Jngrau_telemedicine At Shriners Slc

University of Bath Royal College of Surgeons of Edinburgh Healthcare Informatics

Crossing the Rough Road and Hot Sands via Telephony: An Analysis of Telemedicine Utilization at Shriners Hospitals for Children, Salt Lake City

Jonathan Nicholas Grau

Supervisors: Alice Breton, MSc Director of Studies Royal College of Surgeons of Edinburgh Kristine Ferguson, MSW Director of Outpatient/Care Coordination Services Shriners Hospitals, Salt Lake City

December 2008

This project proposal is submitted in accordance with the requirements for the degree of Masters of Healthcare Informatics of the Royal College of Surgeons of Edinburgh and the University of Bath.

2 Copyright Notice Attention is drawn to the fact the copyright of this project rests with its author. This copy of the project has been supplied on condition that anyone who consults it is understood to recognize that its copyright rests with the author and that no quotation from the project and no information from it may be published without the prior written consent of the author.

Restrictions on Use This project may be made available for consultation within the university Library and may be photocopied or lent to other libraries for the purpose of consultation. Signature: Date: December 31, 2008

Disclaimer The opinions expressed in this work are entirely those of the author except where indicated in the text.

3 Acknowledgements Thanks to the faculty and administrators of this program at University of Bath and Royal College of Surgeons of Edinburgh. Thanks, in particular, goes to Alice Breton and Kris Ferguson and to my fellow students.

4

Abstract The author conducted a qualitative study to explore the factors that have facilitated and prevented optimal use of an established telemedicine program at Shriners Hospitals for Children, Salt Lake City. A focus group was carried out with attending staff to capture data regarding individual and shared experience working with the technology. A simplified extension of Unified Theory of Acceptance and Uses of Technology was applied to the data to identify factors that contributed to staff decision-making regarding how telemedicine was employed. The study revealed how pediatric orthopedic physicians were influenced by individual, technical, and managerial factors and how their individual perception of telemedicine had formed. The results from this study suggest that individual factors such as self-efficacy and planned behavior drove physicians to use telemedicine despite significant technical and managerial barriers. Sensitivity to professional status led them to alter their use of the technology, however, which led to a precipitous decline in its effectiveness. Themes that emerged from this study indicate that physicians require a high level of confidence both in the technology and in the medical skills of remote-side clinicians in order to deliver quality care and that patient satisfaction and expected cost benefits represent powerful normative factors for driving utilization.

5

Contents Abstract.........................................................................................................4 Contents........................................................................................................5 Introduction..................................................................................................6 Statement of Research Problem...............................................................10 Background...............................................................................................10 Research Problem....................................................................................11 Aims and Objectives.................................................................................13 Literature Review.......................................................................................15 Search Terms............................................................................................15 Research Tools.........................................................................................16 Reported Research..................................................................................17 Applied Theory..........................................................................................20 Cost Benefits Analysis..............................................................................23 Research.....................................................................................................25 Consideration of Research Methods........................................................25 Methods....................................................................................................27 Ethical Considerations..............................................................................30 Results......................................................................................................31 Fig. 1 – Focus Group Chart......................................................................34 Evolution of Opinion: Discord and Harmony...........................................34 Stages.......................................................................................................37 Emergent Themes....................................................................................39 Fig. 3 – Themes Chart..............................................................................40 Confidence in Technology and Continuity of Care ..................................40 Patient Satisfaction...................................................................................43 Evident Cost Benefits...............................................................................46 Discussion..................................................................................................48 SLC Staff Physician Perception................................................................50 Individual Factors......................................................................................52 Managerial Factors...................................................................................53 Technical Factors......................................................................................55 Conclusions................................................................................................57 Study Limitations and Areas for Further Study.........................................59 Recommendations....................................................................................60

6

Introduction

Healthcare Informatics (HI) is a broad and diverse field of study. At its most broad, it is the art of communicating health information across large populations. At its most diverse, it includes the science of minimally-invasive, unmanned robotic surgery and the promise of improved outcomes through decision support, data capture, and collaborative treatment technologies. The challenge for the HI researcher is to derive form, function, and value for a specific practice area through critical analysis, comparison, and first-hand experience of a given information system. I chose a study of telemedicine because I felt the technology had received a lot of attention in recent years yet had failed to be fully utilized as an efficient and effective care delivery system. Indeed, as recently as 30 June 2008, Adena Medical Center in Ohio, USA announced an expansion of its telemedicine services fueled by “an increasing need for medical specialists, the decreasing cost of technology and the prospect of lowering health care costs (1).” As I researched this topic I found that many telemedicine programs had likewise been founded on such lofty expectations yet rarely had these expectations actually been met (2). Like HI, telemedicine too has an expansive scope and has evolved rapidly through the years. Many organizations refer to telemedicine as the clinical or care delivery aspect of telehealth. The United States Department of Health and Human Services, Health Resources and Services Administration uses the term telehealth to describe all aspects of:

7 … the use of electronic information and telecommunications technologies to support long-distance clinical health care, patient and professional health-related education, public health and health administration (3). The Association of Telehealth Service Providers, on the other hand, defines telemedicine specifically and more narrowly as, “the provisioning of health care and education over a distance, using telecommunications technology (4).” Common to both definitions is the concept of distance. Telemedicine applications, such as videoconferencing, streaming media, and short message services, have been and will continue to be employed in an effort to conquer distance by giving physicians a longer virtual reach. Enabling providers to span time zones and cross over borders has introduced jurisdictional issues as well. The Joint Commission for the Accreditation of Healthcare Organizations (JCAHO) has established standards to help sort out the credentialing and privileging of telemedicine services in the United States and has set parameters for lawful conduct for licensed independent practitioners. JCAHO holds that telemedicine does not include interpretive or consultative services, for example. Consequently, JHACO standards narrow the definition of telemedicine by limiting it for use by practitioners with total or shared responsibility for the patient. This definition happens to severely constrain the real-world use of telemedicine by radiologists and triage specialists (5). Regardless of the various definitions used to describe or proscribe telemedicine, as a healthcare delivery mechanism it has changed little since the advent of phone communications and the “take two aspirin and

8 call me in the morning” mode of doctoring. Indeed, physicians and nurses have been providing ad-hoc care over the phones for the good part of the last century. It was only in the last twenty years, during the building out of the digital communications infrastructure that many new technologies with healthcare applications came online. With expanded bandwidth came additional capacity for digital visual imaging, two-way video, conferencing, and the like to add to a single voice stream. New telephony technology, for example, had the potential to transform and drive down costs related to emergency medicine, health management, and private practice and still do. Such technology can also help expand the reach of health delivery itself by making specialized care accessible to needy recipients and making localized clinical expertise a common weal asset(6,7). Distance and clinical expertise are therefore the two problems telemedicine is designed to address and Salt Lake City, Utah USA is an excellent example of a test bed for telemedicine services. It has a population of approximately one million yet lies 350 miles in any direction from the next nearest urban center. Salt Lake’s geographic isolation belies deep integration and connectivity into the global information system thanks to its high-tech industries. Shriners Hospitals for Children, Salt Lake City (SLC) benefits from an equally auspicious situation. It delivers free orthopaedic care to all qualified children along the western range of the Rocky Mountains of North America, from Mexico to Canada. It has operated a telemedicine program for nearly ten years to help support this mission. The six surgeons at SLC represent primary attending staff with over 100 years of shared experience at the

9 hospital. Each has at least six years of interaction with the telemedicine suite installed at the facility. The conclusions I drew from this study indicate that successful implementation of any telemedicine system comes with significant investment. Physicians in this study were overwhelmingly concerned with their ability to deliver quality care, to diagnose correctly and to treat effectively, across a virtual network. Barriers to adoption can be overcome by good program design, ongoing outcomes-based feedback, and constant refinement of and investment in/upgrades to the technology. Findings from this study have been drawn using rigorous methodological and theoretical practice and can be used to inform organizational gap analysis prior to design and investment in telemedicine technology.

10

Statement of Research Problem

Background

Shriners Hospitals for Children, Salt Lake City (SLC) is a regional not-for-profit 40-bed hospital specializing in pediatric orthopaedics. Its operating budget in 2005 was just under $22M. SLC treats children from birth to eighteen years of age for a variety of orthopaedic diseases, including but not limited to: scoliosis, clubfoot, and problems caused by disorders like cerebral palsy and spina bifida. Care is provided at no cost to patients and their families. The cost burden is shouldered by the Shriners Hospitals system and by local Shrine temples. In the words of the organization itself: “the Shriners and Shriners Hospitals are separate but inseparable (8).” All of the cost associated with inpatient care, from admitting through global period, is typically incurred by the facility which is funded by the Shriners Hospitals system. The costs associated with transportation for the patient and her family are incurred by the local Shriners Temple organization funded by members. It is this cost structure that defines, for better or worse, the value of the telemedicine program for the organization as each derives separate and unequal returns on investment.a SLC employs a number of ground-breaking technologies to help drive treatment plans. These include plastic and reconstructive surgery, motion analysis, innovative prosthetics and orthotics, and, as of 1998, a fully outfitted a

Families are generally responsible for their own food and lodging save for hardship cases, where the cost for this is assumed by the local Shrine Temple.

