Structuring Advanced Practice Knowledge: Curricular, Practice, and Internet Resource Use


28


Structuring Advanced Practice Knowledge: Curricular, Practice, and Internet Resource Use



Mary Ann Lavin



ACKNOWLEDGMENTS



The author wishes to acknowledge the contributions of: Dr. Joanne Thanovaro, Saint Louis University Association Dean of Graduate Nursing Education for recommending that technology become an integral part of the entire advanced practice nursing curriculum; Dr. Eileen Healy, an advanced practice nurse and informaticist, for her early on support and editing. Sangeeta Agarawal, RN, MSN, BSE, CEO and Founder, Helpsy, Berkeley, CA, a holistic scientist, CS engineer, and award-winning inventor for her AI nurse story; and to Kirk O’Donnell, a board certified family care physician and a board certified flight surgeon, a biotech and medical entrepreneur, and medical advisor to the Cambridge, Massachusetts-based firm, MEETA, for our many COVID-19 PCR and antibody testing discussions.


ABSTRACT



The American Nurses Association text, Nursing informatics: Scope and standards of practice, defines nursing informatics as “the specialty that integrates nursing science with multiple information and analytical sciences to identify, define, manage, and communicate data, information, knowledge and wisdom in nursing practice” (2015, pp. 1–2). This definition delineates the meaning of informatics for the nursing informatics specialist. This chapter, however, is devoted to integrating selected informatics technologies into the educational infrastructure not of the clinical informatics courses but across the curriculum. It examines four curricular threads (nursing theory, practice, research, and legal implications), five selected technologies (social media, wearables, telehealth, structured, and unstructured data applications), and two specialty areas (advanced practice and clinical informatics). It is our hope that, by reading this chapter, advanced practice nurses will heighten their interest in new technologies, embrace innovative thinking, and develop creative applications, making them effective collaborators within the informatics and technology world. It is also hoped that educators will consider restructuring advanced practice nursing curricula, not only integrating technologies across the curriculum but also exploring the development of dual-degree BSN-DNP and Engineering programs. Finally, during this age of emerging global infections, it is hoped that all readers gain insight into the structuring of discreet knowledge guiding: (1) nuanced clinical decision-making rules and (2) the statistical tests deemed essential in the selection of valid diagnostic tests.


INTRODUCTION



Advanced practice nurses are expert clinicians, who provide evidence-based care. Ready access to evidence relies on ready access to accurate information. This chapter examines three sources of information (social media, wearables, and telehealth), two types of data (structured and unstructured), and concludes with suggestions on the restructuring of advanced practice nursing curriculum. However, restructuring the curriculum is not the only way in which multidimensional modeling may be used by students, practitioners, and faculty. Searches may also be structured multidimensionally (Systems and methods for multi-dimensional computer-aided searching, 2017). See Fig. 28.1; each cell represents a unique search. For example, the A1a cell represents a search filter composed of intersecting terms: social media AND clinical nursing AND unstructured data. This may be used within any search engine database, including those of the National Library of Medicine (NLM) PubMed and PMC databases. This A1a advanced search filter is visible within the PubMed search box, once the hyperlink is clicked/activated (Fig. 28.2).


Images


• FIGURE 28.1. Knowledge Object as Multidimensional Search Framework. (Source: TIIKO, patent pending.)


Images


• FIGURE 28.2. PubMed citation retrieval image, National Center for Biotechnology Information, National Library of Medicine, https://pubmed.ncbi.nlm.nih.gov/?term=social+medical+and+clinical+nursing+and+unstructured+data July 12, 2020


Figure 28.1 contains 24 (4×3×2) cells, each containing a unique search. TIIKO software (patent pending) is designed to conduct all 24 searches obtained with one click. Additional examples are available at TIIKO (2018). Results are visualized in what is called a knowledge object. In other words, knowledge may be gleaned from accessing and learning from the intersecting information, made available in the multidimensional display. There is another philosophically based reason for calling the visualization a knowledge object. Multidimensionality confers context and context confers meaning to the search results obtained. Context and meaning imbued information yields knowledge. For additional insight into meaning and knowledge, see the work of Mattingly, Lutkehaus, and Throop (2008).


