tugrul˜u.˜daim nima˜a. behkami orhun˜m.˜kök … · 2020. 5. 5. · nima˜a. behkami...
TRANSCRIPT
Innovation Technology and Knowledge Management
Tugrul U DaimNima A BehkamiNuri BasogluOrhun M KoumlkLiliya Hogaboam
Healthcare Technology Innovation AdoptionElectronic Health Records and Other Emerging Health Information Technology Innovations
Innovation Technology and Knowledge Management
Series Editor Elias G Carayannis George Washington University Washington DC USA
More information about this series at httpwwwspringercomseries8124
Tugrul U Daim bull Nima A Behkami Nuri Basoglu bull Orhun M Koumlk Liliya Hogaboam
Healthcare Technology Innovation Adoption Electronic Health Records and Other Emerging Health Information Technology Innovations
ISSN 2197-5698 ISSN 2197-5701 (electronic) Innovation Technology and Knowledge Management ISBN 978-3-319-17974-2 ISBN 978-3-319-17975-9 (eBook) DOI 101007978-3-319-17975-9
Library of Congress Control Number 2015942128
Springer Cham Heidelberg New York Dordrecht London copy Springer International Publishing Switzerland 2016 This work is subject to copyright All rights are reserved by the Publisher whether the whole or part of the material is concerned specifi cally the rights of translation reprinting reuse of illustrations recitation broadcasting reproduction on microfi lms or in any other physical way and transmission or information storage and retrieval electronic adaptation computer software or by similar or dissimilar methodology now known or hereafter developed The use of general descriptive names registered names trademarks service marks etc in this publication does not imply even in the absence of a specifi c statement that such names are exempt from the relevant protective laws and regulations and therefore free for general use The publisher the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication Neither the publisher nor the authors or the editors give a warranty express or implied with respect to the material contained herein or for any errors or omissions that may have been made
Printed on acid-free paper
Springer International Publishing AG Switzerland is part of Springer Science+Business Media (wwwspringercom)
Tugrul U Daim Department of Engineering
and Technology Management Portland State University Portland OR USA
Nuri Basoglu Department of Industrial Design İzmir Institute of Technology Urla Izmir Turkey
Liliya Hogaboam Department of Engineering
and Technology Management Portland State University Portland OR USA
Nima A Behkami Merck Research Laboratories Boston MA USA
Orhun M Koumlk Ernst and Young Advisory Istanbul Turkey
v
Series Foreword
The Springer book series Innovation Technology and Knowledge Management was launched in March 2008 as a forum and intellectual scholarly ldquopodiumrdquo for globallocal transdisciplinary transsectoral publicndashprivate and leadingldquobleedingrdquo edge ideas theories and perspectives on these topics
The book series is accompanied by the Springer Journal of the Knowledge Economy which was launched in 2009 with the same editorial leadership
The series showcases provocative views that diverge from the current ldquoconven-tional wisdomrdquo that are properly grounded in theory and practice and that consider the concepts of robust competitiveness 1 sustainable entrepreneurship 2 and demo-cratic capitalism 3 central to its philosophy and objectives More specifi cally the aim of this series is to highlight emerging research and practice at the dynamic intersection of these fi elds where individuals organizations industries regions and nations are harnessing creativity and invention to achieve and sustain growth
1 We defi ne sustainable entrepreneurship as the creation of viable profi table and scalable fi rms Such fi rms engender the formation of self-replicating and mutually enhancing innovation networks and knowledge clusters (innovation ecosystems) leading toward robust competitiveness (EG Carayannis International Journal of Innovation and Regional Development 1(3) 235ndash254 2009) 2 We understand robust competitiveness to be a state of economic being and becoming that avails systematic and defensible ldquounfair advantagesrdquo to the entities that are part of the economy Such competitiveness is built on mutually complementary and reinforcing low- medium- and high- technology and public and private sector entities (government agencies private fi rms universities and nongovernmental organizations) (EG Carayannis International Journal of Innovation and Regional Development 1(3) 235ndash254 2009) 3 The concepts of robust competitiveness and sustainable entrepreneurship are pillars of a regime that we call ldquo democratic capitalism rdquo (as opposed to ldquopopular or casino capitalismrdquo) in which real opportunities for education and economic prosperity are available to all especiallymdashbut not onlymdashyounger people These are the direct derivatives of a collection of topdown policies as well as bottom-up initiatives (including strong research and development policies and funding but going beyond these to include the development of innovation networks and knowledge clusters across regions and sectors) (EG Carayannis and A Kaloudis Japan Economic Currents p 6ndash10 January 2009)
vi
Books that are part of the series explore the impact of innovation at the ldquomacrordquo (economies markets) ldquomesordquo (industries fi rms) and ldquomicrordquo levels (teams indi-viduals) drawing from such related disciplines as fi nance organizational psychol-ogy research and development science policy information systems and strategy with the underlying theme that for innovation to be useful it must involve the shar-ing and application of knowledge
Some of the key anchoring concepts of the series are outlined in the fi gure below and the defi nitions that follow (all defi nitions are from EG Carayannis and DFJ Campbell International Journal of Technology Management 46 3ndash4 2009)
GlobalSystemicmacro level
Democraticcapitalism
Structural andorganizationalmeso level
Innovationnetworks
Entrepreneurialuniversity Globallocal
Individualmicro level
Local
Creativemilieus
Academicfirm
Democracyofknowledge
Mode 3 Quadruplehelix
Knowledgeclusters
Sustainableentrepreneurship
Entrepreneuremployeematrix
Conceptual profi le of the series Innovation Technology and Knowledge Management
bull The ldquoMode 3rdquo Systems Approach for Knowledge Creation Diffusion and Use ldquoMode 3rdquo is a multilateral multinodal multimodal and multilevel systems approach to the conceptualization design and management of real and virtual ldquoknowledge-stockrdquo and ldquoknowledge-fl owrdquo modalities that catalyze accelerate and support the creation diffusion sharing absorption and use of cospecialized knowledge assets ldquoMode 3rdquo is based on a system-theoretic perspective of socio-economic political technological and cultural trends and conditions that shape the coevolution of knowledge with the ldquoknowledge-based and knowledge-driven globallocal economy and societyrdquo
bull Quadruple Helix Quadruple helix in this context means to add to the triple helix of government university and industry a ldquofourth helixrdquo that we identify as the ldquomedia-based and culture-based publicrdquo This fourth helix associates with ldquomediardquo ldquocreative industriesrdquo ldquoculturerdquo ldquovaluesrdquo ldquolife stylesrdquo ldquoartrdquo and per-haps also the notion of the ldquocreative classrdquo
Series Foreword
vii
bull Innovation Networks Innovation networks are real and virtual infrastructures and infratechnologies that serve to nurture creativity trigger invention and cata-lyze innovation in a public andor private domain context (for instance govern-mentndashuniversityndashindustry publicndashprivate research and technology development coopetitive partnerships)
bull Knowledge Clusters Knowledge clusters are agglomerations of cospecialized mutually complementary and reinforcing knowledge assets in the form of ldquoknowledge stocksrdquo and ldquoknowledge fl owsrdquo that exhibit self-organizing learning- driven dynamically adaptive competences and trends in the context of an open systems perspective
bull Twenty-First Century Innovation Ecosystem A twenty-fi rst century innovation ecosystem is a multilevel multimodal multinodal and multiagent system of sys-tems The constituent systems consist of innovation metanetworks (networks of innovation networks and knowledge clusters) and knowledge metaclusters (clus-ters of innovation networks and knowledge clusters) as building blocks and orga-nized in a self-referential or chaotic fractal knowledge and innovation architecture 4 which in turn constitute agglomerations of human social intel-lectual and fi nancial capital stocks and fl ows as well as cultural and technologi-cal artifacts and modalities continually coevolving cospecializing and cooperating These innovation networks and knowledge clusters also form reform and dissolve within diverse institutional political technological and socioeconomic domains including government university industry and non-governmental organizations and involving information and communication tech-nologies biotechnologies advanced materials nanotechnologies and next-generation energy technologies
Who is this book series published for The book series addresses a diversity of audiences in different settings
1 Academic communities Academic communities worldwide represent a core group of readers This follows from the theoreticalconceptual interest of the book series to infl uence academic discourses in the fi elds of knowledge also carried by the claim of a certain saturation of academia with the current concepts and the postulate of a window of opportunity for new or at least additional con-cepts Thus it represents a key challenge for the series to exercise a certain impact on discourses in academia In principle all academic communities that are interested in knowledge (knowledge and innovation) could be tackled by the book series The interdisciplinary (transdisciplinary) nature of the book series underscores that the scope of the book series is not limited a priori to a specifi c basket of disciplines From a radical viewpoint one could create the hypothesis that there is no discipline where knowledge is of no importance
2 Decision makers mdash private academic entrepreneurs and public ( governmental subgovernmental ) actors Two different groups of decision makers are being addressed simultaneously (1) private entrepreneurs (fi rms commercial fi rms
4 EG Carayannis Strategic Management of Technological Learning CRC Press 2000
Series Foreword
viii
academic fi rms) and academic entrepreneurs (universities) interested in opti-mizing knowledge management and in developing heterogeneously composed knowledge-based research networks and (2) public (governmental subgovern-mental) actors that are interested in optimizing and further developing their poli-cies and policy strategies that target knowledge and innovation One purpose of public knowledge and innovation policy is to enhance the performance and com-petitiveness of advanced economies
3 Decision makers in general Decision makers are systematically being supplied with crucial information for how to optimize knowledge-referring and knowledge- enhancing decision-making The nature of this ldquocrucial informationrdquo is conceptual as well as empirical (case-study-based) Empirical information highlights practical examples and points toward practical solutions (perhaps remedies) conceptual information offers the advantage of further driving and further-carrying tools of understanding Different groups of addressed decision makers could be decision makers in private fi rms and multinational corporations responsible for the knowledge portfolio of companies knowledge and knowl-edge management consultants globalization experts focusing on the interna-tionalization of research and development science and technology and innovation experts in universitybusiness research networks and political scien-tists economists and business professionals
4 Interested global readership Finally the Springer book series addresses a whole global readership composed of members who are generally interested in knowl-edge and innovation The global readership could partially coincide with the communities as described above (ldquoacademic communitiesrdquo ldquodecision makersrdquo) but could also refer to other constituencies and groups
Elias G Carayannis
Series Foreword
ix
Pref ace
Healthcare costs have been increasing dramatically over the last years This volume explores the adoption of health technology innovations designed to streamline the service delivery and thus reduce costs and increase quality
The fi rst part reviews theories and applications for the diffusion of healthcare technology innovations The second and third parts focus on electronic health records (EHR) which is the leading technology innovation in the healthcare sector The second part develops evaluation models and the third part analyzes an adoption case These models and the case provide a set of factors which need further attention by those responsible for implementing such technologies
Portland OR USA Tugrul U Daim Boston MA USA Nima A Behkami Izmir Turkey Nuri Basoglu Istanbul Turkey Orhun M Koumlk Portland OR USA Liliya Hogaboam
xi
Part I A Dynamic Capabilities Theory-Based Innovation Diffusion Model for Spread of Health Information Technology in the USA Nima A Behkami and Tugrul U Daim
1 Introduction to the Adoption of Health Information Technologies 3 Nima A Behkami and Tugrul U Daim 11 The Healthcare Crisis in the United States 3 12 Government Efforts and HIT Meaningful-Use Initiative 4
121 State of Diffusion Research General and Health IT 5 References 7
2 Background Literature on the Adoption of Health Information Technologies 9 Nima A Behkami and Tugrul U Daim 21 Overview of the Healthcare Delivery System 9 22 A Methodological Note 10 23 The Critical Stakeholders and Actors 10
231 Care Providers 11 232 Government 12 233 Patients and Their Family and Care Givers 13 234 Payers 13 235 HITInnovation Suppliers 14
24 Attributes of the Stakeholders 15 25 Important Factors Effecting Diffusion and Adoption for HIT 15
251 Barriers and Infl uences 17 252 Tools Methods and Theories 19 253 Policy Making 20 254 Hospital Characteristics and the Ecosystem 21 255 Adopter Attitudes Perceptions and Characteristics 22 256 Strategic Management and Competitive Advantage 23
Contents
xii
257 Innovation Champions and Their Aids 23 258 Workfl ow and Knowledge Management 24 259 Timing and Sustainability 24 2510 Modeling and Forecasting 25 2511 Infusion 25 2512 Social Structure and Communication
Channels 25 26 The Need for Multiple Perspectives in Research 26 27 Linstonersquos Multiple Perspectives Method 26 28 The ldquo4 + 1 Viewrdquo Model for Software Architectures 28 29 Categorization of Important Factors in HIT Adoption
Using Multi-perspectives 28 References 30
3 Methods and Models 37 Nima A Behkami and Tugrul U Daim 31 Proposed Model Overview and Justifi cation 37 32 Modeling Approach 39 33 Diffusion Theory 40
331 An Innovation 41 332 Recent Diffusion of Innovation Issues 42 333 Limitations of Innovation Research 44
34 Other Relevant Diffusion and Adoption Theories 45 341 The Theory of Reasoned Action 46 342 The Technology Acceptance Model 46 343 The Theory of Planned Behavior 48 344 The Unifi ed Theory of Acceptance
and Use of Technology 48 345 Matching Person and Technology Model 49 346 Technology-Organization-Environment
Framework (TOE) 49 347 Lazy User Model 50
35 Resource-Based Theory Invisible Assets Competencies and Capabilities 50 351 Foundations of Resource-Based Theory 51 352 Seminal Work in Resource-Based Theory 52 353 Invisible Assets and Competencies Parallel Streams
of ldquoResource-Based Workrdquo 53 354 A Complete List of Terms Used to Refer to Factors
of Production in Literature 54 355 Typology and Classifi cation of Factors of Production 55
36 Modeling Component Descriptions 55 361 Model 56 362 Diagram 56 363 View 56
Contents
xiii
364 Domain 56 365 Modeling Language 56 366 Tool 57 367 Simulation 57
37 Modeling Technique Trade-Off Analysis for Proposed HIT Diffusion Study 57 371 Soft System Methodology 60 372 Structured System Analysis and Design Method 61 373 Business Process Modeling 61 374 System Dynamics (SD) 61 375 System Context Diagram and Data Flow Diagrams
and Flow Charts 62 376 Unifi ed Modeling Language 64 377 SysML 66
38 Conclusions for Modeling Methodologies to Use 66 39 Qualitative Research Grounded Theory and UML 67
391 An Overview of Qualitative Research 67 392 Grounded Theory and Case Study Method Defi nitions 68 393 Using Grounded Theory and Case Study Together 70 394 Grounded Theory in Information Systems (IS)
and Systems Thinking Research 71 395 Criticisms of Grounded Theory 72 396 Current State of UML as a Research Tool and Criticisms 73 397 To UML or Not to UML 73 398 An Actual Example of Using Grounded Theory
in Conjunction with UML 73 References 76
4 Field Test 83 Nima A Behkami and Tugrul U Daim 41 Introduction and Objective 83 42 Background Care Management Plus 84
421 Signifi cance of the National Healthcare Problem 84 422 Preliminary CMP Studies at OHSU 85
43 Research Design 86 431 Overview 86 432 Objectives 86 433 Methodology and Data Collection 87 434 Analysis 90 435 Results and Discussion 91 436 Simulation A System Dynamics Model
for HIT Adoption 100 References 110
Contents
xiv
5 Conclusions 113 Tugrul U Daim and Nima A Behkami 51 Overview and Theoretical Contributions 113 52 Recommended Proposition for Future Research 123 References 123
Part II Evaluating Electronic Health Record Technology Models and Approaches Liliya Hogaboam and Tugrul U Daim
6 Review of Factors Impacting Decisions Regarding Electronic Records 127 Liliya Hogaboam and Tugrul U Daim 61 The Adoption of EHR with Focus on Barriers and Enablers 127 62 The Selection of EHR with Focus on Different Alternatives 133 63 The Use of EHR with Focus on Impacts 137 References 144
7 Decision Models Regarding Electronic Health Records 151 Liliya Hogaboam and Tugrul U Daim 71 The Adoption of EHR with Focus on Barriers and Enables 151
711 Theory of Reasoned Action 151 712 Technology Acceptance Model 152 713 Theory of Planned Behavior 154
72 The Selection of EHR with Focus on Different Alternatives 159 721 Criteria 160
73 The Use of EHR with Focus on Impacts 172 References 178
Part III Adoption Factors of Electronic Health Record Systems Orhun M Koumlk Nuri Basoglu and Tugrul U Daim
8 Adoption Factors of Electronic Health Record Systems 189 Orhun Mustafa Koumlk Nuri Basoglu and Tugrul U Daim 81 Introduction 18982 Literature Review 191
821 Electronic Health Records 191822 Technology Adoption Models 192823 Health Information System Adoption 195
83 Framework 19984 Methodology 206
841 Qualitative Study 206842 Expert Focus Group Study 207843 Pilot Study 207844 Quantitative Field Survey 208
Contents
xv
85 Findings 209851 Qualitative Study Findings 209852 Expert Focus Group Findings 213853 Pilot Study Findings 214854 Quantitative Field Survey Study Findings 217
86 Conclusion 230861 Limitations 231862 Implications 231
87 Appendices 232871 1 Interview Questions 232872 2 Expert Focus Group Questionnaire 233873 3 Factor Analysis Results for Pilot 236 874 4 Factor Analysis Results 238875 5 Regression Results 242
References 245
Contents
Part I A Dynamic Capabilities Theory-Based
Innovation Diffusion Model for Spread of Health Information Technology in the USA
Nima A Behkami and Tugrul U Daim
