer- Chen, Y., … Wilkie, D. (2012). Data mining nursing care
ence for researchers and entrepreneurs, one can expect plans of end-of-life patients: A study to improve health-
the emergence of many evidence-based interventions care decision making. International Journal of Nursing
aided by smartphones, tablets, and minis in the near Knowledge, 24(1), 15–24.
Andersen, T., & Ruland, C. (2009). Cancer patients’ questions
future. With the federal government emphasis on big
and concerns expressed in an online nurse-delivered mail
data, the enhancements made to aid researchers to tackle service: Preliminary results. In K. Saranto M. Tallberg, A.
existing data sets, and the cloud to house and connect Ensio, P. Flatley, & H. Park. (Eds.), Connecting health and
researchers and subjects from distances, one can expect humans. Fairfax, VA: IOS Press.
a proliferation of research techniques and acceptable evi- ATLAS.ti 7. (2014). Thousand Oaks, London: SAGE.
dence using secondary data. In addition, research on com- Retrieved from http://www.atlasti.com
puter applications and informatics is a growing body of Bakken, S. (2013, December 3). Why a nursing terminology?
science that will continue to appear in the literature with Presentation at the CCC Workshop, Nashville, TN.
a renewed purpose of collecting and storing meaningful Bakken, S., Stone, P., & Larson, E. (2008). A nursing informat-
data for multiple uses. ics research agenda for 2008–2018: Contextual influences
and key components. Nursing Outlook, 56, 206–214.
Carrington and Tiase (2013) state, “Health informa-
Banxia. (2014). Decision Explorer. Banxia Software Ltd,
tion technology (HIT), and more specifically, nursing Kendal, Cumbria, UK. Retrieved from http://www.
informatics, is the driving force for the transformation of banxia.com/dexplore
health care in the United States…Nurses’…role in the use Berger, A. M., & Berger, C. R. (2004). Data mining as a tool
of technology [has evolved]” (p. 136). Nursing informat- for research and knowledge development in nursing.
ics research has been influential in shaping health. In a Computers, Informatics, Nursing, 22(3), 123–131.
review of literature published in one year, Carrington and Berry, D. L., Halpenny, B., Wolpin, S., Davison, J., Ellis, W.,
Tiase describe topics ranging from EHR, monitoring, and Lober, W. B., … Wulff, J. (2010). Development and evalu-
human–computer interaction to bar code administration. tion of the personal patient profile-prostate (P3P), a
The studies were 31% qualitative, 54% quantitative, and web-based decision support system for ment newly diag-
nosed with localized prostate cancer. Journal of Medicine
15% mixed methods.
Internet Research, 12(4), e67.
Conclude that the design, implementation, and evalua- Berry, D. L., Wang, Q., Halpenny, B., & Hong, F. (2012).
tion of nursing technologies show promise in nursing effi- Decision preparation, satisfaction and regret in a mult-
ciencies that affect patient safety and healthcare cost, the center sample of men with newly diagnosed localized
two driving forces of policies that shape healthcare. prostate cancer. Patient Education Counseling, 88(22),
The contextual influences on nursing research are 262–267.
discussed by Bakken and colleagues (2008). The nursing Biostat. (2014). Power and precision. Englewood, New Jersey.
informatics research agenda for the next 10 years must Retrieved from http://www.power-analysis.com/home.htm
expand users of interest to include inter-disciplinary Borenstein, M., Hedges, L., Higgins, J., & Rothstein, H.
researchers; build upon the knowledge gained in genomic (2009). Introduction to metaanalysis. Chichester, UK:
John Wiley & Sons. Retrieved from http://www.meta-
and environmental data; guide the re-engineering of
analysis.com
nursing practice; harness new technologies to empower Brennan, P. F., Zielstorff, R. D., Ozbolt, J. G., & Strombom, I.
patients and their caregivers in collaborative knowledge (1998). Setting a national research agenda in nursing
development, particularly related to the social network informatics. Medinfo, 9, 1188–1191.
technologies; and evaluate innovative methodologies that Broome, M. (2014). Open access publishing: A disruptive
recognize human–computer interface factors. innovation. Nursing Outlook, 62(2), 125–127.

Nursing Informatics Chapter 47

Uploaded by

Copyright:

Available Formats

Nursing Informatics Chapter 47

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nursing Informatics Chapter 47

Uploaded by

Copyright:

Available Formats

47

c47.indd 663 20-02-2015 19:04:31

c47.indd 664 20-02-2015 19:04:31

c47.indd 665 20-02-2015 19:04:31

Paper and Pencil Questionnaires. Paper and book-

c47.indd 671 20-02-2015 19:04:32

c47.indd 672 20-02-2015 19:04:32

c47.indd 674 20-02-2015 19:04:33

c47.indd 676 20-02-2015 19:04:33

c47.indd 681 20-02-2015 19:04:34

You might also like