There are two caveats that concern the construction of a multidimensional search system. First, implementing such a system and method, capable of performing 1…n searches with one click, will require high-performance computing, accessing 1…n searches per minute and caching data to conduct the TIIKO searches without interfering with the operations of the original database. Second, multidimensional searches are not automatically constructed. Rather the user identifies the metatopic, domains, and subdomains (within-domain) categories. This information, predefined and entered by the user, serves not only to organize the searches but to give shape to the visualization.


Uses of Multidimensional Modeling


Citation searches are not the only applications for multidimensional search software. One other example is retrieval of diagnostic, intervention, and/or outcome-specific narrative text from 1…n electronic medical/health records. Advertisers or research recruiters may want to use the framework within a public database to identify potential buyers or research participants. Computational biology and pharmacogenomic industries may use multidimensional search models to retrieve saved documents from within a standalone database. As an example, see Fig. 28.3. It illustrates the retrieval of microscopic images showing varying degrees of cellular absorption over 24 hours, given two different drug compounds and two drug compound concentrations. The variety of examples presented in this one paragraph indicates that multidimensional searches of unstructured data allow for flexible, user predefined retrieval capability.


Images


• FIGURE 28.3. Example of TIIKO Multidimensional, Unstructured Data Search Model for Retrieval from a Standalone, Nonrelational Document Database of Microscopic Images of Cellular Absorption at Hours 0–24 of Two Different Drug Compounds at Two Different Drug Concentrations. (“http://www.TIIKO-IT.comwww.TIIKO-IT.com methodology used to develop this figure.)


Multidimensional modeling is conceptual modeling as well, where the term “thinking” may be used in lieu of the term “modeling.” This type of thinking goes beyond that undergirding “what is” and “how to” queries. Multidimensional modeling reflects the kind of thinking required by learning today in fields such as big data analytics, given its data variety, velocity, and volume. It is also the kind of human-computer interactional thinking that serves as a bridge to machine learning and artificial intelligence. For more information on investigations of various kinds or styles of thinking, visit the homepage of Harvard Professor Krzysztof Gajos (2019). You may also want to have access to his Lab in the Wild.


The multiple ways that we think helps us gain insight into the DIKW nursing framework, where DIKW stands for data, information, knowledge, and wisdom (Matney, 2013; Matney et al., 2015). These components serve as the basis for the definition of nursing informatics as “the specialty that integrates nursing science with multiple information and analytical sciences to identify, define, manage, and communicate data, information, knowledge and wisdom in nursing practice” (Matney et al., 2015, pp. 1–2). Although these components are represented graphically as a two-dimensional, layered model, they may also be considered more dynamically within a Tesseract, a four-dimensional model, using a multifunctional Boolean mathematical approach (Systems and methods for multi-dimensional computer-aided searching, 2017). Imagine one rendition, where data is the fourth dimension available, on a selective basis, to the myriad of interactions that occur among the other three dimensions: information, knowledge, and wisdom. Such a mathematical approach may facilitate its application in big data analytics.


STRUCTURED AND UNSTRUCTURED DATA



Readers new to the terms “structured data” and “unstructured data” may find their definitions and related terminology useful (Table 28.1). Becoming comfortable with such terms requires exposure. For that reason, the terms were linked not only to their use within computer science and computing in general but also to their use within nursing and nursing literature. As with any new language or terminology, definitions of terms and examples of use only go so far. Actual use is required. So, take time to discuss structured and unstructured data with a colleague. Attempt to convey their meaning to someone who is knowledgeable in the field. Say something like, “I am learning about structured and unstructured data. My understanding of the term x is [fill in the blank] . Is that your understanding, too?” In brief, learning any new language/terminology requires reading, hearing, and speaking. Using this approach, it is highly likely that you will not only learn the terms but become competent in their use.