Abstract Real adoption (aka successful adoption) of an innovation occurs when an adopter has become aware of the innovation the conditions for using it make sense and the adopter has developed the capabilities to truly and meaningfully implement and use the innovation While making critical contributions existing diffusion the-ory research have not examined capabilities and conditions as part of the adoption framework this proposal helps bridge this gap This has been done by developing a new conceptual model based on Rogersrsquo classical diffusion theory with new exten-sions for capabilities The effort included selecting and integrating the appropriate methodology for data collection (case study) analysis (multi-perspectives) model development (diffusion theory dynamic capabilities) model analysis and documen-tation (Unifi ed Modeling Language) and simulation (system dynamics) In this research the new extensions to diffusion theory are studied in the context of health information technology (HIT) innovation adoption and diffusion in the USA According to the US Department of Health and Human Services (HHS) defi -nition HIT allows comprehensive management of medical information and its secure exchange between healthcare consumers and providers The promise of HIT adoption lies in reducing the cost of care delivery while increasing the quality of patient care therefore its accelerated rate of diffusion is of top priority for the gov-ernment and society
Chapter 1 introduces the crisis in the US healthcare system defi nition of HIT and the motivations for studying and advocating acceleration of HIT diffusion sup-ported especially by the government of the USA Chapter 2 describes an overview of the health delivery system and the critical stakeholders involved The stakehold-ers and their attributes are described in detail This chapter also identifi es factors effecting HIT diffusion and reviews research literature for example for factors such as barriers infl uences adopter characteristics and more The other main point dis-cussed in Chap 1 is that in order to make analysis comprehensive there is a need to look at the research area from a multi-perspective point of view The two popular methodologies of ldquo Linstonersquos Multi-perspectives rdquo and the ldquo 4+1 View Model rdquo for software architectures are examined Finally in Chap 1 important factors identifi ed
2
earlier in the chapter are categorized using Linstonersquos perspectives to show appropriateness of using multi-perspective for analysis
Chapter 3 describes the proposed model and the justifi cations for using the theories and methodologies used to support the research First a detailed description of the new proposed extensions to diffusion theory is presented that include dynamic capabilities and conditions The proposed is supported and reasoned for using fi ve main sections in the chapter that include describing diffusion theory in detail com-paring and evaluating other potential adoption theories exploring resource-based theory and capability research modeling technique trade-off analysis and quality research methods including usage of grounded theory with UML
Chapter 4 is the description of the fi eld study conducted to demonstrate the fea-sibility of research proposal The fi eld study was conducted for examining the adop-tion process for a care management product built and dissemination through Oregon Health and Science University named CMP (Care Management Plus) CMP is a HIT-enabled care model targeted for older adults and patients with multiple chronic conditions CMP components include software clinical business processes and training For this research secondary data from site (clinic) readiness survey and in- person expert interviews were used to collect data Through case study and the-matic analysis methods the data was extracted and analyzed An analysis model was built using data collected that demonstrated the structural and behavioral aspects of the system using UML and a classifi cation of capabilities Later in the chapter to demonstrate the usefulness of system dynamics a simple Bass diffusion model for spread of innovations through advertising was used to estimate dissemination of CMP using data from contact management at OHSU
Chapter 5 concludes the report and the feasibility study with the discovery that through examination of HIT adoption data indeed there is a need for extension of diffusion theory to explain organizational adoption more accurately Dynamics capabilities are an appropriate candidate for integration into diffusion theory Coupling the types of case study andor grounded theory methods with using UML makes valuable strides in studying organization and societal processes And fi nally that system dynamics method can successfully be used as a partner for scenario analysis and forecasting for a wide range of purposes This chapter concludes the report by stating propositions for future research
A Dynamic Capabilities Theory-Based Innovation Diffusion Model for Spreadhellip
3copy Springer International Publishing Switzerland 2016 TU Daim et al Healthcare Technology Innovation Adoption Innovation Technology and Knowledge Management DOI 101007978-3-319-17975-9_1
Chapter 1 Introduction to the Adoption of Health Information Technologies
Nima A Behkami and Tugrul U Daim
11 The Healthcare Crisis in the United States
Due to changing population demographics and their state of health the healthcare system in the United States is facing monumental challenges For example patients suffering from chronic illnesses account for approximately 75 of the nationrsquos healthcare-related expenditures A patient on Medicare with fi ve or more illnesses will visit 13 different outpatient physicians and fi ll 50 prescriptions per year (Friedman Jiang Elixhauser amp Segal 2006 ) As the number of a patientrsquos condi-tions increases the risk of hospitalizations grows exponentially (Wolff Starfi eld amp Anderson 2002 ) While the transitions between providers and settings increase so does the risk of harm from inadequate information transfer and reconciliation of treatment plans A third of these costs may be due to inappropriate variation and failure to coordinate and manage care (Wolff et al 2002 ) As costs continue to rise the delivery of care must change to meet these costs
This has brought about a renewed interest from various government public and private entities for proposing solutions to the healthcare crisis (Technology health care amp management in the hospital of the future 2003 ) which is helping fuel dif-fusion research in healthcare Technology advances and the new ways of bundling technologies to provide new healthcare services is also contributing to interest in Health Information Technology (HIT) research (E-Health Care Information Systems An Introduction for Students and Professionals 2005 ) The promise of applying technology to healthcare lies in increasing hospital effi ciency and accountability and decreasing cost while increasing quality of patient care
N A Behkami Merck Research Laboratories Boston MA USA
T U Daim () Portland State University Portland OR USA e-mail tugruludaimpdxedu
4
( HealthIT hhs gov ) Therefore itrsquos imperative to study how technology in particu-lar HIT is being adopted and eventually defused in the healthcare sector to help achieve the nationrsquos goals Rogers in his seminal work has highlighted his concern for almost overnight drop and near disappearance of diffusion studies in such fi elds as sociology and has called for renewed efforts in diffusion research (Rogers 2003 ) Others have identifi ed diffusion as the single most critical issue facing our modern technological society (Green Ottoson Garciacutea amp Hiatt 2009 )
According to the US Department of Health and Human Services defi nition Health Information Technology allows comprehensive management of medical information and its secure exchange between health care consumers and providers ( HealthIT hhs gov ) Information Communication Technology (ICT) and Health Information Technology (HIT) are two terms that are often used interchangeably and generally encompass the same defi nition It is hoped that use of HIT will lead to reduced costs and improved quality of care (Heinrich 2004 ) Various policy bod-ies including Presidents Obamarsquos administration ( Organizing for America ) and other independent reports have called for various major healthcare improvements in the United States by the year 2025 ( The Commonwealth Fund ) In describing these aspirations almost always a call for accelerating the rate of HIT adoption and diffu-sion is stated as one of the top fi ve levers for achieving these improvement goals ( Organizing for America ) Hence it is of critical importance to study and understand upstream and downstream dynamics of environments that will enable successful diffusion of HIT innovations
12 Government Efforts and HIT Meaningful-Use Initiative
In order to introduce signifi cant and measurable improvements in the populations health in the United States various government and private entities seek to trans-form the healthcare delivery system by enabling providers with real-time access to medical information and tools to help increase quality and safety of care ( US Department of Health and Human Services ) Performance improvement pri-orities have focused on patient engagement reduction of racial disparities improved safety increased effi ciency coordination of care and improved popula-tion health ( US Department of Health and Human Services ) Using these priori-ties the Health Information Technology (HIT) Policy Committee a Federal Advisory Committee (FACA) to the US Department of Health and Human Services (HHS) has initiated the ldquomeaningful userdquo intuitive for adoption of Electronic Health Records (EHR)
Fueled by the $19 billion investment available through the American Recovery and Reinvestment Act of 2009 (Recovery Act) efforts are in full swing to accelerate the national adoption and implementation of health information technology (HIT) ( Assistant Secretary for Public Affairs ) The Recovery act authorizes the Centers for Medicare amp Medicaid Services (CMS) to provide payments to eligible physicians
NA Behkami and TU Daim
5
and hospitals who succeed in becoming ldquomeaningful usersrdquo of an electronic health record (EHR) Incentive payments begin in 2011 and phase out by 2015 nonadopt-ing providers will be subject to fi nancial penalties under Medicare ( US Department of Health and Human Services ) Medicare is a social insurance program adminis-tered by the United States government providing health insurance to people aged 65 and over or individuals with disabilities Similarly Medicaid provides insurance for low-income families ( US Department of Health amp Human Services Centers for Medicare amp Medicaid Services )
CMS will work closely with the Offi ce of the National Coordinator and other parts of HHS to continue defi ning incentive programs for meaningful use The Healthcare Information and Management Systems Society (HIMSS) recommend that a mature defi nition for ldquomeaningful use of certifi ed EHR technologyrdquo includes at least the following four attributes (Merrill 2009 )
1 A functional EHR certifi ed by the Certifi cation Commission for Healthcare Information Technology (CCHIT)
2 Electronic exchange of standardized patient data with clinical and administrative stakeholders using the Healthcare Information Technology Standards Panelrsquos (HITSP) interoperability specifi cations and Integrating the Healthcare Enterprisersquos (IHE) frameworks
3 Clinical decision support providing clinicians with clinical knowledge and intelligently- fi ltered patient information to enhance patient care and
4 Capabilities to support process and care measurement that drive improvements in patient safety quality outcomes and cost reductions
While existence of such programs as the meaningful-use initiative is a motiva-tion to consider using an EHR historically adoption has been slow and troublesome (Ash amp Goslin 1997 ) One often cited obstacle is the high cost of implementing Electronic Health Records Since usually incentives for adoption often go to the insurer recouping the cost is diffi cult for providers (Middleton Hammond Brennan amp Cooper 2005 Cherry 2006 Menachemi 2006 ) Other challenges existing in the United States healthcare system include variations in practices and proportion of income realized from adoption (Daim Tarman amp Basoglu 2008 Angst 2007 )
121 State of Diffusion Research General and Health IT
Health Information Technology (HIT) innovations are considered to have great potential to help resolve important issues in healthcare The potential benefi ts include enhanced accessibility to healthcare reduced cost of care and increased quality of care (COECAO 1996 ) However despite such potential many HIT innovations are either not accepted or not successfully implemented Some of the reasons cited include poor technology performance organizational issues and legal barriers (Cho Mathiassen amp Gallivan 2008 ) In general there is agreement amongst
1 Introduction to the Adoption of Health Information Technologies
6
researchers that we donrsquot fully understand what it takes for successful innovations to diffuse into the larger population of healthcare organizations
Diffusion of Innovation (DOI) theory has gained wide popularity in the Information Technology (IT) fi eld for example one study found over 70 IT articles published in IT outlets between 1984 and 1994 that relied on DOI theory (Teng Grover amp Guttler 2002 ) Framing the introduction of new Information Technology (IT) as an organizational innovation information systems (IS) researchers have studied the adoption and diffusion of modern software practices spreadsheet soft-ware customer-based inter-organizational systems database management systems electronic data interchange and IT in general (Teng et al 2002 ) These studies have been conducted at several levels (1) at the level of intra-fi rm diffusion ie diffu-sion of innovation within an organization (2) inter-fi rm diffusion at the industry level (3) overall diffusion of an innovation throughout the economy
The main models used for diffusion of innovation were established by 1970 The main modeling developments in the period 1970 onwards have been in modifying the existing models by adding greater fl exibility to the underlying model in various ways The main categories of these modifi cations are listed below (Meade amp Islam 2006 )
bull The introduction of marketing variables in the parameterization of the models bull Generalizing models to consider innovations at different stages of diffusions in
different countries bull Generalizing the models to consider the diffusion of successive generations of
technology
In most of these contributions the emphasis has been on the explanation of past behavior rather than on forecasting future behavior Examining the freshness of contributions the average age of the marketing forecasting and ORmanagement science references is 15 years the average age of the businesseconomics reference is 19 years (Meade amp Islam 2006 ) Scholars of IT diffusion have been quick to apply the widespread DOI theory to IT but few have carefully analyzed whether it is justifi able to extend the DOI vehicle to explain the diffusion of IT innovations too Similar critical voices have been raised recently against a too simplistic and fi xed view of IT (Robinson amp Lakhani 1975 )
Figure 11 shows the research publications trend in HIT and Diffusion studies (Behkami 2009a 2009b ) which shows a steep increase in interest over the last few years While adopter attitudes adoption barriers and hospital characteristics have been studied in depth other components of DOI theory are under-studied No research had attempted to explain diffusion of innovation through dynamic capabili-ties yet There also have been less than a handful of papers forecasting diffusion with system dynamics methodology Figure 12 summarizes the frequency of themes that emerged from a study that analyzed publications related to HIT Diffusion 80 of the 108 articles examined were published between the years 2004 and 2009 (Behkami 2009a )
NA Behkami and TU Daim
7
References
Angst C (2007) Information technology and its transformational effect on the health care indus-try Dissertation Abstracts International Section A Humanities and Social Sciences
Ash J amp Goslin L (1997) Factors affecting information technology transfer and innovation dif-fusion in health care Innovation in Technology ManagementmdashThe Key to Global Leadership PICMETrsquo97 Portland International Conference on Management and Technology (pp 751ndash754)
Assistant Secretary for Public Affairs Process begins to defi ne ldquomeaningful userdquo of electronic health records
400
300
200
100
0
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
500articles not in PubMed
articles from PubMed (mostly Biomedical Informatics)
600
700
Fig 11 Cumulative trend of HIT diffusion research publications over the last three decades
0 5 10 15 20 25 30
Social Structure amp Communication Channels
Modeling amp Forecasting
Infusion
Workflow amp Knowledge Management
Timing amp Sustainability
Innovation Champions amp their Aids
Strategic Management amp Competitive Advantage
Adopter Attitudes Perceptions amp Characteristics
Hospital Characteristics amp the Ecosystem
Policy Making
Tools Methods amp Theories
Factors Barriers amp Influences
Fig 12 Number of published articles that address themes generated from review
1 Introduction to the Adoption of Health Information Technologies
8
Behkami N (2009a) Literature review Diffusion amp organizational adoption of healthcare related information technologies amp innovations
Behkami N (2009b) Methodological analysis of Health Information Technology (HIT) diffusion research to identify gaps and emerging topics in literature
COECAO (1996) Telemedicine and IO Medicine Telemedicine A guide to assessing tele-communications for health care Washington National Academies Press
Cherry B (2006) Determining facilitators and barriers to adoption of electronic health records in long-term care facilities UMI Dissertation Services ProQuest Information and Learning Ann Arbor MI
Cho S Mathiassen L amp Gallivan M (2008) From adoption to diffusion of a Telehealth innova-tion Proceedings of the Proceedings of the 41st Annual Hawaii International Conference on System Sciences (p 245) Los Alamitos CA IEEE Computer Society
Daim T U Tarman R T amp Basoglu N (2008) Exploring barriers to innovation diffusion in health care service organizations An issue for effective integration of service architecture and information technologies Hawaii International Conference on System Sciences (p 100) Los Alamitos CA IEEE Computer Society
E-Health care information systems An introduction for students and professionals San Francisco CA Jossey-Bass 2005
Friedman B Jiang H Elixhauser A amp Segal A (2006) Hospital inpatient costs for adults with multiple chronic conditions Medical Care Research and Review 63 327ndash346
Green L W Ottoson J M Garciacutea C amp Hiatt R A (2009) Diffusion theory and knowledge dissemination utilization and integration in public health Annual Review of Public Health 30 151ndash174
HealthIThhsgov Home Heinrich J (2004) HHSrsquos efforts to promote health information technology and legal barriers to
its adoption Meade N amp Islam T (2006) Modelling and forecasting the diffusion of innovationmdashA 25-year
review International Journal of Forecasting 22 519ndash545 Menachemi N (2006) Barriers to ambulatory EHR Who are lsquoimminent adoptersrsquo and how do
they differ from other physicians Informatics in Primary Care 14 101ndash108 Merrill M (2009) HIMSS publishes lsquomeaningful usersquo defi nitions Healthcare IT News Middleton B Hammond W E Brennan P F amp Cooper G F (2005) Accelerating US EHR
adoption How to get there from here Recommendations based on the 2004 ACMI retreat Journal of the American Medical Informatics Association 12
Organizing for America|BarackObamacom|Health Care Robinson B amp Lakhani C (1975) Dynamic price models for new-product planning Management
Science 21 1113ndash1122 Rogers E (2003) Diffusion of innovations (5th ed) New York Free Press Technology health care and management in the hospital of the future Praeger Publishers 2003 Teng J Grover V amp Guttler W (2002) Information technology innovations General diffusion