11 telemedicine suite. SLC serves patients within an area that includes six states in the U.S. and the Mexican states of Sonora and Chihuahua, covering approximately 800,000 square miles of North America. Its telemedicine suite has been in operation since 1998 and is one of the most widely used within the Shriners Hospitals system. The e-MedSoft telemedicine solution installed by SLC enables “physicians to deliver remote examination, diagnosis and treatment to patients via secure, private interactions … as well as allowing for the maintenance of a [Cerner Millenium] electronic patient record, whereby each physician can access their patient's file, complete with physician notes, lab results and X-rays (9).”

Research Problem

SLC Telemedicine clinics have been used most commonly to perform wound checks, post-op and yearly follow-up, and to assess the use and fit of prosthetics and braces. Each physician is expected to conduct a single clinic of at least 3 hours in the telemedicine suite per month and to evaluate at least 15 children during each clinic (10). Patient surveys conducted by the office of the telemedicine coordinator suggest that families enjoy the time savings and convenience that local telemedicine clinics afford. Growing demand for more telemedicine clinics, however, has been met by an ambivalent if not reluctant SLC physician corps. SLC physicians claim they have struggled with the technology and have employed it minimally.

12 SLC administrators, in turn, believe use of telemedicine technology suboptimal and have struggled to drive utilization beyond this benchmark of one clinic per doctor per month. SLC administrators have acknowledged that through the years SLC physicians blamed telemedicine technology for negatively impacting their clinical practice which has led to underutilization. Orthopaedic surgeons, perhaps more than other specialists, are governed by their sense of touch and their ability to observe range of motion, gate, and overall movement. The disjunction of video interface and audio delay, staff claims, has delivered immediate and overwhelming diminishment of good diagnostic practice not to mention interpersonal communication. Furthermore, Incompatibility between Picture Archiving and Communications Systems (PACS) often compounded the already poor image quality of x-rays. In order for SLC physicians to work with absolute confidence, they would often schedule subsequent in-person visits to work around the telemedicine clinic. SLC administration also believes itself to have a distinct disadvantage in fomenting adoption and utilization among its staff thanks to its culture. Staff physicians are measured on a more traditional clinical/outcomes-based scale not solely on their work quotient and ability to contain cost. Additionally, funding for day-to-day hospital operations comes from the Shriners Hospitals system, including all overhead for staff, technology, research, inpatient room/board, etc., while local Shrine temples finance the transportation costs. This shared cost structure is a product of the Shrine’s overall organizational structure: one part philanthropy, one part fraternity. This “split shareholder” arrangement creates indirect operating accountability with unclear repercussions should technology investment lose traction or fail.

13 Consequently, staff physicians appear to enjoy a relatively powerful position in the facility compared to their counterparts in fee-for-service institutions which may be a cause for poor program adherence.

Aims and Objectives

There is no question that SLC has a unique operating dynamic thanks to its not-for-profit status and philanthropic mission. Given the particulars of the culture and structure of SLC, this study ran the risk of being too narrow and lacking in resonant, applicable analysis. However, as Dr. Carolyn Clancy, Director of the Agency for Healthcare Research and Quality, declared, in her May 2005 presentation to the United States Congress Subcommittee on Health Committee on Veterans Affairs: Widespread adoption of individual telemedicine applications in the private sector will continue to grow slowly … unless creative ways are found to speed the development of solid, scientifically generalizable findings of their effectiveness. She went on to say that: While there are many promising initiatives underway, there are few mature telemedicine programs and few good scientific evaluations. There are, nonetheless, lessons learned that may prove useful to the VA. However, there is an obvious need to work collaboratively to identify best practices (11).

14 Following Dr. Clancy’s admonishment, my aim was to examine a mature telemedicine program to better understand drivers of telemedicine utilization and to identify real or perceived barriers that existed for physicians who interfaced with the system. In order to achieve this aim, I set out to learn specifically: -

why SLC physicians were initially impelled to utilize telemedicine;

-

how they felt telemedicine impacted their practice;

-

what the motivations were for continued utilization;

-

and, finally, what they would suggest to reinforce its wider use.

15

Literature Review The strategy employed for literature review was a search for research that has been conducted worldwide relating to usability and adoption of telemedicine technology, telemedicine’s impact on professional autonomy and perceived efficacy, and provider perception of outcomes related to its use. From this broad systematic search, I attempted to isolate research done within the last eight to ten years that involved provider acceptance of telemedicine within an orthopedic and/or pediatric setting. I adhered to this strategy as much as possible yet I followed a number of circular research paths proffered by GoogleScholar and PubMed functionality that supported links ranked by citation statistics, key authors, and other helpful criteria. Overall, this review of literature provided what I believe to be an exhaustive survey of subject matter and theory from recent years to help frame this study within parameters defined by relevance and currency.

Search Terms The search strategy involved matching combinations of the following topics and terms in an attempt to fine tune results:

Topics telemedicine

Terms acceptance, adoption, diffusion, efficacy, orthopedics, pediatrics, provider, satisfaction,

theory

“shriners”, utilization acceptance, adoption, diffusion, efficacy,

cost efficiency

qualitative, technology, telemedicine, utilization acceptance, adoption, efficacy, pediatrics, provider, technology, telemedicine, utilization

16

I performed a standard search followed by an exact-term search on each topic individually and each topic/term combination. Within telemedicine, “telemedicine adoption” returned the broadest set of results, followed by “telemedicine efficacy” and “telemedicine satisfaction”. To generate more precise results related to provider utilization in pediatrics, I targeted my search based on specialty, using terms such as “pediatric telemedicine” and “pediatric orthopedic telemedicine”. Finally, I searched for “shriners telemedicine adoption” which returned a single reported study. Within the theory topic, “diffusion theory” and “qualitative theory” had the broadest set of results while “telemedicine technology acceptance theory” returned three of the reported studies. The cost efficiency topic required search on combinations such as “telemedicine cost efficiency” and “telemedicine provider acceptance cost efficiency” in order to deliver relevant returns.

Research Tools I employed the search capabilities of Google, PubMed, and IEEE Search to survey the universe of literature. I also targeted specific peer-reviewed medical society and professional association Web sites for anything that may not have turned up in the systematic review. I found that GoogleScholar results consistently returned papers and articles of consequence and that a more targeted review and cross-reference using PubMed and IEEE had excellent results. Accordingly, when an appropriate publication, article, or dissertation was found via GoogleScholar, I would earmark the author for targeted searches using PubMed and IEEE. I followed citation and key author

17 leads as often as I could to ensure that I had not missed a source for relevant information. Finally, I went outside of the search engine framework and visited professional associations and medical societies that hosted their own symposia and member meetings. The American Medical Informatics Association (AMIA), through its Classic Paper Series, published a “Collection of Recommended Papers and Information Sources” (12) sourcing a very broad set of papers related to technology adoption. Additionally, the American Telemedicine Association (ATA) served as a virtual hub for vast amounts of research, including unpublished papers presented at ATA’s national and regional meetings.

Reported Research Journal articles provided the strongest material for this study. The most promising research focused on the dynamics of health information systems adoption and the roles, responsibilities, and level of influence held by various stakeholders. Lapointe and Rivard, for example, in their Getting Physicians to Accept New Information Technology (13) posited that individual, professional and organizational factors directly influence adoption and utilization of hospital information systems. They confirmed in their study that physicians who resist using technology held the most power within an institution and that these resistant physicians were responsible both for taking on the learning curve and for validating the technology’s efficacy before quitting it. They also determined that if the technology were implemented correctly, with broad resources and support, any continued problems associated with its utilization

18 represented larger organizational and structural issues; “if the implementers are the object of resistance, the computer system becomes extraneous to the problem.”

The most recent work I was able to find came out of ATA’s 2007 and 2008 symposia. While not directly related to telemedicine, the subject matter and applied theory were quite complementary. Whitten et al.’s Provider Acceptance of a Remote Trauma Consultation Network (14), Reynolds et al.’s The Role of Provider Perceptions in Deploying an Electronic Medical Record at Michigan State University (15), and Sanford Melzer et al.’s Experience with Pediatric Telemedicine Services in a Regional Network (16), for example, suggested that physicians must be convinced of the full spectrum of expected benefits and outcomes early in technology adoption and be regularly briefed on ongoing findings related to such measures in order for perceptions to be changed and/or reinforced. Reynolds, in particular, found that physicians’ perception of the impact of the technology on their practice had a greater effect on its utilization than did demographics such as age and sex. The Reynolds study was the only one of the three that appeared to use a rigorous quantified approach. This method delivered a broad set of data related to satisfaction and success rates for EMR adoption but left a number of important questions unanswered relative to telemedicine, such as: effects on quality, effects on physician and patient relationships, and finally effects on clinician autonomy.