TABLE 28.1. Structured and Unstructured Database Glossary and Related Terms, Organized within Terminology Clusters


Images


Structured Data and Clinical Applications


One example of the clinical applications of structured data within a relational database is the CRIB database, developed at CRIB, the Washington University—St. Louis Center for Research Innovation in Biotechnology (Washington University—St. Louis, Center for Research Innovation in Biotechnology, n.d.). Three screenshot iterations are provided in Figs. 28.4, 28.5, and 28.6). Figure 28.4, the first screenshot, explains the database. When the button “New Molecular Entities” in Fig. 28.4 is activated, the screenshot presented in Fig. 28.5 appears. Note that the first column represents generic drug names. When “Influenza Vaccine, Adjuvanted” is clicked, the next screen opens (Fig. 28.6), which provides information on the influenza vaccine with the brand name of Fluad. Developed and manufactured by Novartis, it was approved by the FDA for use in the United States on November 24, 2015.


Images


• FIGURE 28.4. Screenshot Indicating Access to New Molecular Entities, within a Center for Research Innovation in Biotechnology Database (CRIBdb) Housing Structured Data (Retrieved on January 8, 2019 from http://cribdb.wustl.edu/).


Images


• FIGURE 28.5. Screenshot Showing CRIB Database Distribution of Generic Drugs by Approval Date, Type, Indications, and Target. (Reproduced, with permission, from Michael Kinch, PhD, CRIB.)


Images


• FIGURE 28.6. Screenshot Showing Partial Cribdb Background Information Available On Influenza Vaccine, Adjuvanted. (Reproduced, with Permission, From Michael Kinch, Phd, Crib.)


Michael S. Kinch, PhD, founded the Center for Research Innovation in Biotechnology (CRIB) and Drug Development (CCD) at Washington University in St. Louis (Kinch, n.d.). The mission of CRIB is to “determine the sources of scientific and medical innovation in drug and vaccine development” (Washington University—St Louis, Center for Research Innovation in Biotechnology, n.d.). Its database (CRIBdb) allows researchers and industry professions to track not just the background on one drug, but to examine the history of drug innovation for one or more groups of drugs. An article outlines this approach and serves as a user case scenario in this regard (Griesenauer & Kinch, 2017). It traces the history of vaccine development from Jenner’s development of the smallpox vaccine in 1796 to trends in vaccine development today and their related implications for vaccine development tomorrow.


Some features of the CRIBdb may be applied to advanced practice. First, it is a structured database with a limited amount of data, easily searched and user friendly. Second, the database highlights the value of knowing the history of innovations. Third, once users know the history or market trends of any one intervention or innovation, it is easier to plan for future development needs. These features are applicable outside the realm of pharmaceuticals. If advanced practice nurses looked at the historical development of the technologies associated with an advanced practice intervention, then they may see a trend and more readily grasp the potential for creating and testing a newer technology. Such newer technologies may include use of sensors, robots, telemetry, or applications still lying fallow in our subconscious.


Unstructured Data and Clinical Applications


Highly organized structured data differs from unstructured data. Some terms associated with the term unstructured data are narrative, text, natural language, or natural language processing (NLP) data. Of these terms, narrative and text data more closely convey the meaning of unstructured data. This is because both structured, relational databases use natural language to label column and row data and often, as in the case of the CRIBdb, to fill in the cells. These natural language terms are, however, fixed with regard to their placement within a structured database.


In contrast, natural language, as that used within a narrative account or within a text, is unstructured. Placement is not fixed. Commonly used forms of unstructured data are text messages themselves and social media communications, e.g., those on Facebook, Twitter, Pinterest; or, the digitalized narrative text in books, articles, poetry, and diaries. Within healthcare, unstructured data include digitalized progress notes, nursing comments/notes, consultation reports, imaging reports, admission history and physical examination reports, incident reports, discharge summaries, and digitalized patient communications with providers whether by e-mail, text, or an e-chart client/patient-to-provider messaging platform.