patterns and its relationships to innovation characteristics IEEE Transactions on Engineering Management 49 13ndash27
The Commonwealth FundmdashHealth policy health reform and performance improvement US Department of Health amp Human Services Centers for Medicare amp Medicaid Services US Department of Health amp Human Services HealthIThhsgov Health IT Policy Committee Wolff J Starfi eld B amp Anderson P G (2002) Expenditures and complications of multiple
chronic conditions in the elderly Archives of Internal Medicine 162 (20) 2269ndash2276
NA Behkami and TU Daim
9copy Springer International Publishing Switzerland 2016 TU Daim et al Healthcare Technology Innovation Adoption Innovation Technology and Knowledge Management DOI 101007978-3-319-17975-9_2
Chapter 2 Background Literature on the Adoption of Health Information Technologies
Nima A Behkami and Tugrul U Daim
21 Overview of the Healthcare Delivery System
The Healthcare Delivery System is defi ned as the comprehensive collection of actors stakeholders and the relationships amongst them which when in action deliver care to the patients create economic value for the participants serve govern-ment interests and service societal needs When thinking about the healthcare deliv-ery system itrsquos benefi cial to think in terms of a value chain Lacking this integrated view in research leads to a one dimensional assessment or fails to consider views of all the stakeholders in illustrating the problem space (Chaudhry et al 2006 ) Figure 21 is an illustration of the Healthcare Delivery System in context of usage adoption and diffusion of HIT centered on the patient provider and payer The fol-lowing sections will describe in detail the signifi cance impact and infl uence of each of the components as it partitions to delivery of healthcare and diffusion of Health Information Technology
N A Behkami Merck Research Laboratories Boston MA USA
T U Daim () Portland State University Portland OR USA e-mail tugruludaimpdxedu
10
22 A Methodological Note
In order to provide a complete description of the healthcare delivery system the model is built and analyzed through three components Objects Relationships and Views The behavior of the system results when these elements collaborate towards a system goal This approach to analysis and decomposition is necessary for effec-tive systems thinking (Sterman amp Sterman 2000 ) To refl ect the static structure and dynamic behavior of these collaborating objects various models can be created and a range of notations can be used to describe and communicate the models such as the Unifi ed Modeling Language (UML) But in this section for simplicity a boxes-and- arrows notation has been used followed by more formal modeling languages representations in the following sections of this document In effect what has been attempted here is to produce a conceptual model of the system in a casual manner
23 The Critical Stakeholders and Actors
The critical stakeholders in the Healthcare delivery system in the United States include the providers the government the payers the patients and the suppliers In the following sections each of these categories of stakeholders is described in more detail
Fig 21 The healthcare delivery system
NA Behkami and TU Daim
11
231 Care Providers
The term Provider is used to refer to the source of care that provides treatment to patients It is important to differentiate between the two instantiations of the Provider one as an Individual and another as an Organization The individual Provider is for example the Physician Nurse or someone with similar medical training that provides often one-on-one care to the patient The organization type of Provider is the clinic or hospital which is the business unit housing the physician or nurse whom provide the care
2311 Physicians Nurses and Medical Assistants
Physicians are individuals who through training experience and certifi cation are allowed to provide care to patients with a variety of illnesses A Physician can be a general practitioner such a primary care physician or a specialist Typically physi-cians are employed by a hospital or clinic Nurses similar to Physicians have been through healthcare education and often under physician supervision (and at times independently) are expected to provide care to patients Medical Assistants (MA) typically poses job-specifi c training mainly to assist physicians and nurses with routine and less education dependent activities of providing care around the clinic
During daily operations physicians nurses and MAs are typically consumers of various forms of technology-based tools and they have been subjects of various research studies (Dorr Wilcox Donnelly Burns amp Clayton 2005 Dorr et al 2006 Eden 2002 Eley Soar Buikstra Fallon amp Hegney 2009 Ford McAlearney Phillips Menachemi amp Rudolph 2008 Jha et al 2007 May et al 2001 Simpson 2007 Wilcox et al 2007 ) Research has shown that each of these types of individ-ual provides based on attributes of their work place andor their own personal char-acteristics experience various levels of technology use Their use of technology can range from simply using electronic mail or calendars to sophisticated usages such as design patient selection algorithms from EHR data
Studying this type of stakeholder is critical since they are the daily users of tech-nology and can have a profound effect on adoption of HIT Innovations They can also often act as the champion or decision makers when it comes to adopting an innovation in their clinics or hospitals As shown in Fig 21 the providers provide care to patients are employed in the clinic provide feedback to the IT vendors they use products from adopt amp use HIT innovations and collaborate with other provid-ers for providing care
2312 The Hospital or Clinic
The hospital or clinic is where patients would receive care and they are type of a provider This type of provider can range from a single physician clinic in a rural community to a large multi-system hospital in a large city Research has shown that
2 Background Literature on the Adoption of Health Information Technologies
12
these two types of providers operate drastically different from one another and when it comes to adoption of HIT they have different needs barriers and facilitators(David 1993 Fonkych 2006 Hikmet Bhattacherjee Menachemi Kayhan amp Brooks 2008 May et al 2001 Menachemi 2007 Menachemi Brooks amp Simpson 2007 Menachemi Burke amp Brooks 2004 ) In general hospitals can have various attributes that distinguishes how they participate in the healthcare dev-ilry ecosystem for example affi liation tax status number of beds technology usage culture location and more
It is important to study this type of Provider separate from the individual Provider such as a physician since their priorities are organizational where physicians are individual contributors For example a physician may feel that using an EHR at any price is justifi ed while the priorities and budget conditions of the hospital may not allow for that (Katsma Spil Light amp Wassenaar 2007 Lobach Detmer amp Supplement 2007 ) As shown in Fig 21 the hospital employeersquos physicians pays the HIT vendor for products and adopts innovations
232 Government
The role of government in the health delivery system of the United States is enor-mous (Aalbers van der Heijden Potters van Soest amp Vollebergh 2009 Bower 2005 Cherry 2006 ) Government plays this role in two ways (1) payer (meaning providing insurance through Medicaid and Medicare ( U S Department of Health Human Services Centers for Medicare Medicaid Services ) for the low income and elderly) (2) policy setter and enforcer (Rosenfeld Bernasek amp Mendelson 2005 ) As a payer the government expenditure for providing insurance through Medicare alone reached $440 billion in 2007( Centers for Medicare Medicaid Services National Health Expenditure Data ) Such volume of business makes the government have an active interest in cost reduction through adoption of HIT ( HealthIT hhs gov ) As a policy setter especially under the current Obama administration through the American Recover Act (HR 1 American recovery and reinvestment act of 2009 ) the government of the United States has taken the driver seat to implement Healthcare reform Government hopes that much of this improved in care and reduction in cost will be realized through meaningful use of HIT ( Assistant Secretary for Public Affairs ) and faster and wider spread of technology adoption
Research that have reviewed the role of government have found that it can posi-tively infl uence and sometimes accelerate more effective HIT adoption (Fonkych 2006 ) It is important to note that in the United State with a decentralized health system the government infl uences the ecosystem both at the federal level and at the regionalstate levels Hence when modeling the system it is critical to consider the multiple perspectives As shown in Fig 21 the government pays providers infl u-ences adoption decisions of providers infl uences the physicians in general invests in support agencies and encourages nationwide standards
NA Behkami and TU Daim
13
233 Patients and Their Family and Care Givers
The patient is one of the most critical actors in the healthcare delivery system Patients once ill seek care through providers In 2006 Americans made a total of 902 million healthcare visits and 49 were with primary care physicians (Ambulatory medical care utilization estimates for 2006 ) Family or other care givers are one of the main support networks for the patient Research fi nds that patients with family or a network are more likely to recover As active participants in the care process patients and their familycaregivers can be a large infl uencer for HIT adoption by their providers or even use HIT themselves ( Ash 1997 Dorr et al 2005 Hersh 2004 Leonard 2004 May et al 2001 Robeznieks 2005a ) The patient family also uses HIT by using Personal Health Records (PHR) (Tang Ash Bates Overhage amp Sands 2006 ) As shown in Fig 21 this stakeholder pays pro-viders for service seeks care from physicians can provide feedback to HIT ven-dors cares for patients and use HIT innovations
234 Payers
The payers are the stakeholders who pay for the care that the patients receive They fall in the three categories of the government private insurance and the patients themselves In 2006 43 million Americans were enrolled in Medicare and 53 mil-lion enrolled in Medicaid ( Centers for Medicare Medicaid Services National Health Expenditure Data ) Medicare is an insurance program administered by the United States government providing health insurance to people aged 65 and over or indi-viduals with disabilities Similarly Medicaid provides insurance for low income families ( U S Department of Health Human Services Centers for Medicare Medicaid Services )
By having Private health coverage people can protect themselves from fi nical cost and guaranteed to have access to health care when needed (Claxton 2002 ) In order to make private healthcare affordable to individual citizens payers pool the risk of healthcare cost across large number of people This affords individuals (usu-ally through their employers) to pay a premium that is equal to the average cost of medical care for the group of people It is this spreading of the risk that makes healthcare affordable to most people in the society
Public sources of healthcare coverage include Medicare Medicaid federal and state employee health plans the military and the Veterans Administration Private health coverage is primarily through employee sponsored benefi t plans Private Citizen can also obtain individual health insurance from the free market in 2002 about 12 million nonelderly people purchased health insurance on their own (Claxton 2002 ) Examples of health insurance coverage include commercial health insurers Blue Cross and Blue Shield plans Health Maintenance Organizations (HMOs) Self-Funded Employee Health Benefi t Plans
2 Background Literature on the Adoption of Health Information Technologies
14
With such numbers and revenue it is not surprising that Payers exercise a lot of power and leverage in the healthcare delivery system In fact the change agents in care delivery are often the demands of the payers instead those of the patients (Healthcare payers and providers Vital signs for software development 2004 ) Effectively payers are able to manipulate providers through such mechanisms as co-payments and negotiated rates for procedures It is this infl uence from payers that is pushing hospitals to invest in Health IT For example in order to deliver care more effi ciently integrating their various isolated repositories of patient data is a priority for the payers Providers fear that this push for investment in HIT can erode their already thin revenues However it is believed that if the providers are able show effective use of IT through meaningful usage Payers would be willing to compensate for infrastructure investment through future contract negations that would be more favorable and provide more revenue for the providers (Healthcare payers and providers Vital signs for software development 2004 )
235 HITInnovation Suppliers
In context of the proposed research Suppliers are either the entities that build sup-port or service the HIT innovation that are used by the providers and the patients and sometimes paid for by the payers for the purpose of delivering patient care For example the General Electric Corporation is the vendor that builds one of the most popular EHR on the market and in this case is considered a Supplier in the ecosystem Another type of Supplier is government organizations that support HIT use for pro-viders such as a Regional Health Information Organization (RHIO) discussed below
2351 HIT Vendors
HIT vendors develop and offer technical services for a variety of HIT applications such as Health Records e-prescribing and others Vendors typically specialize in serving certain size physician practices Their products are often licensed by physi-cian or user They charge maintenance and support fees and usually charge for prod-uct upgrades They offer some limited service policies and guarantees
In case of products such as Electronic Health Records (EHR) a vendorrsquos product may be certifi ed for interoperability through the Certifi cation Commission for Health Information Technology (CCHIT) (Certifi edreg 2011 ) The vendors often charge for their products to interface with other products or sources of information at the adopting hospital In some case third-party modules or components are bun-dled with a product and the customer may need to pay for them separately Implementation and training services add to the adoption cost Since usually adop-tion requires a large investment from the provider a healthy relationship is desired
NA Behkami and TU Daim
15
with the vendors As shown in Fig 21 vendors receive feedback from providers and patients and try to stay competitive in the market place
2352 Regional Health Information Organizations
According to the defi nition from National Alliance for Health Information Technology a Regional Health Information Organization (RHIO) or also referred to as Health Information Exchange (HIE) is ldquoA health information organization [HIO] that brings together health care stakeholders within a defi ned geographic area and governs health information exchange [HIE] among them for the purpose of improv-ing health and care in that communityrdquo ( NAHIT releases HIT defi nitions News Healthcare Informatics ) RHIOs are the fundamental building blocks of the pro-posed National Health Information Network (NHIN) initiative presented by the Offi ce of the National Coordinator for Health Information Technology (ONCHIT) It is understood that to build an interoperable national health record network a strat-egy that initiates from the local and state levels is critical
HIE will focus on the areas of technology interoperability standards utilization and business information systems The goal of HIE is to make possible access to clinical data in an effective and timely manner Another goal of the HIEs will be to make available secondary data through implementation of infrastructure to be used for purposes of public health and consumer health research
24 Attributes of the Stakeholders
The Stakeholders described in the previous sections each have multiple attributes For example an attribute of the Hospital as a stakeholder maybe its affi liation is it affi liated with an academic university or is it purely for profi t organization These attributes determine how a stakeholder participates and infl uences the healthcare delivery ecosystem Table 21 summarizes the critical attributes associated with each healthcare system stakeholder extracted from research literature
25 Important Factors Effecting Diffusion and Adoption for HIT
While stakeholders and their attributes determine some of the characteristics of the healthcare delivery system there are other factors that also infl uence the ecosystem The categories of these factors include Barriers amp Infl uences theories amp methodologies policy making ecosystem characteristics adopter attitudes market competition inno-vation champions clinic workfl ow timing modeling infusion and social structures
2 Background Literature on the Adoption of Health Information Technologies
16
Table 21 Stakeholders and attributes
Stakeholder Attribute(s)
Providersphysiciansnurses bull Attitudes toward technology bull Education bull Age bull Comfort with computers bull Leadership style bull Personality bull Workload and productivity bull Stage in career bull Previous experience with adoption bull Specialization bull Role in team bull Continuing education
ProvidersHospital bull Payer mix bull IT concentration bull Patient demographics bull Geography bull Affi liation (academic or other) bull IT operations bull Budget availability bull Type of care provided bull Size bull Affl uence of customer base bull Decision making processes bull Tax status bull Partnerships bull Previous adoption experience bull Org structure style
Government bull Standards bull Regulation bull Education bull Government assistance bull Reimbursement bull Financial incentives
Patient and family bull Quality of care bull Biographic data bull Size of support network bull Education bull Experience with technology bull Extent of illness bull Family and marital status bull Age bull Attitudes towards technology
Payers bull Patient demographics bull Type (public private) bull Executive team bull Mix of patients
(continued)
NA Behkami and TU Daim
17
251 Barriers and Infl uences
Evaluating facilitators and barriers to adoption of electronic health records in long- term care facilities reviled the following barriers costs training implementation processes and compatibility with existing systems (Cherry 2006 ) Physicians EHR adoption patterns show those practicing in large groups in hospitals or medical centers and in the western region of the United States were more likely to use electronic health records (DesRoches et al 2008 ) Less likely are those hospitals that are smaller more rural non-system affi liated and in areas of low environmen-tal uncertainty (Kazley amp Ozcan 2007 ) Another study fi nds support for a positive relationship between IT concentration and likelihood of adoption (Angst 2007 ) Academic affi liation and larger IT operating capital and staff budgets are associ-ated with more highly automated clinical information systems (Amarasingham et al 2008 ) Hospital EMR adoption is signifi cantly associated with environmental uncertainty type of system affi liation size and urban-ness The effects of competi-tion munifi cence ownership teaching status public payer mix and operating mar-gin are not statistically signifi cant (Kazley amp Ozcan 2007 )