19 Regarding telemedicine adoption, specifically, the most widely cited study was published by Gagnon, Lamothe, et al. who performed a quantitative study on physicians working within the provincial telemedicine network of Quebec. They found that telemedicine, like other clinical decision support and computer physician order entry technologies, was a technology that highly impacted physician perception of professional identity. They found that “physicians who perceived professional and social responsibilities regarding adoption of telehealth in their clinical practice had stronger intention to use the technology.” Their most profound finding was that physicians' intention to use telemedicine was better predicted if their self-perception as telemedicine users was considered (17).”

Within orthopaedics, evaluation of provider interfacing and systems utilization is limited to studies related to clinical decision-making and the use of imaging technology to support remote diagnosis and treatment (18,19,20). A single study regarding the telemedicine experience at Shriners Hospitals for Children, Honolulu was identified. Though similar in nature, the Ono and Lindsey (21) study at Honolulu, published in 2004, focused primarily on implementation and adoption within the Hawaii facility with recommendations on operations and ongoing maintenance for that specific program. Provider interfacing with and utilization of the system were not within scope. The Ono and Lindsey study and the findings of the other reported researchers represent a solid but small body of work in the analysis of physician perceptions of technology with an emphasis on telemedicine.

20

Applied Theory Prevailing social science in the area of systems adoption is dominated by Diffusion Theory, Technology Acceptance Model theory, and Social Cognitive Theory, championed by such researchers as Everett Rogers, F.D. Davis, and Albert Bandura, respectively. In 2003, a so-called Unified Theory of Acceptance and Uses of Technology (UTAUT) was proposed as a synthesis of all previous models of technology acceptance. While UTAUT and all of these theories separately can be used to analyze technology adoption by physicians, Mezni and Zeribi’s very recent Determinants of the Individual Acceptance of the Telemedicine (22) which includes the proposal to merge UTAUT’s (at least) eight different adoption factors into a simpler model seems to be most promising. Mezni and Zeribi’s Individual Acceptance model is especially applicable to this study as it draws from the strengths of studies using Self Efficacy and Planned Behavior to specifically analyze telemedicine use among highly autonomous practitioners. Diffusion Theory, regarded as a benchmark for analyzing technology adoption, works best at a macro-level thanks to its application-neutral view of the technology and of its adopters. Though Rogers’s innovation-decision process focuses on an individual’s acknowledgement of a new technology as a potentially productive or useful tool, the process through which “an individual or other decision-making unit” passes first from knowledge of an innovation to forming an attitude toward the innovation, to making a decision to adopt or reject, to implementation of the new idea, and finally to confirmation of this decision does not account for technologies that disrupt or alter significantly

21 the status-quo, specifically in a healthcare setting. Diffusion Theory therefore serves as a suitable benchmark but could not be used to analyze in whole or in part any of the data gathered in this study(23).

Technology Acceptance Model (TAM) theory has been applied to technology in a healthcare setting but not specifically to telemedicine. Succi et al. in their 1999 Theory of User Acceptance of Information Technologies, discussed how TAM could be used to analyze physician utilization of disruptive healthcare technologies such as decision-support and electronic health records. They found that the addition of specific determinants, such as “the perceived impact of using the technology on Professional Status,” was a key component to successful technology adoption by physicians (24). Most recently, Jen-Her Wu et al. in their Testing the Technology Acceptance Model for Evaluating Healthcare Professionals' Intention to Use an Adverse Event Reporting System (25) applied TAM extensions Subjective Norm and Perceived Ease-ofUse to the study of physician adoption of a medical reporting system. Interestingly, both sets of researchers suggested that TAM was limited and required extensions to address the more highly individuated perspective of end-users in the medical profession. There does not appear to have been any subsequent research done using TAM or any of its Subjective Norm, Perceived Ease-of-Use, and Professional Status extensions on telemedicine.

Definitive work using Social Cognitive Theory (SCT) was conducted by a group of researchers again led by Marie-Pierre Gagnon in Quebec, Canada in 2003. According to Gagnon, et al., “physicians who perceived professional

22 and social responsibilities regarding adoption of telehealth in their clinical practice had stronger intention to use [the] technology (26).” Their study employed a theoretical framework established by Hu and Chau who used SCT extensions, Planned Behavior and Self Efficacy, to evaluate the dominant psychosocial drivers of telemedicine adoption. Hu and Chau found that “technology that could interfere with physicians' traditional practice could affect their perception of their professional role (27).” The importance of a physician’s “perception of their professional role” within an equivalent network of “social responsibilities” begins to describe the challenges associated with telemedicine studies using TAM. Gagnon’s application of SCT to telemedicine therefore established a foundation for future research designed to explore the universe of behavioral factors.

As mentioned, Unified Theory of Acceptance and Uses of Technology (UTAUT), while exceptionally robust due to its inclusion of such TAM and SCT extensions as performance expectancy, effort expectancy, social influence and facilitating conditions, had yet to be applied to telemedicine until earlier this year. It was not done without significant alteration (28). Mezni and Zeribi’s very recently introduced Individual Acceptance model, a heavily modified UTAUT model, was used to analyze telemedicine use in Tunisia.b Areas Mezni and Zeribi identified as essential to telemedicine analysis, specifically, the components that comprise a physician’s perception of her capability vis a vis telemedicine represent three areas of “unified theory”: A) Individual Factors, B) Managerial Factors, and C) Technical Factors. The following b

A review of their proposal reveals a quite loose affiliation with UTAUT insofar as it uses the concept of “unified theory” to rationalize its synthesis of SCT drivers that were already proven by Hu and Gagnon to be most directly applicable to telemedicine studies.

23 schema illustrates how those factors combine to contribute to a Physician’s Perception:

Individual Factors Self-efficacy Planned Behavior

Managerial Factors Executive-level Champion Ongoing Training

Physician’s Perception

Individual Acceptance of Telemedicine

Technical Factors Quality Measurements Level of Disruption

Individual Factors are composed of SCT factors Self Efficacy and Planned Behavior which attempt to define how the physician feels technology can enable him to achieve his ends effectively and efficiently. Managerial Factors suggest that an executive level champion must enable the use of the technology and provide training for staff to follow. Finally, Technical Factors include establishing quality measures that can be used to determine the level of compatibility with existing processes. All of these factors combine to influence a Physician’s Perception of a technology’s Ease-of-Use and Expected Performance. The author believes Individual Acceptance model correlates with the situation at SLC and can be used to better understand the captured data.

Cost Benefits Analysis

24 As of December 2008, The American Telemedicine Association (29) documented approximately 200 networks operating in the United States, linking over 2,500 institutions nationwide. Often overlooked in systematic reviews of telemedicine adoption research are the business drivers that have justified build-out of such an expansive telemedicine infrastructure. As mentioned at the outset, fee-for-service providers in the United States generally arrive at decisions to invest and adopt such technology based purely on competitive advantage and cost benefit. The same business drivers can and do exist for non-profits. Indeed budgetary pressures at non-profits can represent significant risk to their mission-driven operations. Understanding how telemedicine has been viewed as a business investment, therefore, and learning whether returns on investment have actually been realized are important factors in evaluating the adoption cycle.

Douglas Smith’s 2005 article, The Influence of Financial Factors on the Deployment of Telemedicine, provided a thorough breakdown of the five financial factors that are most often considered when a telemedicine program is being evaluated. They include: 1. Initial or capital investment; 2. Operating or ongoing costs; 3. Profitability or net income; 4. Cash flow; and 5. Reimbursement. Smith noted that the business case for telemedicine is often very strong at the outset but, “sustainability … is an issue of concern. Although governmental or other agencies may provide some funding for initial deployment, the ongoing operating funds are more uncertain (30).” Broens et al. leveraged Smith’s study to propose “a layered implementation model in which the primary focus on individual determinants changes throughout the

25 development life cycle of the telemedicine implementation (31).” This business driven adoption model would serve to move pilot studies rapidly into robust deployment and continued success.

Much of the research in this area indicts the healthcare economy of the United States which effectively requires the Federal Government to underwrite new technology by approving reimbursement for such services in its Medicare Physician Fee Schedule. At the time of this writing, the United States Congress had just passed into law the first measure for payment of telemedicine services in skilled nursing facilities, in-hospital dialysis centers and community mental health centers (32). This legislation comes after nearly two decades of wide telemedicine utilization in the industry. This legislation could set the stage for more consistent approval and payment for services rendered via telephony by commercial health plans to make telemedicine a viable long-term delivery solution.