Unlike the relative ease with which data is retrieved from structured databases, unstructured data retrieval is more problematic. The reasons are several. First, some words or phrases are abbreviated, such as myocardial infarction (MI), range of motion (ROM), and increased intracranial pressure (↑ ICP). Other words have multiple synonyms, such as pain, discomfort, uncomfortable, hurt, ache, throbbing, and cramp. Some words spell and sound the same but have different meanings, such as digits or digital. As a result of these complexities, retrieval of unstructured data within its original context is difficult. Consequently, considerable amounts of nursing and medical information and knowledge lie hidden within the electronic health record (EHR) (Monga & Singh, 2018, October 18).


The NIH National Center for Biomedical Computing funds a few institutions that host structured and unstructured healthcare data for research purposes. One is the Harvard Center of Informatics for Integrating Biology and the Bedside, also known as the i2b2 Informatics Center.


Its “About Us” page provides a description of the center (I2b2, Informatics for Integrating Biology & the Bedside, 2019). In brief, the center recognizes the bidirectionality of biologic and bedside data, the value added by interdisciplinary research, and the need to integrate biologic and bedside data to advance both research and clinical care. Both structured and unstructured data retrieval strategies and analyses are employed. In addition, the i2b2 center allows eligible researchers from the broader community access to several datasets of deidentified, unstructured datasets, e.g., discharge summaries (i2b2, Informatics for Integrating Biology & the Bedside, 2019).


Discharge summary data contain cues to identifying high readmission risk patients, who are also patients with high care coordination need. One way to identify high readmission risk is by disease diagnoses, e.g., stroke, uncontrolled hypertension, and heart failure (Feldman, McDonald, Barrón, Gerber, & Peng, 2016; O’Connor et al., 2015; Feldman et al., 2015). Experimental care coordination interventions aimed at reducing readmission rates among such patients have yielded inconclusive results, however. It may be that readmission risk and care coordination need are not sufficiently defined by disease diagnosis alone. Symptom severity and nursing sensitive indicators also contribute. These considerations led to the creation of a multidimensional High Care Coordination Need Model (Fig. 28.7). It is a theoretical design to retrieve i2b2 discharge summary documents, distributing them among the 12 (3×2×2) cells of the knowledge object. The particular knowledge object consists of the following three domains and their within-domain categories: (1) primary discharge diagnosis (CHF or T2DB); (2) disease-related risks (hyper-/hypoglycemia, peripheral edema, pulmonary edema); and (3) nursing sensitive indicators (high fall risk and nosocomial infection). With regard to disease-related risk, an assumption was made that not every sign or symptom a patient experienced during a hospitalization would have made it into a discharge summary but only the more severe ones; and these, therefore, are the disease-related issues that place the patient at greater readmission risk.


Images


• FIGURE 28.7. A TIIKO Multidimensional Search Method and System Model for Stratifying Risk and Care Coordination Need. (www.TIIKO-IT.com methodology used to develop this figure.)


The observant reader will have noted that the way in which the High Care Coordination Model (Fig. 28.7) is structured will result in some discharge summaries appearing in more than one cell. For example, the patient with Type 2 diabetes mellitus may not only have hyperglycemia but also peripheral edema and a nosocomial infection and be a high fall risk. In fact, the greater the number of cells in which a patient’s discharge summary appears, the greater the readmission risk and the higher the discharge care coordination need; and the total number of duplicate discharge summaries per patient may be rank ordered and high care coordination resources allocated accordingly. With a more precise measure of readmission risk and care coordination need and with more highly targeted resources, perhaps the readmission rates will decline. Although theoretical, this multidimensional High Care Coordination Model possesses face validity and awaits TIIKO software development and testing.


SOCIAL MEDIA



There are many doors or portals to information. Social media (SM) is one. This section examines five aspects: (1) How advanced practice nurses may seek evidence to answer SM-related questions; (2) social media, information, and misinformation; (3) online support groups; (4) social media and research recruitment; and (5) advanced practice nurse SM presence.