Shared electronic records are not plug-in technologies They are complex inno-vations that must be accepted by individual patients and staff and also embedded in organizational and inter-organizational routines (Greenhalgh et al 2008 ) Physicians located in counties with higher physician concentration were found to be more likely to adopt EHRs Health maintenance organization penetration rate and poverty level were not found to be signifi cantly related to EHR adoption However practice size years in practice Medicare payer mix and measures of technology readiness were found to independently infl uence physician adoption (Abdolrasulnia et al 2008 ) Organizational variables of ldquodecision makingrdquo and ldquoplanningrdquo have signifi cant impacts and successfully encouraging usage of the CPR entails attention and resources devoted to managing the organizational aspects of implementation ( Ash 1997 )
Table 21 (continued)
Stakeholder Attribute(s)
SuppliersHIT vendors bull Portfolio bull Expertise bull Cost Structure bull Marketing bull Partnerships bull Reputation bull Brand positioning
SuppliersHealth information exchange bull Standards bull Regulation bull Geography bull Cost structure
2 Background Literature on the Adoption of Health Information Technologies
18
Hospitals that place a high priority on patient safety can more easily justify the cost of Computerized Physician Order Entry (CPOE) Outside the hospital fi nan-cial incentives and public pressures encourage CPOE adoption Dissemination of data standards would accelerate the maturation of vendors and lower CPOE costs (Poon et al 2004 ) Adoption of functionalities with fi nancial benefi ts far exceeds adoption of those with safety and quality benefi ts (Poon et al 2006 ) The ideal COPE would be a system that is both customizable and integrated with other parts of the information system is implemented with maximum involvement of users and high levels of support and is surrounded by an atmosphere of trust and collabora-tion (Ash Lyman Carpenter amp Fournier 2001 )
Lack of clarity about the value of telehealth implementations is one reason cited for slow adoption of telemedicine (Cusack et al 2008 ) Others have looked at potential factors affecting telehealth adoption (Gagnon et al 2004 ) and end user online literature searching the computer-based patient record and electronic mail systems in academic health sciences centers in the United States ( Ash 1997 ) Successful diffusion of online end user literature searching is dependent on the visibility of the systems communication among rewards to and peers of possible users who promote use (champions) ( Ash 1997 ) Adoption factors on RFID deployment in healthcare applications have also been researched (Kuo amp Chen 2008 )
Technology and Administrative innovation adoption factors that have been iden-tifi ed include the job tenure cosmopolitanism educational background and organi-zational involvement of leaders (Kimberly amp Evanisko 1981 ) Hospitals that adopted a greater number of IT applications were signifi cantly more likely to have desirable quality outcomes on seven Inpatient Quality Indicator measures (Menachemi Saunders Chukmaitov Matthews amp Brooks 2007 ) Factors found to be positively correlated with PSIT (patient safety-related IT) use included physi-cians active involvement in clinical IT planning the placement of strategic impor-tance on IT by the organization CIO involvement in patient safety planning and the perception of an adequate selection of products from vendors (Menachemi Burke amp Brooks 2004 )
Patientrsquos fears about having their medical records available online is hindering not helping the push for electronic medical records Specifi c concerns include com-puter breaches and employers having access to the records(Robeznieks 2005b ) Public sector support is essential in fi ve main aspects of child health information technology namely data standards pediatric functions in health information systems privacy policies research and implementation funding and incentives for technology adoption(Conway White amp Clancy 2009 )
Financial barriers and a large number of HIT vendors offering different solu-tions present signifi cant risks to rural health care providers wanting to invest in HIT (Bahensky Jaana amp Ward 2008 ) The relative costs of the interventions or technologies compared to existing costs of care and likely levels of utilization are critical factors in selection (Davies Drummond amp Papanikolaou 2001 ) Reasons for the slow adoption of healthcare information technology include a misalign-ment of incentives limited purchasing power among providers and variability in
NA Behkami and TU Daim
19
the viability of EHR products and companies and limited demonstrated value of EHRs in practice (Middleton Hammond Brennan amp Cooper 2005 ) Community Health Centers (CHC) serving the most poor and uninsured patients are less likely to have a functional EHR CHCs cited lack of capital as the top barrier to adoption (Shields et al 2007 ) Increasing cost pressures associated with managed-care environments are driving hospitalsrsquo adoption of clinical and administrative IT systems as a means for cost reduction (Menachemi Hikmet Bhattacherjee Chukmaitov amp Brooks 2007 )
252 Tools Methods and Theories
A hospitalrsquos clinical information system requires a specifi c environment in which to fl ourish Clinical Information Technology Assessment Tool (CITAT) which mea-sures a hospitalrsquos level of automation based on physician interactions with the infor-mation system has been used to explain such environment (Amarasingham et al 2008 ) Multi-perspectives and Hazard Modeling Analysis have been used to study impact of fi rm characteristics on diffusion of Electronic Medical Records (Angst 2007 ) Elaboration Likelihood Model and Individual Persuasion model to study presence of privacy concerns in adoption of Electronic Medical Records (Angst 2007 ) Physician Order Entry (POE) adoption has been studied qualitatively using observations focus groups and interviews (Ash et al 2001 )
Other research has built conceptual models to lay out the relationships among factors affecting IT diffusion in health care organizations (Daim Tarman amp Basoglu 2008 ) Yet others have adapted diffusion of innovation (DOI) framework to the study of information systems innovations in healthcare organizations (Wainwright amp Waring 2007 ) and build a causal model to describe the development path of telemedicine internationally (Higa 1997 ) There have been attempts to extend the model of hospital innovation in order to incorporate new forms of inno-vation and new actors in the innovation process in accordance with the Schumpeterian tradition of openness (Djellal amp Gallouj 2007 ) Health innovation has been described as complex bundles of new medical technologies and clinical services emerging from a highly distributed competence base (Consoli amp Mina 2009 )
User acceptance of a Picture Archiving and Communication System has been studied through unifi ed theory of acceptance and use of technology (UTAUT) in a radiological setting (Duyck et al 2006 ) Technology Acceptance Model (TAM) and Trocchia and Jandarsquos interaction themes enabled exploring factors impacting the engagement of consumers aged 65 and older with higher forms of IT primarily PCs and the Internet (Hough amp Kobylanski 2009 ) One Electronic Medical Record (EMR) study examined the organizational and environmental correlates using a Resource Dependence Theoretical Perspective (Kazley amp Ozcan 2007 ) Since Healthcare today is mainly knowledge-based and the diffusion of medical knowl-edge is imperative for proper treatment of patients a study of the industry explored
2 Background Literature on the Adoption of Health Information Technologies
20
barriers to knowledge fl ow using a Cultural Historical Activity Theory framework (Deng amp Poole 2003 Lin Tan amp Chang 2008 )
Diffusion of innovation framework has also been used to discuss factors affect-ing adoption of telemedicine (Menachemi Burke amp Ayers 2004 Park amp Chen 2007 ) Smartphone userrsquos perceptions in a healthcare setting have been studied based on technology acceptance model (TAM) and innovation attributes (Park amp Chen 2007 ) A study of Information Technology Utilization in Mental Health Services utilization adopted two theoretical framework models from Teng and Calhounrsquos computing and communication dimensions of information technology and Hammer and Mangurianrsquos conceptual framework for applications of communi-cations technology (Saouli 2004 )
To identify factors that affect hospitals in adopting e-signature the Technology-Organization- Environment (TEO) have been adopted (Chang Hwang Hung Lin amp Yen 2007 ) An examination of factors that infl uence the healthcare profession-alsrsquo intent to adopt practice guideline innovation combined diffusion of innovation theory and the theory of planned behavior (TPB) (Granoff 2002 ) To identify the concerns of managers and supervisors for adopting a managerial innovation the Concerns-Based Adoption Model and the Stages of Concern (SoC) were utilized (Agney 1997 )
253 Policy Making
There is a gap in our knowledge on how regulatory policies and other national health systems attributes combine to impact on the utilization of innovation and health system goals and objectives A study found that strong regulation adversely affects access to innovation reduces incentives for research-based fi rms to develop innovative products and leads to short- and long-term welfare losses Concluding that policy decision makers need to adopt a holistic approach to policy making and consider potential impact of regulations on the uptake and diffusion of innovations innovation systems and health system goals (Atun Gurol-Urganci amp Sheridan 2007 ) Recommendations have been made to stimulate adoption of EHR including fi nancial incentives promotion of EHR standards enabling policy and educational marketing and supporting activities for both the provider community and health-care consumers (Blumenthal 2009 Middleton et al 2005 ) Proposed manners on how the government should assist are a reoccurring topic (Bower 2005 )
Economic issues for health policy and policy issues for economic appraisal have concluded that a wide range of mechanisms exist to infl uence the diffusion and use of health technologies and that economic appraisal is potentially applicable to a number of them (Drummond 1994 ) Other conclusions calls for greater Centers for Medicare and Medicaid Service (CMS) involvement and reimbursement models that would reward higher quality and effi ciency achieved (Fonkych 2006 ) Medicare should pay physicians for the costs of adopting IT and assume that future savings to Medicare will justify the investment The Medicare Payment Advisory Commission
NA Behkami and TU Daim
21
(MedPAC) recommended establishing a budget-neutral pay-for-performance pro-gram to reward physicians for the outcomes of use instead of simply helping them purchase a system (Hackbarth amp Milgate 2005 Menachemi Matthews Ford amp Brooks 2007 )
As the largest single US purchaser of health care services Medicare has the power to promote physician adoption of HIT The Centers for Medicare and Medicaid Services should clarify its technology objectives engage the physician community shape the development of standards and technology certifi cation crite-ria and adopt concrete payment systems to promote adoption of meaningful tech-nology that furthers the interests of Medicare benefi ciaries (Powner 2006 Rosenfeld et al 2005 )
Imminent adopters perceived EHR barriers very differently from their other colleges For example imminent adopters were signifi cantly less likely to consider upfront cost of hardwaresoftware or that an inadequate return on investment was a major barrier to EHR Policy and decision makers interested in promoting the adop-tion of EHR among physicians should focus on the needs and barriers of those most likely to adopt HER (Menachemi 2006 ) Ensuring comparable health IT capacity among providers that disproportionately serve disadvantaged patients will have increasing relevance for disparities thus monitoring adoption among such provid-ers should be a priority (Shields et al 2007 ) In the health information security arena results suggest that signifi cant non-adoption of mandated security measures continues to occur across the health-care industry (Lorence amp Churchill 2005 )
254 Hospital Characteristics and the Ecosystem
Academic affi liation and larger IT operating capital and staff budgets are associ-ated with more highly automated clinical information systems (Amarasingham et al 2008 ) Despite several initiatives by the federal government to spur this devel-opment HIT implementation has been limited particularly in the rural market (Bahensky et al 2008 ) Study of a small clinic found that the EHR implementation did not change the amount of time spent by physicians with patients On the other hand the work of clinical and offi ce staff changed signifi cantly and included decreases in time spent distributing charts transcription and other clerical tasks (Carayon Smith Hundt Kuruchittham amp Li 2009 )
Health IT adoption for medication safety indicate wide variation in health IT adoption by type of technology and geographic location Hospital size ownership teaching status system membership payer mix and accreditation status are associ-ated with health IT adoption although these relationships differ by type of technol-ogy Hospitals in states with patient safety initiatives have greater adoption rates (Furukawa Raghu Spaulding amp Vinze 2008 ) Another study examined geographic location (urban versus rural) system membership (stand-alone versus system- affi liated) and tax status (for-profi t versus non-profi t) and found that location is systematically related to HIT adoption (Hikmet Bhattacherjee Menachemi
2 Background Literature on the Adoption of Health Information Technologies
22
Kayhan amp Brooks 2008 ) Others studies have also considered hospital characteris-tics (Jha Doolan Grandt Scott amp Bates 2008 Koch amp Kim 1998 )
Although top information technology priorities are similar for all rural hospitals examined differences exist between system-affi liated and stand-alone hospitals in adoption of specifi c information technology applications and with barriers to infor-mation technology adoption (Menachemi Burke Clawson amp Brooks 2005 ) Hospitals adopted an average of 113 (452 ) clinical IT applications 157 (748 ) administrative IT applications and 5 (50 ) strategic IT applications (Menachemi Chukmaitov Saunders amp Brooks 2008 )
There are concerns that psychiatry may lag behind other medical fi elds in adopt-ing information technology (IT) Psychiatristsrsquo lesser reliance on laboratory and imaging studies may explain differences in data exchange with hospitals and labs concerns about patient privacy are shared among all medical providers (Mojtabai 2007 ) Some innovations in health information technology for adult populations can be transferred to or adapted for children but there also are unique needs in the pedi-atric population (Conway et al 2009 )
255 Adopter Attitudes Perceptions and Characteristics
Studies have been conducted on perceptions and attitudes of healthcare profession-als towards telemedicine technology (Al-Qirim 2007a ) A diffusion study of a community-based learning venue demonstrated that about half of this senior popu-lation was interested in using the Internet as a tool to fi nd credible health informa-tion (Cortner 2006 ) Societal trends are transforming older adults into lead adopters of a new 247 lifestyle of being monitored managed and at times motivated to maintain their health and wellness A study of older adults perception of Smart Home Technologies uncovered support of technological advance along with a vari-ety of concerns that included usability reliability trust privacy stigma accessibil-ity and affordability (Coughlin DrsquoAmbrosio Reimer amp Pratt 2007 ) Factors impacting the engagement of healthcare consumers aged 65 and older with higher forms of IT primarily PCs and the Internet have been examined (Hough amp Kobylanski 2009 )
Principal uses for the Information Technology by the nurses are for access to patientsrsquo records and for internal communication However not all aspects of computer introduction to nursing are positive (Eley et al 2009 ) Physicians who cared for large minority populations had comparable rates of EHR use identifi ed similar barriers and reported similar benefi ts (Jha et al 2007 ) Patients have a role in designing Health Information Systems (Leonard 2004 ) and consideration of patient values and preferences in making clinical decisions is essential to deliver the highest quality of care (Melnyk amp Fineout-Overholt 2006 ) Patient characteristics of hospi-tals are related to the adoption of health IT has been under studied Once study pro-posed that children when hospitalized are more likely to seek care in technologically
NA Behkami and TU Daim
23
and clinically advanced facilities However it is unclear whether the IT adopted is calibrated for optimal pediatric use (Menachemi Brooks amp Simpson 2007 )
256 Strategic Management and Competitive Advantage
The diffusion of health care technology is infl uenced by both the total market share of care organizations as well as the level of competition among them Results show that a hospital is less likely to adopt the technology if Healthcare Maintenance Organization (HMO) market penetration increases but more likely to adopt if HMO competition increases (Bokhari 2009 ) Increasing cost pressures associated with managed-care environments are driving hospitalsrsquo adoption of clinical and adminis-trative IT systems as such adoption is expected to improve hospital effi ciency and lower costs (Menachemi Hikmet et al 2007 )
Deployment of health information technology (IT) is necessary but not suffi -cient for transforming US health care The strategic impact of information tech-nology convergence on healthcare delivery and support organizations have been studied (Blumberg amp Snyder 2001 ) Four focus areas for application of strategic management have been identifi ed adoption governance privacy and security and interoperability (Kolodner Cohn amp Friedman 2008 ) While another found little that strategic behavior or hospital competition affects IS adoption (McCullough 2008 )
A study looking at strategic behavior of EHR adopters found that the relevance of EHR merely focuses on the availability of information at any time and any place This implementation of relevance does not meet end-usersrsquo expectations and is insuffi cient to accomplish the aspired improvements In addition the used participa-tion approaches do not facilitate diffusion of EHR in hospitals (Katsma Spil Ligt amp Wassenaar 2007 )
257 Innovation Champions and Their Aids
There is a need for the tight coupling between the roles of both the administrative and the clinical managers in healthcare organizations in order to champion adoption and diffusion and to overcome many of the barriers that could hinder telemedicine success (Al-Qirim 2007b ) Survey of chief information offi cers (CIOs) the indi-viduals who manage HIT adoption effort suggests that the CIO position and their responsibilities varies signifi cantly according to the profi t status of the hospital (Burke Menachemi amp Brooks 2006 )
Acting as aids to change-agents in healthcare settings Clinical engineers can identify new medical equipment review their institutionrsquos technological posi-tion develop equipment-selection criteria supervise installations and monitor
2 Background Literature on the Adoption of Health Information Technologies
24
post- procurement performance to meet their hospitalrsquos programrsquos objectives The clinical engineerrsquos skills and expertise are needed to facilitate the adoption of an innovation (David 1993 ) However Information technology implementation is a political process and in the increasingly cost-controlled high-tech healthcare environment a successful nursing system implementation demands a nurse leader with both political savvy and technological competency (Simpson 2000 ) One study found that prior user testimony had a positive effect on new adaptors (Eden 2002 )