Research

Consideration of Research Methods SLC administrators communicated to me that attending physicians held well differentiated viewpoints and were willing to speak freely on the topic of telemedicine. I therefore took advantage of a standing weekly staff meeting to conduct a focus group, ensuring that I met with the attending staff all at once and in a familiar setting. The choice of an informal session was based on a review of prevailing theory in qualitative data capture with specific attention to

26 Jim Brigsby, Pamela Whitten, and others who found that a “communicative focus which privileges the role of participatory conversation [should be] used to examine and explain the invention, diffusion, and reinvention of telemedicine (33).”

The staff at SLC consisted of six attending physicians each of whom was obliged to conduct telemedicine outreach clinics. Nearly 1200 hours of telemedicine practice had been undertaken among them over the previous eight years. As expert sampling involves the assembly of persons with known or demonstrable experience and expertise in a particular area, so the choice of this sample was made by default and not by convenience.

I conducted a 60-minute focus group to allow for plenty of disclosure, conversation, and potential argument. Every effort was made to capture the tone of voice, body language, and degree of emotional engagement. Simultaneous audio and video devices recorded the conversation following Catterall’s (34) theory that how people talk about a topic and how they respond in a situation where they are exposed to the views and experiences of others is as important as what they say in direct respond to a series of questions. Indeed, the sequence and evolution of the participant responses did seem to enrich the data.

As well as a form of data collection, the focus group was intended to be used to identify relevant content for a more formal quantitative approach (35). My initial proposal involved scaling up this pilot study into an online survey of all

27 doctors in the Shriners Hospitals system. I met with three of the six executives at Shriners Hospitals, each an accomplished researcher and informaticist, who found this proposal promising and generally aligned with the system’s overall vision for telemedicine. Within weeks, however, external pressures related to the management of the non-profit system had forced leadership to effectively close any external research into Shriners Hospitals. This circumstance, understandable and unfortunate, caused me to reassess the pilot study at SLC and to reevaluate its methodology and potential outcomes. Taking into consideration the business drivers and logistical situation, research design for this study was amended to focus exclusively on the qualitative phase. In retrospect, I feel that had I extended the study to a population that had too few telemedicine programs up and running I would have run the risk of capturing far too much unqualified information (36). As late as May 2008, only twelve of the system’s twenty-two hospitals had active programs and of the twelve only Honolulu and Salt Lake City had programs of relative stability and maturity. Based on this study’s findings, the fact that this study remained qualitative does not indicate that it was in any way limited. Recommendations include the possibility for scaling up to include the twelve Shriners hospitals currently employing telemedicine (a modification of this study’s original plan) and also the opportunity to do further Individual Acceptance research on targeted practitioners within the Shriners system based entirely on the results of this study.

Methods

28 Self-contained focus groups have been shown to be a good method for discovering factors or influences that might not be previously known to the interviewer or to the participants themselves. Joss and Durant (37), for example, documented how focus groups have been used to help crystallize individual opinion, heighten group awareness, and strengthen perceptions of shared experience within the confines of a presumably static research objective. As Kitzinger (38) observed, focus groups can give participants the freedom to form and express opinion through interaction and can also help the interviewer better understand how the observed interaction can be used to analyze the results. The focus group also allows for participants to experience and contemplate the group and their role in the group which can precipitate further reflection on their own experience, potential interaction outside of established hierarchy or protocol, and even exploration of an argumentative approach or play “devil’s advocate”. The focus group is by no means a controlled environment, therefore, and is ultimately unpredictable. It has been the source of concern for qualitative researchers such as Fern and Bristol (39) who argued that the group dynamic can suppress or even polarize individual opinion, can undermine the building up of consensus, and can result in unclear outcomes. Conversely, many studies have shown how focus group interaction yields stronger, richer data thanks to unforeseen and spontaneous “synergism” (40). I chose a focus group because it seemed to be the best way to capture substantive data about each physician, including his or her experience with the technology and, especially, the decision-making process behind its use.

29 A common concern going into the focus group data capture is how the moderator can affect the proceedings. “The moderator must be sufficiently involved in the group to fulfill the role of facilitator, but not so dominant as to bias or inhibit discussion (41).” Accordingly, an interview script was designed to support the focus group by encouraging the experts to speak freely about their experiences on specified topics. Open-ended questions were crafted along with somewhat more detailed follow-up questions to help focus on areas of specific interest. The script was tested with the advisors for this study and reduced to three sets of questions.c

Telemedicine Ease-of-Use 1) What would you rank your knowledge of the telemedicine system? 2) How frequently are you brought up to date on changes to system? 3) Who keeps you informed of and debriefs you on changes to the system? 4) How much emphasis is placed on the system? 5) How would you compare your experience with the telemedicine system with other types of electronic systems currently in place such as medical records, etc.?

Clinical Value 6) Has the system been shown to improve patient outcomes? 7) Has the system been shown to reduce your work quotient?

c

Interestingly, these three sets of questions were finalized prior to a review of UTAUT and the introduction of Individual Acceptance model to the literature.

30 8) Is the system always being brought to your attention in staff meetings or committee forums?

Institutional Value 9) Did you receive formal training on how to use the system? 10) Are you expected to train yourself on use of the system? 11) Does staff stay fully updated on system capabilities? 12) Is ongoing analysis (utilization, outcomes, improvements) communicated to staff? 13) Is the system easy to use? 14) Do you feel you are expected to use the system?

While every attempt was made to execute the script, the conversation inevitably followed its own course. Extrapolation, interjection, and reversingof-course overwhelmed the scripted questions; only a few were actually posed. The participants were excited and willing to speak freely. The conversation itself generally held to the intent of the script and I tried to lead the conversation according to the agenda set by the script.

Ethical Considerations This study involved consenting human subjects and took place in the United States of America. Full consideration was required, therefore, by University of Bath, School for Health, School Research Ethics Approval Panel (SREAP) before work could begin. Issues this study addressed in order to gain full

31 SREAP approval included detailing how sample representatives would be recruited and assembled and what blinding measures would be adopted to ensure confidentiality and anonymity. Additionally, I obtained and processed first-person quotations through recorded interviews. Voice and video files have been destroyed in compliance with the UK Data Protection Act. Focus Group participants can gain access to and make requests for alteration of this study should they find it does not satisfactorily protect them from damage.

Results Taken as a whole, the focus group conversation followed a predictable trajectory; one that allowed the most outspoken members to reveal themselves and to share their well-formed opinions early and the more reserved to express their opinions in kind. Regardless of whether he or she volunteered information early in the conversation, it was clear that each respondent had a particular point of view that may or may not have been known by other members of the group before hand. This made for a rich discussion and provided enough data for later analysis.

The study named respondents by the order in which they made their first remarks. Accordingly, Respondent One (R1) and Respondent Two (R2) spoke first and second (and did so in an excited manner). Respondent Three (R3) had a sort of prepared statement and did not engage in further discussion. Respondent Four (R4) and Respondent Five (R5) were influential to the conversation’s narrative and made a number of quite poignant

32 statements. Respondent Six revealed himself to be an outlier. He was not an orthopaedic doctor but a burn doctor. His contribution was important to the flow of the conversation but his responses regarding utilization were not helpful. Respondent Six should not be viewed as a negative result of the methodology rather as a by-product of the focus group format.

To facilitate subsequent analysis of data, respondent contributions were recorded on audio and video media and memoed by the interviewer in realtime. Individual Acceptance theory’s three factor groupings provided the structure for the interpretive coding that took place during analysis. Its managerial, technical, and individual factor groupings neatly aligned with the survey questions and ultimately with the overarching structure of the conversation. The subtleties of the conversation were memoed and later recalled after replay of the interview.

For example, as initial viewpoints were communicated, I notated that subsequent commentary by an individual often grew more reflective and, in at least one case, softened over time. Such evolution of opinion was an expected outcome of the grounded semi-structured approach. The group setting and Time, itself, seemed to function in a normative way. By charting responses according to factor groupings along a Time axis, I was able to visualize the affect the group dynamic had on individual opinion, the conversation’s stages, and ultimately its themes. That is, by displaying each statement according to its factor grouping over the course of the interview, I feel I was able to achieve a greater understanding of the narrative structure

33 (42)

. The following Focus Group Chart (fig.1) illustrates how literal statements

evolved over Time (where the center represents the end of the conversation and the outer-edge the very beginning). Placement indicates when the respondent spoke, what he or she spoke about, and what factor grouping they touched on. If a respondent chose to touch on all three or a synthesis of two, I plotted that point closer to another grouping. Respondent One, for example, spoke about all three factor grouping in his first multi-part statement which are highlighted below:

34

Fig. 1 – Focus Group Chart

Evolution of Opinion: Discord and Harmony According to Sim, Fern and Bristol (39,40), focus group study analysis should include a study of any normative or suppressive factors that may have

35 affected the data. By analyzing each respondent’s viewpoints over the course of the focus group we can see how group interaction may have helped fix or alter the original position and overall sentiment. As mentioned, the first three respondents commented that they were unenthusiastic about the technology and cited a number of specific reasons why it could not be used effectively. They did so in a decidedly pessimistic fashion. Thanks to the group dynamic, however, R1 and R2 seemed to come around to a more broadly accepted opinion. While they still held somewhat antagonistic views their polarized views seemed somewhat neutralized. As previously mentioned, R3 made a prepared statement, engaged in short dialogue, then left. R4 and R5 both started and ended the conversation with positive patient and outcomes-oriented language that did not change over the course of the conversation. If any normalization of opinion occurred during the focus group it served to neutralize the somewhat polarized views of R1 and R2. We can see the evolution of both respondent viewpoints over the course of the conversation. Each of R1’s initial statements, for example, represented a negative viewpoint that was reinforced by R2. Each employed words and phrases like “problem”, “total waste of time,” “ungraceful,” “compromised” to describe their experience with the technology and its administration. After nearly sixty minutes of vigorous conversation, there seemed to be softening of previous held view. Phrases and words like: “It’s better medically,” “more effective,” “incredibly good” were used to summarize their viewpoints near the end in accordance with the overall evolution of the conversation.