Seeking Evidence for SM-related Questions


Not infrequently, advanced practice nurses are asked to give their opinion on information clients gain through social media. Advanced practice nurses should not take this responsibility lightly. When they are uncomfortable in answering, they need to (1) indicate their discomfort and (2) refer the client to someone qualified in the practice area in question. When comfortable with the subject matter, they need to provide an evidence-based answer to the client.


Let’s take, as an example, a relatively new procedure: Deep brain stimulation to control Parkinson’s disease tremor. A patient, whose mother has Parkinson’s disease, asks a nurse practitioner to comment on the following Facebook video: https://www.facebook.com/enda.creaven.9/videos/10156755290113156/ (Creaven, 2018). A message that accompanies the video states, “A life-changing device made in Clonmel, Ireland by Boston Scientific. #Parkinson’s Disease.” The nurse practitioner recognizes that the video and its accompanying comment provide the viewer with a portal or gateway to device information, specifically: the manufacturer and the country where the device was made. Given this information, the nurse next formulates questions or queries that are basic to any inquiry about any new drug or device:


•   What is the product name?


•   Is it FDA-approved for use in the United States?


•   Is it safe and effective?


•   Do data indicate that the benefits outweigh the risks?


For any manufactured medical device, Table 28.2 provides a method for finding answers from the FDA. The method consists of establishing a set of queries. For each query, the steps are determined to answer the query. Finally, the query-specific answers are recorded. The advanced practice nurse now not only has a guide to direct the conversation with the client, but also has a record of the queries posed and the sources of information/evidence used to answer the queries. Because the device is only recently approved in the United States, it is also wise to add:



TABLE 28.2. Queries Designed to Evaluate Manufacturer’s Evidence Presented to the FDA


Images


•   There are several risks associated with the device.


•   The two manufacturer studies upon which the device was approved are well designed.


•   While the results indicate that benefits seem to outweigh risks, both the benefits and the risks need to be communicated thoroughly before any decision is made to proceed with the procedure.


The nurse practitioner or advanced practice nurse also knows that all new product safety and effectiveness information based on manufacturer research, albeit FDAapproved, may be subject to change, given subsequent clinical experience with the product after it is released as well as information derived from additional clinical trials/research. Given these insights into the product, the clinician and client next decide if more detailed information needs to be reviewed together and, if so, when.


The issue of FDA approval of medical devices becomes more complex with devices being sold outside of the United States. The issue is this: Not all U.S. manufactured medical devices are FDA-approved for use in the United States. A related issue is, even though not FDA-approved for use in the United States, a product may be FDA-approved for export for use in other countries, given varying regulatory standards among countries. Still, all buyers and all patients, in the opinion of this author, they have a right, to know whether a U.S. manufactured product is also approved for use in the United States. For additional information on FDA export policies, see the FDA Web site on the exporting of medical devices (U.S. Food and Drug Administration (FDA) | WCMS, 2018, September 27).


Social Media, Information, and Misinformation


As the reader is aware, social media platforms are readily available and frequently used (Murnane, 2018, March 3). These platforms expand the availability of not only information, but also misinformation. Let’s take vaccine safety as an example. The purpose of this subsection is to show the reader how the same question may yield opposing information as a function of the algorithm the SM and/or search engine uses to answer questions.


Let’s take the question, “Are vaccines safe?” We will answer the same question on three platforms: Facebook, Twitter, and Google. Looking first at Facebook, the first postings delivered within “my” Facebook page were five advertisements. Of the five, the first four were antivaccine and the last was a vaccine-supportive advertisement from the Australian Government Department of Health. While the Centers for Disease Control and Prevention (CDC) have its own CDC account and pages on Facebook, it seems that vaccine questions in the Facebook search box do not trigger a CDC response, at least not on my page. While it is recognized that social media has a role to play in public health, in promoting individual and population health, and in preventing disease, a systematic review of systematic reviews on the effectiveness of social media in delivering these messages was inconclusive (Giustini, Ali, Fraser, & Boulos, 2018). Given the amount of misinformation available through social media, it is imperative that health professionals regularly pass the word about vaccine safety, advocate for full immunization, and take a more proactive role in confronting misinformation on Facebook and other social media sites. One way is to simply share CDC evidence-based messages, e.g., “Measles: Make Sure Your Child is Fully Immunized” (Centers for Disease Control and Prevention, 2019). In this way, each health professional, as an individual social media user, assumes direct responsibility for increasing social media health promotion and disease prevention effectiveness.