258 Workfl ow and Knowledge Management
Successful adoption of health IT requires an understanding of how clinical tasks and workfl ows will be affected yet this has not been well described Understanding the clinical context is a necessary precursor to successful deployment of health IT (Leu et al 2008 ) Healthcare today is mainly knowledge-based and the diffu-sion of medical knowledge is imperative for proper treatment of patients (Lin et al 2008 ) For example researchers must determine how to take full advantage of the potential to create and disseminate new knowledge that is possible as a result of the data that are captured by EHR and accumulated as a result of EHR diffusion (Lobach amp Detmer 2007 ) Findings suggest that some small practices are able to overcome the substantial learning barriers presented by EMRs but that others will require support to develop suffi cient learning capacity (Reardon amp Davidson 2007 )
259 Timing and Sustainability
Determining the right time for adoption and the appropriate methods for calculating the return on investment are not trivial (Kaufman Joshi amp OrsquoDonnell 2009 ) Among the practices without an EHR 13 plan to implement one within the next 12 months 24 within the next 1ndash2 years 11 within the next 3ndash5 years and 52 reported having no plans to implement an EHR in the foreseeable future (Simpson 2000 ) The relationship between the timing of adoption of a technologi-cal innovation and hospital characteristics have been explored (Poulsen et al 2001 )
Key factors that infl uence sustainability in the diffusion of the Hospital Elder Life Program (HELP) are Staff experiences sustaining the program recognizing the need for sustained clinical leadership and funding as well as the inevitable modifi cations required to sustain innovative programs can promote more-realist (Bradley Webster Baker Schlesinger amp Inouye 2005 )
NA Behkami and TU Daim
25
2510 Modeling and Forecasting
The future diffusion rate of CPOE systems in US hospitals is empirically predicted and three future CPOE adoption scenarios-ldquoOptimisticrdquo ldquoBest estimaterdquo and ldquoConservativerdquo developed Two of the CPOE adoption scenarios have diffusion S-curve that indicates a technology will achieve signifi cant market penetration Under current conditions CPOE adoption in urban hospitals will not reach 80 penetration until 2029 (Ford et al 2008 ) Using a Bass Diffusion Model EHR adoption has been predicted Under current conditions EHR adoption will reach its maximum market share in 2024 in the small practice setting The promise of improved care quality and cost control has prompted a call for universal EHR adoption by 2014 The EHR products now available are unlikely to achieve full diffusion in a critical market segment within the time frame being targeted by policy makers (Ford Menachemi amp Phillips 2006 ) Others have attempted to model healthcare technology adoption patterns (Carrier Huguenor Sener Wu amp Patek 2008 )
2511 Infusion
Innovation attributes are important predictors for both the spread of usage (internal diffusion) and depth of usage (infusion) of electronic mail in a healthcare setting (Ash amp Goslin 1997 ) In a study two dependent variables internal diffusion (spread of diffusion) and infusion (depth of diffusion) were measured Little correlation between them was found indicating they measured different things (Ash 1999 ) Study of organizational factors which infl uence the diffusion of end user online lit-erature searching the computer-based patient record and electronic mail systems in academic health sciences centers found that Organizational attributes are important predictors for diffusion of information technology innovations Individual variables differ in their effect on each innovation The set of attributes seems less able to pre-dict infusion ( Ash 1997 )
2512 Social Structure and Communication Channels
Resisting and promoting new technologies in clinical practice face a fundamental problem of the extent to which the telecommunications system threatened deeply embedded professional constructs about the nature and practice of care giving rela-tionships (May et al 2001 ) Researchers have also attempted to understand how and why patient and consumer organizations use Health Technology Assessment
2 Background Literature on the Adoption of Health Information Technologies
26
(HTA) fi ndings within their organizations and what factors infl uence how and when they communicate their fi ndings to members or other organizations (Fattal amp Lehoux 2008 )
26 The Need for Multiple Perspectives in Research
In his book ldquoUsing Multiple Perspective to improve performancerdquo Linstone states that the approach of looking at the problem from multiple perspectives will enable ldquo viewing complex systems and decision about them from different perspectives each providing insights not attainable with the others rdquo (Linstone 1999 ) Due to the ever growing complexity of systems many researchers and practitioners have advo-cated the need for viewing building and analyzing systems (especially those used by humans and the society) from multiple views Two methods that are pertinent to the HIT diffusion research being proposed here are Linstonersquos Multiple Perspectives Methodology and the ldquo4 + 1 viewrdquo model originated by Philippe Kruchten ( 1995 ) and popularized in Software Engineering and Software Architecture Domains The next two sections discuss these to methodologies in detail
27 Linstonersquos Multiple Perspectives Method
There are three perspectives that are part of Linstonersquos Multiple Perspectives meth-odology Technical (T) Organizational (O) and Personal (P) (Linstone 1999 )
In the T perspective the technology and its environment are viewed as a system The T perspective is a rational approach to viewing the problem and it represents a quantitative approach to viewing the world in terms of for example alternatives trade-offs optimization data and models (Linstone 1999 )
The O perspective is concerned with less technical matters and more what affects organizations can have The O perspective also describes the culture that has helped form and connects the organization or a society For example an example of an item from this view could be fear of staff in a company about making errors in their work The O perspective can help by identifying pressures on the technology insights into societal abilities to absorb a technology and increase abilities to facili-tate organizationrsquos support for technology
According to Linestone the P perspective can be the hardest view to defi ne and should include any matters relating to individuals that are not included in other views In general the P perspective helps us better understand the O perspective Individuals matter and they can sometimes bring changes to organization with less effort than the whole institution would the P perspective identifi es their character-istic and behavior Perspectives are dynamic and change over time they also can confl ict or support each other Table 22 shows a summary of characteristics for each Linestone perspective (Linstone 1999 )
NA Behkami and TU Daim
27
Tabl
e 2
2 Su
mm
ary
of L
inst
onersquo
s m
ulti-
pers
pect
ives
cha
ract
eris
tics
(Lin
ston
e 1
999 )
Tech
nica
l (T
) O
rgan
izat
iona
l (O
) Pe
rson
al (
P)
Wor
ldvi
ew
Scie
nce-
tech
nolo
gy
Uni
que
grou
p or
inst
itutio
nal v
iew
In
divi
dual
the
sel
f O
bjec
tive
Prob
lem
sol
ving
pro
duct
A
ctio
n p
roce
ss s
tabi
lity
Pow
er i
nfl u
ence
pre
stig
e Sy
stem
foc
us
Art
ifi ci
al c
onst
ruct
So
cial
G
enet
ic p
sych
olog
ical
M
ode
of in
quir
y O
bser
vatio
n a
naly
sis
dat
a an
d m
odel
s C
onse
nsua
l ad
vers
ary
bar
gain
ing
and
com
prom
ise
Intu
ition
lea
rnin
g e
xper
ienc
e
Eth
ical
bas
is
Log
ical
rat
iona
lity
Just
ice
fai
rnes
s M
oral
ity p
erso
nal e
thic
s Pl
anni
ng h
oriz
on
Far
(low
dis
coun
ting)
In
term
edia
te (
mod
erat
e di
scou
ntin
g)
Shor
t for
mos
t (hi
gh d
isco
untin
g)
Oth
er d
escr
ipto
rs
Cau
se a
nd e
ffec
t O
ptim
izat
ion
Qua
ntifi
catio
n tr
ade-
offs
cos
t-be
nefi t
ana
lysi
s Pr
obab
ilitie
s a
vera
ges
sta
tistic
s
expe
cted
val
ue
Prob
lem
sim
plifi
ed a
nd id
ealiz
ed
redu
ctio
nism
N
eed
valid
atio
n r
eplic
abili
ty
Con
cept
ualiz
atio
n s
yste
ms
theo
ries
U
ncer
tain
ties
note
d
Age
nda
(pro
blem
of
the
mom
ent)
Sa
tisfy
ing
Incr
emen
tal c
hang
e R
elia
nce
on e
xper
ts i
nter
nal t
rain
ing
of
prac
titio
ners
Pr
oble
m d
eleg
ated
fac
tore
d is
sues
and
cr
isis
man
agem
ent
Nee
d st
anda
rd o
pera
ting
proc
edur
es
reut
iliza
tion
Rea
sona
blen
ess
Unc
erta
inty
use
d fo
r or
gani
zatio
nal
self
-pre
serv
atio
n
Cha
lleng
e an
d re
spon
se l
eade
rs a
nd
follo
wer
s A
bilit
y to
cop
e w
ith o
nly
a fe
w a
ltern
ativ
es
Fear
of
chan
ge
Nee
d fo
r be
liefs
illu
sion
s m
ispe
rcep
tion
of
prob
abili
ties
Hie
rarc
hy o
f in
divi
dual
nee
ds (
surv
ival
hellip)
Nee
d to
fi lte
r ou
t inc
onsi
sten
t im
ages
C
reat
ivity
vis
ion
by th
e fe
w i
mpr
ovis
atio
n N
eed
for
cert
aint
y
Cri
teri
a fo
r ldquoa
ccep
tabl
e ri
skrdquo
Log
ical
sou
ndne
ss o
penn
ess
to
eval
uatio
n d
ecis
ion
anal
ysis
In
stitu
tiona
l com
patib
ility
pol
itica
l ac
cept
abili
ty p
ract
ical
ity
Con
duci
vene
ss to
lear
ning
foc
us o
n ldquom
e-no
wrdquo
Com
mun
icat
ions
Te
chni
cal r
epor
t br
iefi n
g In
side
r la
ngua
ge o
utsi
ders
rsquo as
sum
ptio
ns
ofte
n m
ispe
rcei
ved
Pers
onal
ity a
nd c
hari
sma
desi
rabl
e
2 Background Literature on the Adoption of Health Information Technologies
28
When using the perspectives to build a real-world model or make a decision so called the ldquo Ultimate decision rdquo by Linstone all inputs from various perspectives should to be integrated The process of integration is never simply adding the infor-mation up from various perspectives The perspectives have to fi t each other some-times reinforcing each other or canceling each other out (Linstone 1999 Linstone Mitroff amp Hoos )
28 The ldquo4 + 1 Viewrdquo Model for Software Architectures
Numerous sources emphasis the importance of modeling business processes and the relevant ecosystems however there seems to be a lack of guidance on how to best capture these architectures Documenting a model is an important sub-disciple of software engineering Architecture allows us to concentrate on the components and relationship at a relevant yet manageable level Dividing a complex problem into parts allows groups to participate in solving a problem In general documenting systems serves three important purposes as a means of education by using it to introduce people to the system a tool for communication between stakeholders and provides appropriate information for analysis
A view represents elements and relationships amongst them within a system When documenting a model a view highlights dimensions of the system architecture while hiding other details Various authors have recommended specifi c views that should be employed when documenting software architectures including Zachman Framework ( The Zachman Framework ) Reference Model for Open Distributed Processing (RM-ODP) ( Reference model of open distributed processing Wiki ) Department of Defense Architecture Framework (DoDAF) ( DoDAF Architecture Framework Version 2 0 ) Federal Enterprise Architecture ( Federal Enterprise Architecture ) and Nominal Set of Views (ANSIIEEE 1471 ) In particular ldquo4 + 1rdquo approach to architecture by Philippe Kruchten of the Rational Corporation (Kruchten 1995 ) has been infl uential used in system building it uses four views (Logical Process Development and Physical) with a fi fth view (Scenarios) that ties the other four together While these are benefi cial views they may not be useful in every system and the ultimate purpose is to separate concerns and document the model for a variety of stakeholders (Bachmann
et al 2001 )
29 Categorization of Important Factors in HIT Adoption Using Multi-perspectives
Recall that Linstonersquos multi-perspectives methodology uses the Technical Perspective (T) Organizational Perspective (O) and the Personal Perspective (P) In Sect 25 infl uencing factors within the healthcare delivery ecosystem were iden-tifi ed In this section using an iterative thematic analysis method the important
NA Behkami and TU Daim
29
factors have been group into T-O-P perspectives showing how the various factors relating to HIT Diffusion can fi t into views and the proposed research
Consistent with Linstone methodology if a factor was related to technology and its focus was an artifi cial construct it was placed under the T column If the factor was from an institutional view and its system focus was social it was placed under O column If the factor was related to an individual or self with a psychological focus it was placed in the P column Table 23 shows the combinations of stakehold-ers and perspectives being considered in this research Table 24 lists each factor in
Table 23 Userperspective matrix
Perspectives
Technical perspectives (T)
Organizational perspective (O)
Personal perspective (P)
Stakeholders Patient X X X Provider X X X Payer X X X Government X X X
Table 24 Classifi cation of HIT diffusion factors by Linstone T-O-P perspectives
Technical perspective (T) Organizational perspective (O) Personal perspective (P)
Increase quality of care Reduce cost Patient family Increase accessibility of care Increase productivity Adoption decision Quality metrics Environment Patient satisfaction HIT innovations Value chain Provider attitude towards Adoption rate Patient coordination Adoption Adoption timeline Adoption decision Provider education Diffusion Adoption attitudes Social structure Meaningful HIT use Adoption barriers and challenges Support network Reimbursement Facilitators Comfort with using
technology Payer model IT decision makers Communication channels Payer mix Financial decision maker Staff roles Demographics Affi liation Staff Education Lock in cost Tax status Support cost Minority population status Standards Social structure Social system Communication channels Social structure Information activities Communication channels Diffusion activities Size
Public opinion IT operations Budget availability
2 Background Literature on the Adoption of Health Information Technologies
30
its relevant T-O-P perspective column at this time they are combined for all the stakeholders in the future factors can be separated by stakeholder
References
Aalbers R van der Heijden E Potters J van Soest D amp Vollebergh H (2009) Technology adoption subsidies An experiment with managers Energy Economics 31 431ndash442
Abdolrasulnia M Menachemi N Shewchuk R M Ginter P M Duncan W J amp Brooks R G (2008) Market effects on electronic health record adoption by physicians Health Care Management Review 33 243
Agney M (1997) Managersrsquo and supervisorsrsquo stages of concern regarding adoption of Total Quality ManagementContinuous Quality Improvement as an organizational innovation in a medical center hospital Dissertation Abstracts International Section A Humanities and Social Sciences
Al-Qirim N (2007a) Realizing telemedicine advantages at the national level Cases from the United Arab Emirates Telemedicine and e-Health 13 545ndash556
Al-Qirim N (2007b) Championing telemedicine adoption and utilization in healthcare organiza-tions in New Zealand International Journal of Medical Informatics 76 42ndash54
Amarasingham R Diener-West M Plantinga L Cunningham A C Gaskin D J amp Powe N R (2008) Hospital characteristics associated with highly automated and usable clinical information systems in Texas United States BMC Medical Informatics and Decision Making 8 39
Ambulatory medical care utilization estimates for 2006 (Center for Disease Control and Prevention)
Angst C (2007) Information technology and its transformational effect on the health care indus-try Dissertation Abstracts International Section A Humanities and Social Sciences
ANSIIEEE Standard 1471ISOIEC 42010 (Recommended Practice for Architectural Description of Software-Intensive Systems)
Ash J (1997) Organizational factors that infl uence information technology diffusion in academic health sciences centers Journal of the American Medical Informatics Association 4 102ndash109
Ash J S (1997) Factors affecting the diffusion of the computer-based patient record Proceedings of the AMIA Annual Fall Symposium 682ndash686
Ash J S (1999) Factors affecting the diffusion of online end user literature searching Bulletin of the Medical Library Association 87 58
Ash J amp Goslin L (1997) Factors affecting information technology transfer and innovation dif-fusion in health care Innovation in technology managementmdashthe key to global leadership PICMET rsquo97 Portland International Conference on Management and Technology (pp 751ndash754)
Ash J S Lyman J Carpenter J amp Fournier L (2001) A diffusion of innovations model of physician order entry Proceedings of the AMIA Symposium 22
Assistant Secretary for Public Affairs Process begins to defi ne ldquomeaningful userdquo of electronic health records
Atun R A Gurol-Urganci I amp Sheridan D (2007) Uptake and diffusion of pharmaceutical innovations in health systems Innovation in the Biopharmaceutical Industry 85
Bachmann F Bass L Clements P Garlan D Ivers J Little R et al (2001) Documenting software architectures Organization of documentation package Pittsburgh PA Software Engineering Institute
NA Behkami and TU Daim
31
Bahensky J A Jaana M amp Ward M M (2008) Health care information technology in rural America Electronic medical record adoption status in meeting the national agenda The Journal of Rural Health 24 101ndash105
Blumberg M R amp Snyder R L (2001) The strategic impact of information technology conver-gence on healthcare delivery and support organizations Biomedical Instrumentation and Technology 35 177ndash187
Blumenthal D (2009) Stimulating the adoption of health information technology New England Journal of Medicine 360 1477
Bokhari F A (2009) Managed care competition and the adoption of hospital technology The case of cardiac catheterization International Journal of Industrial Organization 27 223ndash237
Bower A G (2005) The diffusion and value of healthcare information technology Santa Monica CA Rand Corporation
Bradley E H Webster T R Baker D Schlesinger M amp Inouye S K (2005) After adoption Sustaining the innovation A case study of disseminating the hospital elder life program Journal of the American Geriatrics Society 53 1455ndash1461
Burke D Menachemi N amp Brooks R (2006) Health care CIOs Assessing their fi t in the orga-nizational hierarchy and their infl uence on information technology capability The Health Care Manager 25 167
Carayon P Smith P Hundt A S Kuruchittham V amp Li Q (2009) Implementation of an electronic health records system in a small clinic The viewpoint of clinic staff Behaviour and Information Technology 28 5ndash20
Carrier J M Huguenor T W Sener O Wu T J amp Patek S D (2008) Modeling the adoption patterns of new healthcare technology with respect to continuous glucose monitoring IEEE Systems and Information Engineering Design Symposium 2008 SIEDS 2008 (pp 249ndash254)
Centers for Medicare amp Medicaid Services National Health Expenditure Data CCHIT Certifi ed reg 2011 products|CCHIT Chang I Hwang H Hung M Lin M amp Yen D C (2007) Factors affecting the adoption of
electronic signature Executivesrsquo perspective of hospital information department Decision Support Systems 44 350ndash359
Chaudhry B Wang J Wu S Maglione M Mojica W Roth E et al (2006) Systematic review Impact of health information technology on quality effi ciency and costs of medical care Annals of Internal Medicine 144 742ndash752
Cherry B (2006) Determining facilitators and barriers to adoption of electronic health records in long-term care facilities UMI Dissertation Services ProQuest Information and Learning Ann Arbor MI