36 R1 used a great deal of first-person experience to validate his negative opinion. It was as if R1 wanted to make sure his displeasure with clinics administration was documented in order to justify his unease with or distrust of the technology. After consideration of others, however, and a strong shift to discussion of patient expectation and perceived outcomes, his initial view softened.

Such is also the case with Respondent Two, who after an extremely pessimistic and negative representation of his experience as, “a total waste of time for us to sit there for four or five hours and see six or seven patients,” posited a more constructive approach that suggested, ”a better use of time and effort would be to have a local pediatrician and if he sees a problem to send them down.” R2 was clearly the most dissatisfied with the technology yet he likewise acknowledged the benefit to the patient and the opportunity for improvement.

Respondent Four, on the other hand, began with a clear acknowledgement that he felt the clinics were, “patient-driven and … valuable for the post-op for some simple things.” He continued to discuss the benefits of the technology and concluded by saying, “it’s better than it used to be.” R4 ultimately held to his original view which seemed to affect R1 and R2.

Finally, R5 began and ended with the most positive statements of the lot: “I’ve certainly had some great satisfaction out of a number of cases where you could actually get some things done and the parents were too busy or

37 whatever to come down and that’s been actually … good,” and, “I think that’s just the nature of what we do and how try to kind of triage what’s the most effective use of everybody’s time and telemedicine gets knocked down to a lower rung because we’ve got other things going on. It’s better than nothing in many instances.”

Stages Bales (43) identified that a group goes through a sequence of stages simply as a result of its existence as a group. Therefore, in order to more easily identify and sequence the stages that emerged out of this focus group, I used the respondent chart to pinpoint those contributions that seem to have impacted the tenor or tempo of the discussion. The resulting map (fig. 2) illustrates how the conversation progressed as whole. I was also able to identify which respondents had a transformative affect on the conversation. In effect, I sought to identify the substantive statements that served to evolve the narrative and the normative statements that represented the group’s shared opinion (44). Plotted on a Time axis, these key statements could be used to identify the transitions and connectors of each stage. I applied Tuckman’s (45) theory of group development to help identify the stages I had initially identified and named as expository, interjectory, climactic, and resolving. His concepts forming – storming – norming - performing aligned neatly with the stages I had observed as a period of orientation/testing (expository/forming), then of conflict (interjectory/storming), then of group cohesion (climactic/norming) and finally settling on functional role-relatedness (resolving/performing). The

38 organic emergence of these stages convinced me that the data capture had been effective, that the focus group method was valid and that the data being captured could be analyzed on a number of different levels.

Fig. 2 – Stages Chart

39

Emergent Themes As Catterall and McLaran (46) posited, there is no proven method of analyzing focus group results. There are however many techniques that have been used to analyze the observed interaction of focus group participants, structurally as well as thematically. In addition to reviewing verbatim content for word repetition and indigenous verbiage, the researcher can look for potentially important themes by comparing and contrasting what is said during the course of the conversation, how it is said, and by whom. Further, sometimes important information is communicated often by what is left unsaid during the course of a conversation. As previously shown in figure 2, by plotting the focus group on a factor grouping and Time axis, I had created both a stages narrative and a preliminary correspondence chart that served to identify areas of dissonance and concordance within the focus group (47). A review of these areas of concordance, including the key-words-in-context and (shared) metaphor allowed me to identify at least three emergent themes that I believe were consequential to understanding provider acceptance of telemedicine at SLC. These three themes were identified in figure 3 as: 1. Confidence in Technology and Continuity of Care 2. Patient Satisfaction 3. Evident Cost Benefits

40

Fig. 3 – Themes Chart

Confidence in Technology and Continuity of Care The Confidence in Technology and Continuity of Care theme revealed itself through a series of initial ice-breaking statements by the first four respondents

41 and vigorous discussion throughout the focus group. As the first and strongest concordant theme, it revealed shared experience within technical and individual factor groupings, specifically level of disruption and selfefficacy. As previously mentioned, R1 and R2 worked aggressively to set the tone of the conversation from the outset. It was as if R1 and R2 colluded to bias the conversation at the outset with their negative view on every aspect of the technology, administration of the technology, and its affect on their livelihoods. Comments like: “it’s second class,” “it’s a total waste of time,” “it is sort of not ideal,” and “it is compromised,” set the tone for the theme. Frustration with the technology itself was communicated with clear language and emotion in such statements as:

Respondent One … in these instances you are talking to a treating physician with legal responsibility for that patient and you are offering suggestions rather than trying to treat your own patients in a second-class sort of form where you really don’t have the best information coming in because you either cant see the x-ray or you cant trust the exam.

Respondent Two … on the telemedicine side is the communication issue. The communication sitting in a room is completely different than the communication over a TV screen where you may or may not even be able to see the person you are talking to. And I have a very difficult time keeping a train of thought, keeping … trying to get information.

42

Respondent Four Actually, if someone wanted to transmit the x-rays digitally I don’t think we have any of our systems that do that reproducibly [sic].

Each respondent grappled with technical issues they felt had been working against them as clinicians, specifically incompatible imaging and communications systems. This technical discussion precipitated a more charged discussion surrounding the corresponding concept of Continuity of Care detailed by R3 in the concise statement that follows:

Respondent Three It’s completely dependent on the examiner on the other end. You know, if you have to have someone that you’re talking through how to do an exam it’s useless … So sometimes I feel an emotional disconnect from the patient from that and a really important part of your practice is being able to emotionally connect with the mom and the kid and, um, I can’t do that as well in telemedicine.

Indeed, each respondent commented first on how the technology failed as a tool to extend capabilities then talked about how specific personnel issues undermined their ability to provide quality care by proxy. R1 lamented the fact that, “we practice medicine face-to-face and when forced to practice it over the teleconference … you realize it’s not the same level of quality,” and each respondent agreed that the practitioner at the remote end of the telemedicine

43 must be experienced to ensure quality could be delivered regardless of technology.

Patient Satisfaction As each respondent tried to reconcile their negative experience associated with technical factors, the theme of Patient Satisfaction revealed itself. In particular, the physicians shared concern over how they were perceived as practitioners. Each respondent acknowledged that “it [telemedicine] is better than nothing for the patient.” This most concordant sentiment of the entire conversation was a general acceptance that the technology had the potential to improve outcomes thanks to patient acceptance if nothing else. R5 spoke up for the first time with a highly charged and positive statement which included terms like, “effective,” really like,” and “very, very good.” R1 and R2 seemed surprised by this viewpoint. R5’s influence impacted the Patient Satisfaction theme which remained positive and aspirational throughout. Indeed, many of the respondents felt patient satisfaction was a high priority which in some instances trumped apparent efficacy. On the other hand, the fact that these physicians were pediatric surgeons may reveal a group bias that weighed patient comfort, psychological and physical, over all other factors. Accordingly, the perception that the technology allowed them to be effective as long as the patient was satisfied was shared by the group:

44 Respondent Five As kind of a glorified triage mechanism for problems or calls that parents have, the parents really like it for the main reason that you can sit down and talk much more effectively … I’ve certainly had some great satisfaction out of a number of cases where you could actually get some things done and the parents were too busy or whatever to come down and that’s been actually … good.

Respondent One Its more user-friendly for families who are three hundred four hundred miles away that have questions to say ‘o.k. you go a mile down the road or maybe even fifteen miles or twenty miles to do that telemedicine thing’ and whoever is in clinic just looks at a screen just like they would in the next room you know ‘you’re doing okay’ or ‘maybe you should come in’ or ‘that looks like it’s a wound and needs to be seen.’