Looking next at my account on Twitter, the same query was entered on the same day: “Are vaccines safe?” No advertisements for or against vaccines appeared. Of the first five Tweets available, four supported vaccine use. When following up on one of the tweets, I lost my original page. This time, when I resubmitted my query, I inadvertently skipped the question mark at the end of the query. The responses led off with an antivaccine advertisement and two-thirds of the first few tweets were antivaccine. The only difference in response sets was the presence or absence of a question mark. Apparently, the IT developers responsible for selecting queries most likely to be used by antivaccine proponents assume they are grammatically challenged or indifferent; or, perhaps think that those most susceptible to antivaccine ads are.


Of course, the platform that most use to answer questions is not Facebook or Twitter, but Google and other search engines. Assume, again, a client asks the same question of a nurse, “Are vaccines safe?” One response may be to google the question for the client and show the client the results. Search engine results, even Google’s, are biased, however. Why? Because a search engine algorithm is at play. These algorithms take into account the history and, hence most importantly, the preferences of the user. As a result, search engines provide the authors with pro-vaccine Web site results when its algorithm identifies users as pro-vaccine, whereas it will provide a user it identifies as antivaccine with antivaccine Web sites.


Two questions arise. The first is: What is the best way to evaluate a Web site to avoid misinformation? There are ways to evaluate content. Government (.gov) and education (.edu) domains are advertisement free and hence less subject to bias. Furthermore, there are specific criteria to evaluate Web sites, whether posted on .gov, .edu, .org, or .com domains or any social media site. These criteria are accuracy, authority, objectivity, currency (how current is the message or the data presented), and coverage or, what Cornell, calls “putting it all together” (Cornell, 2019).


The second question is: What are the ethical considerations of social media tracking or surveillance? The answer is complex. The following Web sites provide different perspectives as well as an overview of this aspect of social media: tracking social media algorithms, social media and policing, and ethics in a digital surveillance world (McCourt, 2018, April 3; Bousquet, 2018, April 27; University of Maryland, Robert E. Smith School of Business, 2018, September 27).


Online Support Groups


Social media is a useful portal for supporting people with health challenges, especially chronic illness. Nurses may be asked to suggest or evaluate online support groups. Well-established, nonprofit organizations (.orgs) led by healthcare professional teams as well as for-profit businesses associated with reputable nonprofits with input from health professionals are recommended over informal support groups on SM sites, e.g., Facebook and/or Twitter, which have no official relationship to healthcare organizations.


When discussing online support with clients, point out distinguishing characteristics. For example, there are online organizations that provide professionally led, evidence-based health information. These include the American Heart Association (2019) and the American Cancer Society (2019).


There are online organizations that are more grass roots or community development-oriented in structure. They strategically place healthcare professionals within leadership positions even as they reach out directly to support and improve the quality of life for individuals and families affected. Examples include Lupus Foundation of America (2019) and NAMI, an acronym for the National Alliance for Mental Illness (2019). The former hosts an online LupusConnect™ community for members to join; however, the connection is with Inspire (Inspire, Health and Wellness Support Groups and Communities, 2019).


Inspire.com hosts multiple support groups with more than 1,000,000 members, including patients and caregivers (Inspire, Why Inspire? The Leading Social Network on Health, 2019). Some Inspire groups are devoted to little understood conditions, e.g., Ehlers–Danlos syndrome. This particular online group boasts 69,026 members, worldwide. Because Ehlers-Danlos is not a commonly diagnosed syndrome, members often learn useful information about their own condition from other Inspire members in a manner analogous to which clients with more common illnesses learn about the illness from the experience of members of their own family or community. In fact, Inspire online support groups refer to themselves as online communities, sharing the same interests, concerns, and accrued knowledge.