Claxton G (2002) How private insurance works A primer The Kaiser Family Foundation Consoli D amp Mina A (2009) An evolutionary perspective on health innovation systems Journal
of Evolutionary Economics 19 297ndash319 Conway P H White P J amp Clancy C (2009) The public role in promoting child health infor-
mation technology Pediatrics 123 S125 Cortner D M (2006) Stages of Internet adoption in preventive health An exploratory diffusion
study of a community-based learning venue for 50+ year-old adults Ann Arbor 1001 Coughlin J DrsquoAmbrosio L A Reimer B amp Pratt M R (2007) Older adult perceptions of
smart home technologies Implications for research policy amp market innovations in healthcare Proceedings of IEEE Engineering in Medicine and Biology Society 2007 1810ndash1815
Cusack C M Pan E Hook J M Vincent A Kaelber D C amp Middleton B (2008) The value proposition in the widespread use of telehealth Journal of Telemedicine and Telecare 14 167
Daim T U Tarman R T amp Basoglu N (2008) Exploring barriers to innovation diffusion in health care service organizations An issue for effective integration of service architecture and information technologies In Hawaii International Conference on System Sciences (p 100) Los Alamitos CA IEEE Computer Society
2 Background Literature on the Adoption of Health Information Technologies
32
David Y (1993) Technology evaluation in a US hospital The role of clinical engineering Medical and Biological Engineering and Computing 31 HTA28ndashHTA32
Davies L Drummond M amp Papanikolaou P (2001) Prioritizing investments in health technol-ogy assessment International Journal of Technology Assessment in Health Care 16 73ndash91
Deng L amp Poole M S (2003) Learning through telemedicine networks In Proceedings of the 36th Annual Hawaii International Conference on System Sciences ( HICSSrsquo03 )mdash Track 6mdashVolume 6 (p 1741) IEEE Computer Society
DesRoches C M Campbell E G Rao S R Donelan K Ferris T G Jha A et al (2008) Electronic health records in ambulatory caremdasha national survey of physicians The New England Journal of Medicine 359 50
Djellal F amp Gallouj F (2007) Innovation in hospitals A survey of the literature The European Journal of Health Economics 8 181ndash193
DoDAF Architecture Framework Version 20 Dorr D Wilcox A Burns L Brunker C Narus S amp Clayton P (2006) Implementing a
multidisease chronic care model in primary care using people and technology Disease Management 9 (1) 1ndash15
Dorr D A Wilcox A Donnelly S M Burns L amp Clayton P D (2005) Impact of generalist care managers on patients with diabetes Health Services Research 40 1400ndash1421
Drummond M (1994) Evaluation of health technology Economic issues for health policy and policy issues for economic appraisal Social Science and Medicine (1982) 38 1593
Duyck P Pynoo B Devolder P Voet T Adang L amp Vercruysse J (2006) User acceptance of a picture archiving and communication systemmdashApplying the unifi ed theory of acceptance and use of technology in a radiological setting Nuklearmedizin 45 139ndash143
Eden K B (2002) Selecting information technology for physiciansrsquo practices A cross-sectional study BMC Medical Informatics and Decision Making 2 4
Eley R Soar J Buikstra E Fallon T amp Hegney D (2009) Attitudes of Australian nurses to information technology in the workplace A national survey Computers Informatics Nursing 27 114
Fattal J amp Lehoux P (2008) Health technology assessment use and dissemination by patient and consumer groups Why and how International Journal of Technology Assessment in Health Care 24 473ndash480
Federal Enterprise Architecture Fonkych K (2006) Accelerating adoption of clinical IT among the healthcare providers in United
States Strategies and policies The Pardee Rand Graduate School Ford E W McAlearney A S Phillips M T Menachemi N amp Rudolph B (2008) Predicting
computerized physician order entry system adoption in US hospitals Can the federal mandate be met International Journal of Medical Informatics 77 539ndash545
Ford E W Menachemi N amp Phillips M T (2006) Predicting the adoption of electronic health records by physicians When will health care be paperless Journal of the American Medical Informatics Association 13 106ndash112
Furukawa M F Raghu T S Spaulding T J amp Vinze A (2008) Adoption of health informa-tion technology for medication safety in US hospitals 2006 Health Affairs 27 865
Gagnon M Lamothe L Fortin J Cloutier A Godin G Gagne C et al (2004) The impact of organizational characteristics on telehealth adoption by hospitals In System Sciences 2004 Proceedings of the 37th Annual Hawaii International Conference on 2004 (p 10)
Granoff M J (2002) An examination of factors that infl uence the healthcare professionalsrsquo intent to adopt practice guideline innovation Dissertation Abstracts International Section B The Sciences and Engineering
Greenhalgh T Stramer K Bratan T Byrne E Mohammad Y amp Russell J (2008) Introduction of shared electronic records Multi-site case study using diffusion of innovation theory British Medical Journal 337 a1786
HR 1 American recovery and reinvestment act of 2009 (GovTrackus)
NA Behkami and TU Daim
33
Hackbarth G amp Milgate K (2005) Using quality incentives to drive physician adoption of health information technology Health Affairs 24 1147ndash1149
Healthcare payers and providers Vital signs for software development 2004 HealthIThhsgov Health IT adoption Hersh W (2004) Health care information technology Progress and barriers Journal of the
American Medical Association 292 2273ndash2274 Higa K Shin B amp Au G (1997) Suggesting a diffusion model of telemedicinemdashFocus on
Hong Kongrsquos case In Hawaii International Conference on System Sciences (p 156) Los Alamitos CA IEEE Computer Society
Hikmet N Bhattacherjee A Menachemi N Kayhan V O amp Brooks R G (2008) The role of organizational factors in the adoption of healthcare information technology in Florida hos-pitals Health Care Management Science 11 1ndash9
Hough M amp Kobylanski A (2009) Increasing elder consumer interactions with information technology Journal of Consumer Marketing 26 39ndash48
Jha A K Bates D W Jenter C A Orav E J Zheng J amp Simon S R (2007) Do minority- serving physicians have comparable rates of use of electronic health records AMIA Symposium 993
Jha A K Doolan D Grandt D Scott T amp Bates D W (2008) The use of health information technology in seven nations
Katsma C P Spil T A M Light E amp Wassenaar A (2007) Implementation and use of an electronic health record Measuring relevance and participation in four hospitals
Katsma C P Spil T A Ligt E amp Wassenaar A (2007) Implementation and use of an elec-tronic health record Measuring relevance and participation in four hospitals International Journal of Healthcare Technology and Management 8 625ndash643
Kaufman M Joshi S amp OrsquoDonnell E (2009) Itrsquos all about the timing While implementing technologies throughout your hospitalrsquos supply chain has been identifi ed as an avenue of improvement determining the right time for adoption and the appropriate methods for calculat-ing the return on investment are not quite that easy Supply Chain
Kazley A S amp Ozcan Y A (2007) Organizational and environmental determinants of hospital EMR adoption A national study Journal of Medical Systems 31 375ndash384
Kimberly J R amp Evanisko M J (1981) Organizational innovation The infl uence of individual organizational and contextual factors on hospital adoption of technological and administrative innovations The Academy of Management Journal 24 689ndash713
Koch J amp Kim C (1998) Business objectives hospital characteristics and the uses of advanced information technology In Proceedings Pacifi c Medical Technology Symposium-PACMEDTek Transcending Time Distance and Structural Barriers (Cat No98EX211) Honolulu HI (pp 68ndash78)
Kolodner R M Cohn S P amp Friedman C P (2008) Health information technology Strategic initiatives real progress Health Affairs 27 w391
Kruchten P (1995) Architectural blueprintsmdashThe ldquo4+ 1rdquo view model of software architecture IEEE Software 12 42ndash50
Kuo C amp Chen H (2008) The critical issues about deploying RFID in healthcare industry by service perspective In Hawaii International Conference on System Sciences (p 111) Los Alamitos CA IEEE Computer Society
Leonard K J (2004) The role of patients in designing health information systems The case of applying simulation techniques to design an electronic patient record (EPR) interface Health Care Management Science 7 275ndash284
Leu M G Cheung M Webster T R Curry L Bradley E H Fifi eld J et al (2008) Centers speak up The clinical context for health information technology in the ambulatory care setting Journal of General Internal Medicine 23 372ndash378
Lin C Tan B amp Chang S (2008) An exploratory model of knowledge fl ow barriers within healthcare organizations Information and Management 45 331ndash339
2 Background Literature on the Adoption of Health Information Technologies
34
Linstone H A (1999) Decision making for technology executives Using multiple perspectives to improved performance BostonLondon Artech House
Linstone H A Mitroff I I amp Hoos I R R The challenge of the 21st century State University of New York Press
Lobach D F amp Detmer D E (2007) Research challenges for electronic health records American Journal of Preventive Medicine 32 104ndash111
Lobach D F Detmer D E amp Supplement (2007) Research challenges for electronic health records
Lorence D P amp Churchill R (2005) Incremental adoption of information security in health-care organizations Implications for document management IEEE Transactions on Information Technology in Biomedicine 9 169ndash173
May C Gask L Atkinson T Ellis N Mair F amp Esmail A (2001) Resisting and promoting new technologies in clinical practice The case of telepsychiatry Social Science and Medicine (1982) 52 1889ndash1901
McCullough J S (2008) The adoption of hospital information systems Health Economics 17 649ndash664
Melnyk B M amp Fineout-Overholt E (2006) Consumer preferences and values as an integral key to evidence-based practice Nursing Administration Quarterly 30 123
Menachemi N (2006) Barriers to ambulatory EHR Who are lsquoimminent adoptersrsquo and how do they differ from other physicians Informatics in Primary Care 14 101ndash108
Menachemi N (2007) Hospital adoption of information technologies and improved patient safety A study of 98 hospitals in Florida
Menachemi N Brooks R G amp Simpson L (2007) The relationship between pediatric volume and information technology adoption in hospitals Quality Management in Health Care 16 146ndash152
Menachemi N Burke D Clawson A amp Brooks R G (2005) Information technologies in Floridarsquos rural hospitals Does system affi liation matter The Journal of Rural Health 21 263ndash268
Menachemi N Burke D E amp Ayers D J (2004) Factors affecting the adoption of telemedi-cinemdashA multiple adopter perspective Journal of Medical Systems 28 617ndash632
Menachemi N Burke D amp Brooks R G (2004) Adoption factors associated with patient safety-related information technology Journal for Healthcare Quality 26 39ndash44
Menachemi N Chukmaitov A Saunders C amp Brooks R G (2008) Hospital quality of care Does information technology matter The relationship between information technology adop-tion and quality of care Health Care Management Review 33 51
Menachemi N Hikmet N Bhattacherjee A Chukmaitov A amp Brooks R G (2007) The effect of payer mix on the adoption of information technologies by hospitals Health Care Management Review 32 102
Menachemi N Matthews M C Ford E W amp Brooks R G (2007) The infl uence of payer mix on electronic health record adoption by physicians Health Care Management Review 32 111
Menachemi N Saunders C Chukmaitov A Matthews M C amp Brooks R G (2007) Hospital adoption of information technologies and improved patient safety A study of 98 hospitals in Florida Journal of Healthcare ManagementAmerican College of Healthcare Executives 52 398
Middleton B Hammond W E Brennan P F amp Cooper G F (2005) Accelerating US EHR adoption How to get there from here Recommendations based on the 2004 ACMI retreat Journal of the American Medical Informatics Association 12
Mojtabai R (2007) Datapoints Use of information technology by psychiatrists and other medical providers Psychiatric Services 58 1261
NAHIT releases HIT defi nitions|News|Healthcare Informatics Park Y amp Chen J V (2007) Acceptance and adoption of the innovative use of smartphone
Industrial Management and Data Systems 107 1349
NA Behkami and TU Daim
35
Poon E G Blumenthal D Jaggi T Honour M M Bates D W amp Kaushal R (2004) Overcoming barriers to adopting and implementing computerized physician order entry sys-tems in US hospitals Health Affairs 23 184ndash190
Poon E G Jha A K Christino M Honour M M Fernandopulle R Middleton B et al (2006) Assessing the level of healthcare information technology adoption in the United States A snapshot BMC Medical Informatics and Decision Making 6 1
Poulsen P B Vondeling H Dirksen C D Adamsen S Go P M amp Ament A J (2001) Timing of adoption of laparoscopic cholecystectomy in Denmark and in The Netherlands A comparative study Health Policy 55 85ndash95
Powner D A (2006) Health information technology HHS is continuing efforts to defi ne a national strategy Testimony before the Subcommittee on Federal Workforce and Agency Organization Committee on Government Reform House of Representatives Government Accountability Offi ce (Vol 15 pp 7ndash8)
Reardon J L amp Davidson E (2007) An organizational learning perspective on the assimilation of electronic medical records among small physician practices European Journal of Information Systems 16 681ndash694
Reference model of open distributed processing Wiki Robeznieks A (2005a) Privacy fear factor arises (Cover story) Modern Healthcare 35 6ndash16 Robeznieks A (2005b) Privacy fear factor arises The public sees benefi ts to be had from health-
care IT but concerns about misuse of data emerge in survey Modern Healthcare 35 6 Rosenfeld S Bernasek C amp Mendelson D (2005) Medicarersquos next voyage Encouraging phy-
sicians to adopt health information technology Health Affairs 24 1138ndash1146 Saouli M A (2004) Information technology utilization in mental health services Thesis
(DPA)mdashUniversity of La Verne 2004 Shields A E Shin P Leu M G Levy D E Betancourt R M Hawkins D et al (2007)
Adoption of health information technology in community health centers Results of a national survey Health Affairs 26 1373
Simpson S (2000) Intra-institutional rivalry and policy entrepreneurship in the European union The politics of information and communications technology convergence New Media and Society 2 445
Simpson R L (2007) The politics of information technology Nursing Administration Quarterly 31 354ndash358
Sterman J amp Sterman J D (2000) Business dynamics Systems thinking and modeling for a complex world with CD-ROM Irwin McGraw-Hill
Tang P C Ash J S Bates D W Overhage J M amp Sands D Z (2006) Personal health records Defi nitions benefi ts and strategies for overcoming barriers to adoption Journal of the American Medical Informatics Association 13 121ndash126
The Zachman Frameworktrade The offi cial concise defi nition US Department of Health amp Human Services Centers for Medicare amp Medicaid Services Wainwright W D amp Waring S T (2007) The application and adaptation of a diffusion of inno-
vation framework for information systems research in NHS general medical practice Journal of Information Technology 22 44ndash58
Wilcox A B Dorr D A Burns L Jones S Poll J amp Bunker C (2007) Physician perspec-tives of nurse care management located in primary care clinics Care Management Journals 8 58ndash63
2 Background Literature on the Adoption of Health Information Technologies
37copy Springer International Publishing Switzerland 2016 TU Daim et al Healthcare Technology Innovation Adoption Innovation Technology and Knowledge Management DOI 101007978-3-319-17975-9_3
Chapter 3 Methods and Models
Nima A Behkami and Tugrul U Daim
N A Behkami Merck Research Laboratories Boston MA USA
T U Daim () Portland State University Portland OR USA e-mail tugruludaimpdxedu
31 Proposed Model Overview and Justifi cation
Most classical and modern adoption literature attempts to defi ne awareness of an innovation (aka knowledge) as the main factor effecting diffusion Meaning once awareness occurs followed by a persuasion stage the innovation stands a chance for diffusion This explanation is often incomplete and at best more appropriate for consumer behavior than applicable to organizational (ie hospital) adoption of innovations Therefore a new perspective on diffusion of organizational innovations as product of three parts is needed and this proposal is a step toward such explana-tion awareness plus condition plus capabilities Figure 31 shows questions relevant to each of these three factors and how individual adoptions will accumulate to become diffusion of an innovation Figure 32 compares the data and decision fl ow in existing diffusion models with the one in newly proposed extensions
Figure 33 summarizes the proposed extensions to Rogersrsquo diffusion theory using dynamic capabilities The top part of the diagram shows the stages in the classical Rogersrsquo diffusion theory where adopters move through the stages of knowledge persuasion decision implementation and confi rmation The bottom part of the dia-gram shows the proposed extensions for condition (existence of it) and capability (acquiring and actually using it) Figure 34 shows the state chart for the new diffu-sion view using the proposed extensions Figure 35 shows how using a capability- based view rather than a knowledge-based (awareness) can show precisely how an adopter can be pushed out on the technology adoption life cycle (depending on when the organization is ready to adopt)
38
(Does the organization knowabout this HIT Innovation) Awareness
+
+
+
+
+
+
Awareness
Awareness
Condition Adoption 1
Adoption N
DiffusionAdoption Condition
Condition
Capabilities
Capabilities
Capabilities
(Does the organization have theCompetencies need to adoptthe Innovation)
(Does Adopting the innovationfinanciallyother make sense)
Fig 31 Capability-based diffusion
Fig 32 Flow of diffusion in existing research vs proposed
NA Behkami and TU Daim
39
32 Modeling Approach
In researching the HIT diffusion phenomena using system thinking this proposed research has two overarching goals One is ldquoto understandrdquo and the other is ldquoto improverdquo To understand means and refers to all the activities related to
Fig 33 New extensions to Rogersrsquo DOI theory
Adopter Knowledge Persuasion
Condition Capabilities
Decision Implementaton
DiffusionAdopters
Confirmation
Fig 34 Diffusion state chart with new extensions
Fig 35 Time element of capabilities in diffusion
3 Methods and Models
40
investigating and later describing the problem space To improve means and refers to all the activities to use the description and use it to improve the existing condi-tion or problem Naturally various research traditions tools techniques and theo-ries can be used to assist in achieving these two goals (Forrester 1994 ) Figure 36 shows the phases of research model building using system thinking that are appro-priate for the proposed HIT diffusion study ldquoTo understandrdquo includes prototyping modeling documenting and communicating research models and fi ndings ldquoTo improverdquo includes using documentation and communication simulation and changing through new policy or theories Inside each of the boxes in Fig 36 the artifacts used for that activity are listed For example technology management constructs scientifi c theories and research methods are tools for m odeling In the following sections various methods and tools for modeling simulation theoriz-ing and research methods that were investigated as candidate for this research are described and discussed
33 Diffusion Theory