Within the Patient Satisfaction theme we see the aspirational views of the respondents who are better able to envision the larger possibilities of the technology on their practice regardless of how they felt individually about the technology. Given the perceived limitations of the technology and the frustration with its administration, the respondents did see value in its use and agreed on an ideal application for the technology thanks to the perception of the patient:

45

Respondent Four … they have tele-gait lab 2-D video which was actually pretty good, better than just watching the kid walk, but they have a special hallway with a video from the front and a video form the side where you can see, you know … and they can actually send the video then they had a couple of sessions where they correlate the 2-D video analysis with the 3-D gait lab. It was a pretty good correlation … You can see how they walk and if they look like they are walking great then … They can go ahead and do their gait lab and send it to you and in your free time.

Respondent Two … the telemedicine room would simply be one more room and you would go in there when the light comes on, meaning the patient is ready to be examined rather than sitting in front of this TV camera during all of this downtime … and you go to the next room that’s a live room here with a patient in it and you don’t come back until the next telemedicine patient is really ready. That makes our time more effective.

Such constructive statements indicated that the physicians were curious to learn more about potential outcomes with technical issues resolved. Patient Satisfaction as a theme also paralleled the Performing/Resolution stage of the focus group which confirms this as an important contextual area.

46

Evident Cost Benefits The final theme to emerge was one that involved discussion of SLC administration and apparent cost-effectiveness of telemedicine services. Respondents seemed to feel Administration had over-estimated the value of the technology yet had built a fragile support system around its use. They could understand the rationale for initial investment and remained convinced of its overall value proposition. R1 and R3 articulated the challenge associated with running a cost-efficient operation thusly:

Respondent One I think the problem we run into is the panacea that telemedicine seems to offer. A system like the Shrine generates a huge push to see more and more and from the medical side we’re saying ‘you know we’re not … it’s second class to begin with.

Yes, it does, perhaps, save money but the other shoe is that it’s not state of the art medical care. It is compromised. And you kind of have to make it as safe as you can within the constraints of the technology.

Respondent Three Although I completely appreciate the money it saves the Shrine and I know it’s a lot more convenient for a lot of patients and they’re very thankful for that. I always sort of feel like I hadn’t done my job as well as I could have.

47 Respondent One Telemedicine is encouraged, particularly, from the board level. They look at our numbers and they are discouraged when they go down because they perceive that it saves them money. Whether, in the big picture, it [does,] depends on which pot the money comes out of in order to determine whether it really saves or not because as [R2] pointed out there’s a lot of assets that go into these clinics some of which is visible to our Board of Trustees but nevertheless there’s cost involved … It’s better medically. We’re seeing patients that are more appropriate …

Respondent Four On the other side of that equation I think you know early on … you look at the figures, you look at the travel fund and the costs of paying for the travel and you have to offset that with the cost of running the operation.

While each respondent had frustrating experiences with technology and generally felt their effectiveness had been compromised as clinicians, they all continued to be convinced of the cost benefits that made such a compromised program “better than nothing in many instances.” That telemedicine as a concept could be viewed as a cost-effective tool that might not significantly undermine physician autonomy was an opinion shared by each respondent. That SLC had somehow failed in its delivery and/or support of the technology and had undermined physician autonomy was also a shared opinion. The group held a common understanding of the need for appropriate application of

48 the technology and agreed that while the technology may have been flawed and the administration of the program may have needed alteration, the potential for high-yield outcomes yet remained.

Discussion Given the long history of this group with its telemedicine program, it was not surprising that one of the themes to emerge from the discussion was one focused on pragmatism: cost benefits that justified continued telemedicine use. The emergence of the Evident Cost Benefits theme, therefore, indicates that the physicians were in agreement on the concept of telemedicine, its promise and potential. Despite this overarching sentiment, SLC staff purposefully chose to limit their use of the technology to accomplish the revised, sub-optimal objectives of: 1. meeting institutional mandates; 2. minimizing risk to themselves; and, 3. satisfying patient expectation.

What drove them to this behavior was a multi-faceted distrust of the technology. There is no question that technical and interfacing issues presented problems. Staff physicians knew additional resources could be applied to fix incompatibility and latent communications. Staff was more than a bit frustrated this did not happen sooner and had expected upgrades to occur much more regularly. Yet what was clearly a more ominous component to the technical problem was the lack of confidence in the clinic-side

49 doctoring. Each respondent had negative and possibly frightening experiences with telemedicine. R1 offered an example of his experience, shared by many in the group, which indicated how fearful he was of making an incorrect diagnosis due to the technology: Once a year, a child will come to us with in-toe which is a very simple, very common pediatric problem that gets better with time … Once a year you’re going to find a child with spasticity that the physical therapist examiner on the other end might not actually find. That child has a brain tumor. It’s not typical in-toeing; it’s actually neurological. That’s the fear.

In other institutions, this crisis in confidence in technology and in care continuity may have forced a program to come to a halt. There are no studies to tell us how incompatibility issues with technology and inconsistency in remote skills have affected telemedicine programs. We can only assume that SLC staff physicians continued to work with the technology because they had the authority to make adjustments they felt were necessary. As discussed, the Patient Satisfaction theme represented an acknowledgement that the physicians knew their patients valued the hospital’s use of telemedicine. Together, these complex themes illustrate why recent research has focus on uncovering the complexities of physician acceptance and why Mezni and Zeribi’s model does a good job of isolating the key factors that contribute to Physician Perception.

50

SLC Staff Physician Perception Telemedicine at SLC was perceived by physicians to carry both positive and negative attributes. Each physician acknowledged that they felt telemedicine somehow impacted how they were perceived as professionals. Each felt compelled to use a technology they felt had very low clinical value. Each believed that their patients and their families felt good about telemedicine clinics and that hospital administration, not to mention highly visible members of the local Shrine temples, expected it to be used as much as possible. These themes had direct, literal correspondence to the factor groupings. By charting whether the correspondence was positive or negative, we identified how the factor groupings impacted the situation at SLC. In some cases no data existed to support a correspondence which hinted at potentially larger organizational issues. Identifying these particular correspondences allowed me to produce a thematic overlay to the factor groupings and definitive explanation for the SLC experience. Expressed as positive, negative or noncontributing to the physician decision-making process and the creation of Physician Perception, the key factors are:

51 Thematic Overlay Confidence in Technology / Continuity of Care

Patient Satisfaction

Evident Cost Benefits

+ +

+ +

+ +

Individual Factors Self-efficacy Planned Behavior Managerial Factors Executive-level Champion Ongoing Training Technical Factors Quality Measurements Level of Disruption

-

Taken as a whole, the focus group provided three strong areas of thematic correspondence across the sample. Based on the progression of stages and the normalization of opinion over the course of the conversation, I determined there were too few data to suggest that any single theme was dominant. While individual respondents may have weighted certain factor groupings differently in developing his or her own Physician Perception, none were pronounced enough to resonate as themes across the sample.

52

Individual Factors Individual factors dominated the discussion. Despite problematic experiences with telemedicine, SLC physicians persevered with its use. This behavior indicates that they cared a great deal about how they were perceived as professionals. There also appeared to be a good deal of solidarity among the staff which manifested in competitive (peer) pressure, desire for control/mastery of technology, and concern about professional status within the group. For example, respondents spoke often about their intention to use the technology for its best purpose. Statements like, “I think that’s just the nature of what we do; to try to kind of determine what’s the most effective use of everybody’s time. Telemedicine gets knocked down to a lower rung because we’ve got other things going on,” indicated that each had a definite idea of what that “effective use” should be and over time each acted upon those instincts Staff could not deny the importance the technology seemed to hold for institutional stakeholders, and for each other, so they found ways to work around the clinics requirements. They viewed the telemedicine program as a potential liability and therefore made what they felt were necessary, riskadjusted changes to its application in order to meet hospital administration, Shriners, and patient expectations. Such behavior represented the power of the normative environment within which they operated. Individual factors, therefore, seemed to have had an overwhelming impact on Physician Perception of telemedicine at SLC despite evidence to suggest that the technology and the administration had failed them in many ways.

53

Managerial Factors The telemedicine program at SLC impacted a fairly autonomous and highly professional staff rather profoundly. By 2003, SLC administration had implemented and held staff physicians accountable for clinics requirements of one clinic per month. Whether this mandate was a product of the success hospital administrators believed they were having and a way to capitalize on the program’s momentum or whether it was a response to early resistance they were feeling from staff is unknown. The fact that the mandate was put in place indicates that the hospital was planning to further operationalize the technology. According to Mezni and Zeribi’s model, hospital administration is responsible for recruiting a clinical champion and for providing on-going training/education as they proceed with a telemedicine program. As Peter Yellowlees discussed in his Successfully Developing a Telemedicine System (48)

, the physician or “clinical” champion is the peer and advocate who takes

on the burden of the learning-curve, troubleshoots the technology, and fine tunes its application.