When discussing online support groups with clients, be sure the client realizes rules governing interaction are important. Inform the client to determine if the site has ground rules in place (e.g., no spam nor selling, no bullying) and maintain respect for members and member confidentiality, and value each member’s contributions. Do educate clients to avoid sites where trolling or flaming are allowed or go unaddressed. Trolling is the “deliberate provocation of another”; and, flaming is “mocking or encouraging deliberate self-harm,” both of which have been identified as potential social media harms (Dyson et al., 2016).


Agencies sponsoring online support groups rightly post disclaimers, notifying clients that the sponsoring agency is not responsible for any advice given by members to other members. Such disclaimers usually make clear that the notice includes any and all medication, treatment, diet, or any other healthcare recommendations.


Social Media and Research Recruitment


According to the Pew Research Center, 69% of all Americans use at least one social platform, which is an increase from 50% in 2011 and 5% in 2005. By far the most popular platform is Facebook, with other platforms attracting similar but smaller proportions of the population (Pew Research Center, Internet and Technology, 2018, February 5). SM platforms appeal to researchers, including those in the healthcare industry. The databases provide access to potential research participants; and, there are results indicating SM is a promising strategy to recruit research participants (Gelinas et al., 2017). These same authors discuss the ethics of recruitment, primarily privacy and transparency, from the standpoint of traditional and online recruitment methods. Disclosure of research presence and purpose as well as informed consent is as basic to SM research as it is to research in more traditional settings. If there is a question of the degree of disclosure required between public and private groups, err on the side of caution. Also, the author assumes that online recruitment does not operate under a set of exceptional ethical standards. If there are differences in online recruitment that require additional ethical consideration from the point of view of a researcher or an Internal Review Board (IRB), then the respective researcher or ethics committee needs to assess the online format and draw up additional ethical guidelines that cover the online differences.


Some may ask: What are the differences that may occur with online recruitment that do not occur with traditional recruitment? One is the contract established between the SM platform and its users. Some SM platforms in their terms of service do not give third-party users access to data. Others do. Under “Researchers and Academics,” Facebook states:


We also provide information and content to research partners and academics to conduct research that advances scholarship and innovation that support our business or mission and enhances discovery and innovation on topics of general social welfare, technological advancement, public interest, health and wellbeing (Facebook, Academic Programs, 2019).


Facebook’s research instructions are available online (Facebook, Data Policy, 2019). However, doctoral students or faculty seeking to conduct their own healthcare-related research using Facebook would do well to discuss their own research-specific privacy concerns and ethics directly with Facebook, Inc. (ATTN: Privacy Operations, 1601 Willow Road, Menlo Park, CA 94025) before discussing Facebook’s policy with the university’s Internal Review Board.


Another element in social media research is cost. In Table 28.3, it is obvious, based on willingness to pay, that the acne group researchers were seeking a larger group/sample size than the emphysema group researchers (MD Connect, 2018). While cost per referral was less on Facebook than on Google, the number of referrals per the two platforms varied as a function of the age group involved. Assuming that younger people use Facebook at greater frequency than older people, the finding that Facebook yielded more acne group referrals than Google is logical. Conversely, Google yielded more emphysema group referrals than Facebook. Both groups are costly. Still the cost may be less than traditional methods with its high recruitment staff costs plus high loss to follow-up rates. In addition, SM recruitment may be less subject to bias as compared with the narrower institutional or geographic base from which research participants are more traditionally drawn.



TABLE 28.3. Comparison of Research Recruitment Costs between Google and Facebook for Two Different Research Group Sizes, Two Different Time Periods, and Two Different Conditions (MD Connect, 2018).

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jul 29, 2021 | Posted by in NURSING | Comments Off on Structuring Advanced Practice Knowledge: Curricular, Practice, and Internet Resource Use

Full access? Get Clinical Tree

Get Clinical Tree app for offline access