ldquoDiffusion is the process in which an innovation is communicated through certain channels over time among the members of a social systemrdquo (Rogers 2003) This special type of communication is concerned with new ideas It is through this pro-cess that stakeholders create and share information together in order to reach a shared understanding Some researchers use the term ldquodisseminationrdquo for diffusion that is directed and planned In his classic work (Rogers 2003) Rogers identifi es four main elements in the diffusion process that are virtually present in all diffusion research (1) an innovation (2) communication channels (3) over time and (4) social systems The following sections provide an overview of each of these process elements
Fig 36 Phases of research model building using system thinking
NA Behkami and TU Daim
41
331 An Innovation
An innovation is a new idea or product perceived useful by an individual or an organization Newness is not measured by the time passed since inception of the idea it is rather the point of time that the individual becomes aware of the perceived benefi ts of the innovation The innovation can have a physical form such as the television or a personal computer Or it can also be entirely composed of informa-tion such as a political view a business idea or a software innovation A method-ological diffi culty exists in that it is not easy to track and evaluate information-based innovations (Rogers 2003)
Innovations encounter different adoption rates For example administrating lemon juice to Navy soldiers in order to prevent illness during long voyages take over a 100 years to be adopted by the Western Navies By contrast youtubecom has reached astronomic number of daily users since its inception in 2005 Understanding Rogersrsquo ldquoperceived attributes of innovationsrdquo helps explain this variance in adoption rates
3311 Relative Advantage
Advantage is defi ned in terms of a benefi t gained Therefore relative advantage in this case is the amount of benefi t realized using the new innovation rather than apply-ing the existing and older solutions This relative advantage can be in the form of economic gain or non-tangible gains such as improved perception safety or peace of mind Relative advantage has a positive effect on an innovation rate of adoption The higher the perceived value of an innovation the faster its adoption rate
3312 Compatibility
Compatibility is referred to as how good of a fi t the new innovation is with the cur-rent structure of values past experiences and needs of candidate adopters An idea that is ill fi t for an organization will face slower adoption rate or may never be adopted For an unfi t innovation to be adopted by an organization it requires the culture and value structure of the adopters to change
3313 Complexity
The extent that an innovation is challenging to use or understand is the complexity attribute of the innovation Innovations that can easily be understood by the majority of population donrsquot require specialized skill and knowledge For example a nontech-nical project manager may have diffi culty understanding the need for adopting a cer-tain technology that would provide the company a competitive advantage Ideas that are simpler and require little or no amount of learning achieve faster adoption rates
3 Methods and Models
42
3314 Trialability
New innovations that can be tried within a restricted scope prior to adoption are said to be trialable The easier it is to try out a new idea the higher the chance of its adop-tion by potential participants The concept of trial has become immensely popular with software innovation Many software vendors allow a close to full product dem-onstration of their products over an extended period of time (usually 30 days) The feeling of uncertainty inherent in adopters can be reduced by a trial of a new innova-tion The new learning can lead to a more rapid adoption
3315 Observability
Observability is the extent that results of an adoption of a new innovation are notice-able by other people The more noticeable innovations are adopted more quickly Observability information is mostly communicated through peer-to-peer networks
332 Recent Diffusion of Innovation Issues
Based on a literature review for criticisms and limitations of diffusion theory some of the more recent issues are listed and described in this section
Diffusion research is spreading from industrial settings to public policy setting as well DOI research was started in industrial and service settings and ever since it has been concentrated in areas of study such as agriculture manufacturing and electronics Success in those fi elds has prompted applying DOI research in areas such as public service and policy innovation for example healthcare and education (Nutley amp Davies 2000 )
Diffusion of innovation is not as linear process as most researches suggest Traditional research has described the DOI process as one that fl ows through the fol-lowing steps research creation dissemination and fi nally utilization These steps describe a more or less linear process Studies have shown that in fact often innovations donrsquot spread throughout the population in such a manner and instead experience vari-ous iterations and loops among the stages (Cousins amp Simon 1996 ) Therefore to have a better understanding of the DOI process the entire picture needs to be evaluated
Interests in diffusion research still remains high Wolfe conducted a literature review on diffusion of innovation from 1989 to 1994 and identifi ed 6240 articles on this topic (Wolfe 1994 ) A similar search was performed by Nutley from 1990 to 2002 that identifi ed 14600 articles (Nutley amp Davies 2000 ) This twofold increase highlights the increasing research interest in this area Increase may be contributed to public policy health and energy and consumer diffusion research
NA Behkami and TU Daim
43
Research has not characterized organization innovativeness Structure of inno-vative organizations has been subject of many studies Their ability and attitude toward adopting innovation have been measured in various ways (Damanpour 1988 1991 ) However we yet donrsquot have a characterization of an organization that is more innovative vs one that is slower to adopt innovation (Nutley amp Davies 2000 )
The path diffusion of innovation fl ows is unpredictable Path of diffusion is the stages an innovation passes through from inception to utilization Van de Ven argues that qualitative DOI studies have highlighted that it may be better not to discuss dif-fusion in terms of a predictable or unpredictable path (Vandeven amp Rogers 1988 ) similar to Cousins and Simon argument that diffusion process is not linear To think of the complex process of diffusion in terms of a predictable process may corner us into trying to fi t research into this otherwise incorrect notion of predictability
Innovation type classifi cation To better understand and evaluate the effective-ness of diffusion of innovation itrsquos important to be able to classify types of innova-tions Types can have similarities but also each type may uncover peculiarities that are important to be noted Damanpour and Evens have proposed two simple classi-fi cations fi rst technical vs administrative innovations and second product vs pro-cess innovations (Damanpour amp Evan 1984 ) Wolfe has provided more resolution to innovation types with 17 innovation attributes (Wolfe 1994 ) More recently Osborne has classifi ed social policy innovations (Osborne 1998 )
Innovation adopter decisions are more based on fad and fashion than rationality
A rational decision is one that is made with the desirable outcomes in mind A logi-cal process is followed and is free of peer network pressure and current fashion Research has shown that similar to consumer markets innovation adopters are heav-ily infl uenced by fad and fashion when deciding to adopt (Abrahamson 1991 1996 ) The need for peer acceptance is a large driver of adoption behavior (ONeill Pouder amp Buchholtz 1998 ) To have a correct understanding itrsquos critical to keep this variable in mind when studying and evaluating innovation diffusion
Adoption decision reversal Much of the research has focused on the adoption decision process itself The phenomena of adoption reversal have mostly been neglected Even after making an adoption decision adopters look for continuous reinforcements within their network if they are exposed to negative press they attempt to reverse their adoption decision (Rogers 2003)
Staged diffusion models The most sited model of diffusion is Rogersrsquo fi ve- stage illustration (Rogers 2003) Rogersrsquo model includes the following stages in order knowledge persuasion decision implementation and confi rmation Other authors have proposed variation to Rogersrsquo model to include routinization and infusion (Cooper amp Zmud 1990 ) Routinization occurs when adoption is no longer consid-ered innovative this is normally seen in late adopters Infusion occurs when innova-tion has been adopted by an organization and it has spread strongly within that organization
3 Methods and Models
44
Additional innovation characteristics In his classic work Rogers identifi ed the following innovation attributes relative advantage comparability compatibility trialability and observiblity (Rogers 2003) Building on his work other attributes have been suggested such as adoptability centrality and additional work load (Wolfe 1994 )
Linear-stage model inadequate (innovation journey) Linear models that have so far been defi ned for innovation diffusion are limited Linear models assume tech-nology fl ows from one step to the other in a waterfall manner Based on case studies such as the Minnesota Innovation Research Program (MIRP) the process is more and more being visualized as a journey termed the ldquoinnovation journeyrdquo (Vandeven amp Rogers 1988 ) The new fi ndings show that DOI is non-sequential chaotic and impulsive The new learning highlights that there are no simple solutions but orga-nizations can learn from their past adoption experiences to improve future projects While there are no simple representations of the process and no ldquoquick fi xesrdquo to ensure that it is successful participants who learn from their past experience can increase the odds of their success (Nutley amp Davies 2000 )
Institutional pressure is a large factor in adoption decisions Abrahamson et al introduced administrative innovations as a new type The authors explained how groups adopt or reject administrative innovations They argue that rather than evi-dence institutional pressures coming from certain fads and fashion infl uence the adopter (Abrahamson 1991 1996 Abrahamson amp Fombrun 1994 Abrahamson amp Rosenkopf 1993 )
Decentralized systems are most appropriate (for not highly technical adop-tions) In the newest revision of his book Rogers argues that decentralized systems are best diffused when a high level of new technical learning expertise is not needed and the users are very mixed in expertise and skills (Rogers 2003)
333 Limitations of Innovation Research
According to Nutley (Nutley amp Davies 2000 ) to date Wolfe identifi es the following limitations in innovation research (Wolfe 1994 )
bull Lack of specifi city concerning the innovation stage upon which investigations focus
bull Insuffi cient consideration given to innovation characteristics and how these change over time
bull Research being limited to single-type studies bull Researchers limiting their scope of inquiry by working within single theoretical
perspectives
NA Behkami and TU Daim
45
34 Other Relevant Diffusion and Adoption Theories
A macro-level (market-levelecosystem-level) theory such as diffusion theory is better suited for describing activities of multiple fi rms in a space that can have policy implications (Erdil amp Emerson 2008 Otto amp Simon 2009 ) However for example theories such as the technology acceptance model (TAM) are at the indi-vidual (micro) level which is better suited for analyzing the atomic individual deci-sion (can later be built into a market-level theory such as diffusion models) Therefore for the proposed HIT study diffusion theory is the best fi t Table 31 lists other relevant theories relating to adoption and diffusion that were considered before deciding on using diffusion theory for this research The following sections describe each theory in detail and discuss its strength and weakness as relevant to this research effort (Fig 37 )
Table 31 List of relevant diffusion and adoption theories
Name
Main dependent construct
Main independent construct Originating area
Level of analysis
Technology acceptance model (TAM)
Behavioral intention to use system usage
Perceived usefulness perceived ease of use
Information systems
Individual
Theory of reasoned action (TRA)
Behavioral intention behavior
Attitude toward behavior subjective norm
Social psychology
Individual
Theory of planned behavior (TPB)
Behavioral intention behavior
Attitude toward behavior subjective norm perceived behavioral control
Social psychology
Individual
Unifi ed theory of acceptance and use of technology (UTAUT)
Behavioral intention usage behavior
Performance expectancy effort expectancy social infl uence facilitating conditions gender age experience voluntariness of use
Information systems
Individual
Technology-organization- environment framework (TOEF)
Likelihood of adoption intention to adopt extent of adoption
Technological context Organizational context Environmental context
Organizational psychology
Firmorganization
Matching Person and technology model (MPTM)
Behavior Attitude Social sciences Individual
Lazy user model (LUM)
Behavior Attitude Engineering Individual
3 Methods and Models
46
341 The Theory of Reasoned Action
According to the theory of reasoned action (TRA) an individualrsquos behavior is guided by an individualrsquos attitude along with the subjective norms (Ajzen amp Fishbein 1973 Fishbein 1967 Fishbein amp Ajzen 1975 ) as illustrated in Fig 38 An individualrsquos positive or negative attitude toward conducting a behavior is defi ned as the attitude toward act or behavior Assessing an individualrsquos belief regarding results of acting and desirability of that result determine the attitude Subjective norm is described as whether the individualrsquos environment and other people in it feel itrsquos positive or nega-tive for a behavior to be performed The strength of subjective norm factor on actual behavior of the individual is affected by the level of strength the individual wished to conform to opinions of the others
The TRA model has two important limitations (Eagly amp Chaiken 1993 ) First there can be confusion between attitude and subjective norm since attitudes can often be driven or be products of subjective norms or vice versa The other limita-tion of the model is that it does not consider constraints imposed on individual behavior In other words it assumes free will to behave independent of constraints such as time environment and laws
342 The Technology Acceptance Model
The TAM model is an adaptation of the TRA for the information technology (IT) domain How users reach the point to adopt a technology and use it is explained by TAM TAM hypothesizes that perceived usefulness and perceived ease of use are
Attitude TowardAct or Behavior
BehavioralIntention
Behavior
Subjective Norm
Fig 38 Theory of reasoned action (TRA)
Fig 37 Market level vs fi rm level
NA Behkami and TU Daim
47
the determinants for an individualrsquos intention to use a system or not as shown in the top part of Fig 39 (Davis 1985 1989 Davis Bagozzi amp Warshaw 1989 ) Perceived usefulness is defi ned as the degree that an individual believes using a technology would improve hisher performance Perceived ease of use is defi ned as the level an individual believes using a technology would bring himher effi ciently by saving them effort for otherwise needed work Perceived usefulness can also be directly impacted by perceived ease of use
In order to simplify the TAM model researchers have removed the attitude constrict from the original TRA (Venkatesh et al 2003 ) In the literature various efforts have been made to extend TAM which these efforts generally fall into one of the following three categories adding infl uential parameters from other related models adding brand new parameters to the model not found in other models and fi nally examining various infl uences on perceived usefulness and perceived ease of use (Wixom amp Todd 2005 ) The relationship between usefulness ease of use and system usage have been explored since the original work on TAM (Adams Nelson amp Todd 1992 Davis et al 1989 Hendrickson Massey amp Cronan 1993 Segars amp Grover 1993 Subramanian 1994 Szajna 1994 ) Similar to the limitations of TRA TAM also assumes that intention to act is formed free of limitations and constraints such as time environment and capability In addi-tion triviality and lack of practical value have been recently highlighted as limita-tions of TAM (Chuttur 2009 ) The original TAM has been extended to now include social infl uence and instrumental processes in TAM2 (Viswanath Morris Davis amp Davis 2003 )
A Possible Dynamic Capabilitiesextension to TAM
Classic TAM Model
PeroeivedUsefulness
Peroeived Ease of Use
BehavioralIntention to Use
Capabilities toUse Exists
Actual SystemUse
PeroeivedUsefulness
Peroeived Ease of Use
BehavioralIntention to Use
Source Davis et al (1989) Venkatesh et al (2003)
Actual SystemUse
Fig 39 Theory of technology acceptance model (TAM)
3 Methods and Models
48
As explained earlier for the proposed study the methodology of choice is diffu-sion theory since it provides a macro-level view However dynamic capabilities can also be integrated with the TAM model For example as shown in the bottom part of Fig 39 a new ldquocapabilities to use existrdquo construct can be added to the classic TAM which would infl uence the existing ldquobehavioral intentions to userdquo or ldquoactual system userdquo constructs One of the main diffi culties in this integration is that unlike diffu-sion theory TAM does not provide a way to describe a time element
343 The Theory of Planned Behavior
The theory of planned behavior (TPB) model states that an individualrsquos behavior is powered by behavioral intentions which are infl uenced by attitude subjective norm and perceptions of ease of use as in Fig 310 (Ajzen 1985 1991 ) The originating fi eld for this theory is psychology and it was proposed as an extension to TRA Similar to the components of TRA model an individualrsquos positive or negative attitude toward performing a behavior is defi ned as the attitude toward act or behavior Subjective norm is described as whether the individualrsquos environment and other people in it feel itrsquos positive or negative for a behavior to be performed Behavioral control is described as an individualrsquos perception of how diffi cult it is to perform an act or behavior
344 The Unifi ed Theory of Acceptance and Use of Technology
The unifi ed theory of acceptance and use of technology (UTAUT) was developed to explain the individualrsquos intentions in using an information system and its resulting behavior as in Fig 311 UTAUT was developed based on the combination of com-ponents identifi ed by previous models including theory of reasoned action TAM motivational model theory of planned behavior a combined theory of planned behaviortechnology acceptance model model of PC utilization innovation
Attitude TowardAct or Behavior
Subjective NormBehavioralIntention Behavior
Source Ajzen (1991)
PerceivedBehavioral
Control
Fig 310 Theory of planned behavior (TPB)
NA Behkami and TU Daim
49
diffusion theory and social cognitive theory Its hypostasis that the four constructs of performance expectancy effort expectancy social infl uence and facilitating con-ditions can explain usage intention and resulting behavior (Viswanath et al 2003 ) Gender age experience and voluntariness of use were identifi ed as other important parameters in explaining usage and behavior (Viswanath et al 2003 )
345 Matching Person and Technology Model