SLC had a clinical champion early in the adoption cycle and this factor may have contributed to a sense of competitiveness and of obligation among staff to employ telemedicine early and often, certainly before any clinics requirements had been communicated by administrators. With Chief Medical Officer, Peter Armstrong, SLC had a clinical champion who definitively associated himself with the success or failure of the program. However, within 24 months of initiating and driving the program, Dr. Armstrong was selected to become the Shiners Hospitals corporate Director of Medical

54 Affairs in 2000. His departure unquestionably affected how telemedicine would be devalued and potentially underused by the staff he left behind.

The decision to invest in a telemedicine suite at SLC was made by the Board in 1997 and funded by the Shriners Hospitals system and the local Shrine Temple in 1998. At a symposium in 1999, Dr. Armstrong made the following remarks: The Shriner telemedicine initiative set out to provide supplemental services by reducing the need for travel and lost time away from work and home for patients and their families, as well as for Intermountain [SLC] staff. Beyond supplementing outreach clinics, telemedicine is expected to enable the SHC-Intermountain to reach additional patient populations and provide patient and provider education. Ultimately these services will provide a medium by which Shrine hospitals can exchange information (e.g., outcomes research) (49).

This compelling language shows an activist, internal champion seeking novel ways to realize the institution’s mission. Like many in the industry at that time, he believed the material benefits of streaming voice and data to transform clinical practice were within reach. Yet within six years of the launch of SLC’s telemedicine initiative the story had reached its denouement. Despite a mandate by the administration and high patient satisfaction, utilization plateaued and more and more clinics were being cancelled. Did Armstong’s leadership as clinical champion really have such a profound affect

55 on utilization in the first three years? His name wasn’t mentioned once in the focus group.

Further, SLC administration rarely upgraded the system and seldom provided training. There was informal surveying of the patient population to support outcomes research but the findings were never released. For clinicians who live and die by outcomes research, the dearth of information in support of telemedicine utilization within SLC and in the Shriners Hospitals system atlarge amounts to conceptual gap between purported benefits and actual outcomes. Staff physicians had nothing to turn to within their organization or even their specialty if they wanted to learn more about the technology and its effects on their clinical practice. This Managerial factors gap represents a significant area of speculation and should be a focus for further study.

Technical Factors

Some time early in the adoption process, each staff physician came to their own conclusion about the potential risks associated with use of the telemedicine system as it was originally installed. Again, the shared belief that the technology was conceptually sound and afforded cost benefits to the institution indicates that staff was willing to use an improved technology that never came. Orthopaedic surgeons, more so than other specialists, are governed by their sense of touch and their ability to observe range of motion, gate, and overall movement. The disjunction of video interface and audio delay presented a

56 significant barrier. PACS incompatibility often compounded the poor image quality of x-rays. In order for SLC physicians to work with absolute confidence, they often eschewed telemedicine as a clinical tool and scheduled subsequent visits to the facility for an in-person evaluation. Thus, incompatibility of imaging systems essentially forced postponement of care. The telemedicine clinic came to be used and perceived as a stop-gap measure, a time sink, and a major contributor to discontinuity of care. Staff physicians were unsettled by the lack of dedicated support for the system and commented that, “We rely on our IT people here to tell us what they can do. Neither I nor any of my partners have time to go out and figure out which cameras are the best, I can assure you … there’s no way I have time for any of that.” As Tulu Bengisu described in his, Internet-based Telemedicine: An Empirical Investigation of Objective and Subjective Video Quality (50), new clinical decision-making scales are introduced with virtually every new application of telemedicine technology released into the market today. Without these types of measures in place and programmatic ways to resolve technical issues it is hard to overcome quality concerns. At SLC, there were no documented measurements put into place at implementation and subsequently nothing upon which to base future upgrade decisions other than anecdotal evidence and ad-hoc IT support.

57

Conclusions This study was conducted in order to gain a deeper understanding of the drivers of telemedicine utilization among staff physicians at Shriners Hospitals for Children, Salt Lake City. The findings illustrate how Physician Perception of telemedicine at SLC formed over eight years of use. Using Mezni and Zeribi’s Individual Acceptance model for analysis, we found that Individual factors dominated the decision-making process of SLC staff and that equally important Technical and Managerial factors were simply not present or not sufficiently powerful to influence optimal usage of the technology. This imbalance manifested itself over the lifespan of the telemedicine program. Conscious avoidance of many of the technology’s functions and altered behavior by the attending staff resulted in a precipitous decline in the success of the program.

Themes that emerged out of the focus group indicate that SLC staff physicians were highly concerned with their professional status. Each physician was aware that her mastery of the technology could somehow impact how she would be viewed by administrators, peers, and patients. This sensitivity to professional status represented the most potent driver in a complex decision-making process. However, professional status alone did not determine how telemedicine would go on to be utilized. SLC staff felt compelled to use telemedicine to preserve their autonomy. Yet they capitulated and changed their behavior. This form of self-efficacy was by no means motivated by pride or obstinacy. On the contrary, the change in

58 behavior was an attempt to minimize potential risks staff felt telemedicine introduced to their practice. The tactile nature of orthopedics coupled with the extreme sensitivity of pediatric care, concerned SLC staff a great deal. On balance, they felt the technology severely constrained how they diagnosed and treated their patients. They responded by subverting the process. The cause for this deviation can be found in the theoretical framework within which technology adopters operate. According to Mezni and Zeribi’s model, SLC’s telemedicine program lacked managerial support, in the form of internal champions and training which could serve as a continual troubleshooting/feedback loop for the organization. Additionally, SLC’s failure to establish quality metrics to help baseline and resolve technical issues plagued the program from its inception.

Over the lifespan of the telemedicine program, enough counter-intuitive and costly experience had accumulated so that the staff, as a whole and in tacit acknowledgement of its autonomy, began to change how telemedicine would be employed. For treating new patients, for example, telemedicine was changed from a screening technology to a triage mechanism and finally to an appointment setting tool. Its effectiveness as a simple follow-up and fitments tool was challenged due to the poor image and voice quality and constant concern over clinic-side medical skills. Through it all, SLC staff remained convinced of the program’s cost-effectiveness. Statements were made suggesting that the Shrine Temples saved money year-over-year on transportation and that the hospital was able to reallocate budget for other purposes thanks to cost-savings in the clinics. Yet, there is little evidence to

59 support these statements. SLC administration never published its operating results or established cost-benefit metrics to track these operations and any material cost savings were anecdotal at best.

Study Limitations and Areas for Further Study

There are two limitations that need to be acknowledged and addressed regarding the present study. The first limitation concerns the sample size of this research project. As detailed, this study was based on a single focus group with a small, yet complete expert sample. It did not necessarily reflect the experience of other physicians within Shriners Hospitals system. Within the Shriners Hospitals system, telemedicine programs may vary considerably in terms of the staff expertise and the extent to which telemedicine programs have been developed and supported by regional leadership. Additionally, I purposefully excluded the input offered by Respondent Six who did not meet the criteria for the study yet may have had additional insights into telemedicine acceptance at SLC. Future research, therefore, should more fully explore the variety of program practices at different Shriners Hospitals to include practitioners outside of the orthopedic specialty isolated for this study. The second limitation has to do with the extent to which the findings can be generalized beyond this study. While the applied theory and emergent themes were effective in helping identify universal characteristics found in the telemedicine experience at SLC, this study was statistically too small for broad generalizations. However, the provider organizations performing gap

60 analysis in anticipation of launching telemedicine can benefit from these findings. Additionally, while provider acceptance of the technology has been shown to be an important component to successful telemedicine adoption, in many respects the patient is an equal and formidable actor in this delivery system. Future work should be done to introduce qualitative and quantitative studies of acceptance of telemedicine and it potential impact of outcomes from patient perspective. The findings from this study certainly suggest that patient satisfaction drove physician adherence. More research should be done to understand the subtle influence the patient has on utilization.

Recommendations Studies show that a great deal of power resides in the technology-resistant physician (51). SLC administration, therefore, should not take the physician’s perception of her status within a substantial normative environment for granted. Indeed, administrators should attempt to channel a physician’s desire to innovate and improve by introducing technologies that, at least conceptually, can be shown to enhance clinical capabilities. Further, hospital administrators should become equal and opposite partners in technology adoption by supporting: 1. collaborative end-user feedback and internal leadership via a clinical champion 2. outcome-based research and user-acceptance testing to support on-going innovation

61 3. continuous investment in technology upgrades and user training to overcome process disruption

Without effective training, support, and outcomes research to show that benefits can indeed be derived from proscribed use of the technology, so much of the evidence to support investment is lost. SLC’s experience is an example of how a promising program lost momentum due to a number of gaps such as the loss of its clinical champion and poor execution in identifying and resolving technical issues. While not every provider organization will share the same viewpoints and experiences as found in this study, a thorough exploration of staff expectations and values could go a long way in determining how the Individual Acceptance model can be applied to any organization’s telemedicine program.

62

References

1

()

2

()

3

()

4

()

5

()

6

()

7

()

8

()

9

()

10

()

11

()

12

()

13

()

Hannah, J. 2008. Use of Telemedicine Expanding in Ohio's Rural Areas, Appalachia. Chillicothe Gazette. May, C., et al. 2005. Telemedicine and the 'Future Patient': Risk, Governance and Innovation. The Economic & Social Research Council. United States Department of Health and Human Services, H. R. a. S. A. Retrieved December 4, 2008, from http://www.hrsa.gov/telehealth/ . Association of Telehealth Service Providers. Newsletter, T. T. T. e. Retrieved December 4, 2008, from http://www2.telemedtoday.com/glossary/index.html . Joint Commission Resources.2003. Existing Requirements for Telemedicine Practitioners Explained. Joint Commission Perspectives 23(2):pp.4. Field, M. J. ed., 1996. Telemedicine: A Guide to Assessing Telecommunications for Healthcare. Washington, D.C.: National Academy Press. McConnochie, K. 2008. Telemedicine Could Eradicate Many Expensive ED Visits. Health-e-Access. Rochester, University of Rochester Medical Center. Shriners Headquarters 2008. What is the Relationship Between Shriners and Shriners Hospitals? from http://www.shrinershq.org/Shrine/Membership . Business Editors. 2000. e-MedSoft.com Designated `Vendor of Choice' by Shriners Hospitals for Children Telemedicine Program. Business Wire: Shriners HQ. 2006. Pamphlet: Delivering Care from a Distance. Shriners Hospitals for Children from www.shrinershq.org. United States House of Representatives, 2005. Telemedicine Activities at the Department of Health and Human Services Subcommittee on Health Committee on Veterans Affairs. Agency for Healthcare Research and Quality. Rockville, MD. American Medical Informatics Association. Classic Paper Series: POI-WG Collection of Recommended Papers and Information Sources. from http://www.amia.org/mbrcenter/wg/poi/docs/recommended_papers.asp Lapointe, Liette and Rivard, Suzanne. 2006. Getting Physicians to Accept New Information Technology: Insights from Case Studies. Canadian Medical Association Journal 174(11), pp.1583-1584.

14 ()

Whitten, P. et al. 2007. The Role of Provider Perceptions in Deploying an Electronic Medical Record at Michigan State University. American Telemedicine Association 2007 International. Nashville, TN.

15

()

16

()

Reynolds, B., et al. 2007. Provider Acceptance of a Remote Trauma Consultation Network. American Telemedicine Association 2007 International. Nashville, TN. Melzer, S. et al. 2008. Experience with Pediatric Telemedicine Services in a

Regional Network. American Telemedicine Association 2008 International. Seattle, WA. 17

()

18

()

19

()

20

()

21

()

22

()

23

()

24

()

25

()

Gagnon, M-P, et al. 2003. An Adaptation of the Theory of Interpersonal Behaviour to the Study of Telemedicine Adoption by Physicians. International Journal of Medical Informatics (2-3),pp.103-115. Lau, C.et al. 2001. Web-based Home Telemedicine System for Orthopaedics (Univ. of Washington). SPIE: The International Society for Optical Engineering 4319. 693-698. Ricci, W.M. and Borrelli J. 2004. Teleradiology in Orthopaedics. Clinical Orthopaedic and Related Research April, 421, pp.64-69. Vuolio, S. et al 2003. Videoconferencing for Orthopaedic Outpatients: One-year Follow-up. Journal of Telemedicine and Telecare. Royal Society of Medicine Press 9(1), pp.8-11. Ono, C.M. and Lindsey, J.L. 2004. Shriners Hospitals for Children, Honolulu's Experience with Telemedicine: Program Implementation, Maintenance, Growth, and Lessons Learned. Hawaii Medical Journal 63(October), pp.296-9. Mezni, H. and Zeribi B.O. 2008. Determinants of the Individual Acceptance of the Telemedicine. Information and Communication Technologies: From Theory to Applications, 3rd International Conference on ICTTA. Rogers, E. M. 2003. Diffusion of Innovations. New York, NY:Free Press.

Succi, M. J. et al 1999. Theory of User Acceptance of Information Technologies: An Examination of Health Care Professionals. 32nd Hawaii International Conference on System Sciences. Hawaii: IEEE Computer Society. Wu, J.H., et al. 2008. Testing the Technology Acceptance Model for Evaluating Healthcare Professionals' Intention to use an Adverse Event Reporting System. International Journal for Quality in Healthcare. 20(2), pp.123-9.

26 ()

Gagnon, M-P. Lamothe, L. et al., 2004. The Impact of Organizational Characteristics on Telehealth Adoption by Hospitals. Proceedings of the 37th Annual Hawaii International Conference on System Sciences, IEEE.

27

()

28

()

29

()

Hu, P. J. and Chau., P.Y. et al. 1999. Physician Acceptance of Telemedicine Technology: An Empirical Investigation. Topics in Health Information Management 19,pp.20-35. Venkatesh, V., Morris, M, et al. 2003. User Acceptance of Information Technology : Toward a Unified View. Management Information Systems Quarterly. 27(3), pp.425478. http://www.americantelemed.org/i4a/pages/index.cfm?pageID=3308

30

()

31

()

32

()

33

()

34

()

35

()

36

()

37

()

38

()

39

()

Smith, Douglas L., 2005. The Influence of Financial Factors on the Deployment of Telemedicine. Journal of Health Care Finance, Fall 32(1), pp.16-27. Broens, T.H. et al., 2007. Determinants of Successful Telemedicine Implementations: a Literature Study. Journal of Telemedicine and Telecare. Royal Society of Medicine Press.13, pp.303-309. Congressional Budget Office, 2008. Budget Report: H.R. 6331: Medicare Improvement for Patients and Providers Act of 2008. U. S. H. o. R. Committee on Ways and Means. Grigsby, J., et al., 2002. The Diffusion of Telemedicine. Telemedicine Journal and e-Health 8(1), pp.79-94. Catterall, M. and Maclaran, P., 1997. Focus Group Data and Qualitative Analysis Programs: Coding the Moving Picture as Well as the Snapshots. Sociological Research Online. 1(1). Fowler, F. J. ,1993. Survey Research Methods. Thousand Oaks, CA: Sage Publications Inc. Patton, M. Q., 2002. Qualitative Research and Evaluation Methods. London:Sage. Joss, S. and Durant, J, ed. 1995. Citizen Participation in Science: the Role of Consensus Conferences in Europe London:Science Museum. Kitzinger, J., 1994. The Methodology of Focus Groups: The Importance of Interaction Between Participants. Sociology of Health and Illness 16(1), pp.103 – 121. Bristol, T. and Fern, E., 2003. The Effects of Interaction on Consumers' Attitudes in Focus Groups. Psychology and Marketing. 20(5), pp.433-454.

40 ()

Sim, J., 2001. Collecting and Analysing Qualitative Data: Issues Raised by the Focus Group. Journal of Advanced Nursing. 28(2), pp.345 - 352.

41

()

42

()

43

()

44

()

45

()

Goldman, A. E. "The Group Depth Interview." Journal of Marketing. 1962.

Sarker, S., Lau, F., and Sahay, S., 2000. Building an Inductive Theory of Collaboration in Virtual Teams: An Adapted Grounded Theory Approach. Proceedings of the 33rd Hawaii International Conference on System Sciences. IEEE Hawaii. Bales, R. F. ,1953. The Equilibrium Problem in Small Groups. Parsons, Shils, et al ed. Working Papers in the Theory of Action. Free Press. pp.111-61. Glaser, B. and Strauss, A., 1967. The Discovery of Grounded Theory: Strategies for Qualitative Research. Chicago: Aldine Publishing. Tuckman, B. W., 1965. Developmental Sequence in Small Groups. Psychological Bulletin 63(6), pp.384-399.

46

()

Catterall, M. and Maclaran, P., 1997. Focus Group Data and Qualitative Analysis Programs: Coding the Moving Picture as Well as the Snapshots. Sociological Research Online 1(1).

47 ()

Ryan, G.W. and Russell B.H., 2000. Data Management and Analysis Methods. Denzin, Norman, and Lincoln. ed. Handbook of Qualitative Research, 2nd Edition. Thousand Oaks, California: Sage.

48

()

49

()

50

()

51

()

Yellowlees, P. M., 2005. Successfully Developing a Telemedicine System. Journal of Telemedicine and Telecare. Royal Society of Medicine Press 11, pp.331-335. Armstrong, P., 1999. Shriners Hospitals for Children Telemedicine Initiative. Telemedicine Journal and e-Health 1(March), pp.76-119. Bengisu, T., 2006. Internet-based Telemedicine: An Experimental Study to Provide Decision Support Using Real-time Assessment of Video Quality. Claremont Graduate School. Patton, G. A., 2001. The Two-edged Sword: How Technology Shapes Medical Practice. Physician Executive. March.