Matching person and technology model (MPTM) is a way to organize infl uences on the successful adoption and use of technologies in systems in settings such as the workplace home and healthcare settings Research has shown that a well- intentioned technology may not arrive at its full potential if the important personal-ity preference psychosocial characteristics or necessary environmental support critical are not considered An MPTM assessment can help match individuals with the most appropriate technologies for their intended use (Scherer 2002 )
346 Technology-Organization-Environment Framework (TOE)
TOEF framework identifi es technological organizational and environmental contexts as the components of the processes by which fi rms adopt and use technological inno-vations (Tornatzky amp Fleischer 1990 ) The scope of technological context includes both external and internal artifacts relevant to the fi rm Both physical equipments and processes are part of the technological context Organizational context includes the
UseBehavior
BehavioralIntention
Voluntarinessof Use
ExperienceAgeGender
PerformanceExpectancy
EffortExpectancy
SocialInfluence
FancilitatingConditions
Fig 311 The unifi ed theory of acceptance and use of technology
3 Methods and Models
50
characteristics of the fi rm fi rm size degree of centralization managerial structure and the likes The environment context can include the size and structure of the market ecosystem including competition regulations and more
347 Lazy User Model
Similar to the TAM lazy user model (LUM) attempts to describe the process that individuals use to select a solution for satisfying a need from a series of alternatives (Collan amp Teacutetard 2007 ) LUM hypothesizes that from a set of available solutions the user always attempts to select the one with the least amount of effort
The model starts by assuming that the user has a need that is defi nable and satisfi able Then the set of possible solutions are defi ned by the user need Each solution in the set has its own characteristics which meet the user need in varying degrees The user state further determines the available solutions For example to check an address for a restaurant an individual can use the Internet or a tele-phone But if this individual is driving and is without an Internet connection heshe can either call the phone directory to get the restaurant phone number or phone a friend for directions Therefore as in this example the user state is deter-mined by the users and their situation characteristics at any given time
The LUM model assumes that after the user need and user state have defi ned the set of possible solutions the user will select a solution Worth mentioning that if the set is empty the user does not have a way to satisfy the need The LUM hypothesizes that the use will select a solution from the limited set based on lowest level of effort Effort is defi ned as aggregate of monetary cost + time needed + physical andor mental efforts necessary to satisfy the user need (Tetard amp Collan 1899 )
35 Resource-Based Theory Invisible Assets Competencies and Capabilities
As described in the earlier sections of this document dynamic capabilities are one of the main constructs that are being proposed for extending diffusion theory for HIT adoption What is specifi cally referred to as dynamic capabilities is also generally discussed by researchers through other explanations such as competencies factors of production assets and more The roots of almost all of these variations can be traced back to resource-based theory (RBT) Before deciding on dynamic capabili-ties it was important to review and compare all the variations of so-called factors of production Almost any of the variations would be useable for the proposal since itrsquos merely intended to demonstrate the existence of organizational ability (capabil-ity) However since adoption of HIT would require obtaining new abilities or recon-fi guring existing abilities this is most consistent with the dynamic qualifi cation of dynamic capabilities
NA Behkami and TU Daim
51
Strategic management researchers attempt to understand differences in fi rm per-formance by asking the question ldquoWhy do some fi rms persistently outperform othersrdquo(Barney amp Clark 2007 ) Understanding this point has traditionally been looked at from a strategic management point of view in the context of creating com-petitive advantage or diversifying the corporate portfolio But interesting enough studying the differences in this performance can also help us understand diffusion of innovation In this context one of the major goals of research industry society and especially government is the accelerated diffusion of information in healthcare technology So knowing how why and which fi rms outperform others would allow the stakeholders involved to make better policy and plan more precisely It is in this context that this research proposes using dynamic capabilities to model diffusion of HIT In order to better understand its importance it is useful to look at the history of this research how it developed and what alternative candidates to dynamic capa-bilities there are This is done in the following sections by reviewing the foundations of RBT seminal work in the area variations classifi cations and limitations
351 Foundations of Resource-Based Theory
Firmsrsquo outperforming other fi rms has been explained using two explanations in the literature (Barney amp Clark 2007 ) The fi rst is attributed to Porter (Porter 1981 Porter Michael 1979 ) and is based on structure-conduct-performance (SCP) theory from industrial organization economics (Bain 1956 ) This perspective argues that a fi rmrsquos market power to increase prices above a competitive level creates the superior performance (Porter 1981 ) The second explains superior performance through the differential ability of those fi rms to more rapidly and cost effectively react to cus-tomer needs (Demsetz 1973 ) This perspective suggests that it is resource intensive for fi rms to copy more effi cient fi rms hence this causes the superior performance to persist between the haves and the have-nots (Rumelt amp Lamb 1984 )
In RBD Barney acknowledges that these two explanations are not contradictory and each applies in some settings While also acknowledging the roll of market power in explaining sustained superior performance Barney chooses to ignore it and instead focus on ldquoeffi ciency theories of sustained superior fi rm performancerdquo (Barney amp Clark 2007 )
Four sources contribute to theoretical underpinnings of RBD (Barney amp Clark 2007 ) (a) distinctive competencies research (b) Ricardorsquos analysis of land rents (c) Penrose 1959 (Penrose 1959 ) and (d) studies of antitrust implications of economics Of the four parts only distinctive competencies and Penrosersquos work are related to this proposed research and will be explained in more detail in the following subsections
3511 Distinctive Competencies
A fi rmrsquos distinctive competencies are the characteristics of the fi rm that enable it to implement a strategy more effi ciently than other fi rms (Hitt amp Ireland 1985a 1986 Hrebiniak amp Snow 1982 Learned Christensen Andrews amp Guth 1969 ) One of
3 Methods and Models
52
the early distinctive competencies that researchers identifi ed was ldquogeneral manage-ment capabilityrdquo The thinking was that fi rms that employ high-quality general man-agers often outperform fi rms with ldquolow-qualityrdquo general managers However it is now understood that this perspective is severely limited in explaining performance difference among fi rms First the qualities and attributes that constitute a high- quality general manager are ambiguous and diffi cult to identify (a platter of research literature has shown that general managers with a wide array of styles can be effec-tive) Second while general management capabilities are important itrsquos not the only competence critical in the superior performance of a fi rm For example a fi rm with high-quality general managers may lack the other resources ultimately necessary to gain competitive advantage (Barney amp Clark 2007 )
3512 Penrose 1959
In the work The Theory of the Growth in 1959 Penrose attempted to understand the processes that lead to fi rm growth and its limitations (Penrose 1959 ) Penrose advocated that fi rms should be conceptualized as follows fi rst an administrative framework that coordinates activities of the fi rm and second as a bundle of produc-tive resources Penrose identifi ed that the fi rmrsquos growth was limited by opportuni-ties and the coordination of the fi rm resources In addition to analyzing the ability of fi rms to grow Penrose made two important contributions to RBD (Barney amp Clark 2007 ) First Penrose observed that the bundle of resources controlled can be different from fi rm to fi rm in the same market Second and most relevant to this research proposal Penrose used a liberal defi nition for what might be considered a productive resource including managerial teams top management groups and entrepreneurial skills
352 Seminal Work in Resource-Based Theory
Four seminal papers constituted the early work on RBT these included Wernerfelt (1984) Rumelt (1984) Barney (1986) and Dierickx (1989) (Barney 1986 Dierickx amp Cool 1989 Rumelt amp Lamb 1984 Wernerfelt 1984 ) These papers made it pos-sible to analyze fi rmrsquos superior performance using resources as a unit of analysis They also explained the attributes resource must have in order to be source of sus-tained superior performance
Using the set of resources a fi rm holds and based on the fi rmrsquos product market position Wernerfelt developed a theory for explaining competitive advantage (Wernerfelt 1984 ) that is complementary to Porters (Porter 1985 ) Wernerfelt labeled this idea resource-based ldquoviewrdquo since it looked at the fi rmrsquos competitive advantage from the perspective of the resources controlled by the fi rm This method argues that the collection of resources a fi rm controls determines the collections of product market positions that the fi rm takes
NA Behkami and TU Daim
53
Around the same time as Wernerfelt Rumelt published a second infl uential paper that tried to explain why fi rms exist based on being able to more effi ciently generate economic rents than other types of economic organizations (Rumelt amp Lamb 1984 ) An important contribution of Rumelt to RBD was that he described fi rms as a bun-dle of productive resources
In a third paper similar to Wernerfelt Barney recommended a superior perfor-mance theory based on attributes of the resources a fi rm controls (Barney 1986 Wernerfelt 1984 ) However Barney additionally argued that a theory based on product market positions of the fi rms can be very different than the pervious and therefore a shift from resource-based view to the new RBD (Barney amp Clark 2007 ) In a fourth paper Dierickx and Cool supported Barneyrsquos argument by explaining how it is that the resources already controlled by fi rm can produce economic rents for it (Dierickx amp Cool 1989 )
353 Invisible Assets and Competencies Parallel Streams of ldquoResource-Based Workrdquo
While RBD was shaping into its own other research streams were developing theories about competitive advantage that have implications to this proposed research since they were also looking at competencies and capabilities The most infl uential were the theory of invisible assets by Itami and Roehl ( 1987 ) and competence-based theo-ries of corporate diversifi cation (Hamel amp Prahalad 1990 Prahalad amp Bettis 1986 )
Itami described sources of competitive power by classifying physical (visible) assets and invisible assets Itami identifi ed information-based resources for exam-ple technology customer trust and corporate culture as invisible assets and the real source of competitive advantage while stating that the physical (visible) assets are critical to business operations but donrsquot contribute as much to source of competitive advantage Firms are both accumulators and producers of invisible assets and since it is diffi cult to obtain them having them can lead to competitive advantage Itami classifi ed the invisible assets into environment corporate and internal categories Environmental information fl ows from the environment to the fi rms such as cus-tomer information Corporate information fl ows from the fi rm to its ecosystem such as corporate image Internal information rises and gets consumed within the fi rm such as morale of workers
In another parallel research stream Teece and Prahalad et al (Prahalad amp Bettis 1986 Teece 1980 ) had started looking at resource-based logic to describe corporate diversifi cation Prahalad in particular stresses the importance of sharing intangible assets and its impact on diversifi cation Prahalad and Bettis called these intangible assets the fi rmrsquos dominant logic ldquoa mindset or a worldview or conceptualization of the business and administrative tools to accomplish goals and make decisions in that busi-nessrdquo Hamel and Prahalad ( 1990 ) extended dominate logic into the corporation ldquocore competence rdquo meaning ldquothe collective learning in the organization especially how to coordinate diver production skills and integrate multiple streams of technologiesrdquo
3 Methods and Models
54
354 A Complete List of Terms Used to Refer to Factors of Production in Literature
For the purposes of this proposal the various forms of factors of production have been extracted from literature and presented here in Table 32 The table includes the name of the view its source and some brief notes
Table 32 List of names used for factors of production in literature
Nameunit Source Notes
1 Firmrsquos distinctive competencies
Learned et al ( 1969 ) Hrebiniak and Snow ( 1982 ) Hitt and Ireland ( 1985a 1985b ) Hitt and Ireland ( 1986 )
Aka general management capability
2 Factors of production
Ricardo ( 1817 ) For example the total supply of land
3 Bundle of productive resources
Penrose ( 1959 ) Managers exploit the bundle of productive resources controlled by a fi rm through the use of the administrative framework that had been created in a fi rm
4 Invisible assets and physical (visible) assets
Itami and Roehl ( 1987 )
Invisible assets are necessary for competitive success Physical (visible) assets must be present for business operations to take place
5 Shared intangible assets (called fi rmrsquos dominant logic)
Prahalad and Bettis ( 1986 )
A mindset or a worldview or conceptualization of the business and administrative tools to accomplish goals and make decisions in that business
6 Corporationrsquos ldquocore competencerdquo
Hamel and Prahalad ( 1990 )
The collective learning in the organization especially how to coordinate diverse production skills and integrate multiple streams of technologies
7 Resources Barney ( 1991 Wernerfelt ( 1984 )
Simply called these assets ldquoresourcesrdquo and made no effort to divide them into any fi ner categories
8 Capabilities Stalk Evans and Shulman ( 1992 )
Argued that there was a difference between competencies and capabilities
9 Dynamic capabilities
Teece Pisano and Shuen ( 1997 )
The ability of fi rms to develop new capabilities
10 Knowledge Grant ( 1996 Liebeskind 1996 Spender and Grant 1996 )
Knowledge-based theory
11 Firm attributes Barney and Clark ( 2007 )
A causal reference to factors of production
12 Organizational capabilities (organizational routines)
Nelson and Winter ( 1982 )
Organizational routines are considered basic components of organizational behavior and repositories of organizational capabilities
NA Behkami and TU Daim
55
355 Typology and Classifi cation of Factors of Production
A variety of researchers have created typologies of fi rm resources competencies and capabilities (Amit amp Schoemaker 1993 Barney amp Clark 2007 Collis amp Montgomery 1995 Grant 1991 Hall 1992 Hitt Hoskisson amp Kim 1997 Hitt amp Ireland 1986 Thompson amp Strickland 1983 Williamson 1975 )
36 Modeling Component Descriptions
During research when modeling ecosystems or problem domains for the purposes of system analysis a variety of complementary and sometimes redundant methods exist Choosing the right combination is important and is a multistep process First the need for problem analysis or modeling has to be clear Second a set of alterna-tive solutions needs to be developed and third well-suited combination of tools needs to be picked to demonstrate the problemsolution In order to be able to effectively execute these three steps the researcher needs to be familiar with the tools of the trade Figure 312 shows the building blocks of these tools and the relationships among them A description of each of these building blocks follows in this section
Fig 312 Research and modeling components and their relationships
3 Methods and Models
56
361 Model
A model is a miniature representation or description created to show the structural components of a problem and their interactions They are often limited replicas of real-ity and are used to assist in understanding complex ideas for further studies Models come in a variety of formats including textual mathematical graphical and hybrid
362 Diagram
A diagram is a symbolic representation of information used for visualization pur-poses A diagram is almost always graphical and shows collection(s) of objects and relationships Often the terms model and diagram are incorrectly used in an inter-changeable manner Diagrams can be part of a model however models are usually collection of multiple types of information including text and graphics Models are used to understand problems and are multiple perspectives while diagrams are used to show a specifi c window on an issue
363 View
A view is a representation of a system from a particular perspective Views or view-point frameworks are techniques from systems engineering and software engineer-ing which describe a logical set of related matters to be used during systems analysis and development A view can be part of a model and diagrams can be used to help further elaborate a view However views donrsquot exist without being part of a model or are rendered meaningless that way
364 Domain
Domain is a set of expertise or applications that assist us in defi ning and solving everyday problems Software engineering and healthcare are two examples of domains
365 Modeling Language
A modeling language is an artifi cial language that describes a set of rules which are used to describe structures of information or systems The rules are what provide meaning and description to the various artifacts for example in a graphical
NA Behkami and TU Daim
57
diagram Modeling languages are usually graphical or textual Diagrams contain-ing symbols and lines are usually graphical modeling languages such as Unifi ed Modeling Language (UML) and textual modeling languages use mechanisms such as standardized keywords or other constructs to create understandable expressions
An important point to keep in mind is that not all modeling languages are execut-able For example although UML can be used to generate parts of code itrsquos not executable whereas graphical models such as stock and fl ow diagrams from system dynamics models (even though analysis wise much less descriptive than UML dia-grams) are an executable model Executable models are given values as inputs and after calculations they are able to provide results as outputs
366 Tool
In a general sense a tool is an object that interfaces between two or more domains It enables a useful action from one domain on another For example a system dynamics model which is a tool from the engineering domain can act as an interface for a problem in the healthcare domain
367 Simulation
Simulation is the reproduction of a concept that may be rooted in reality a process or an organization etc Simulation requires modeling key behavior and characteris-tics of the targeted system Simulation is often used to show eventual results of alternative paths or solutions
37 Modeling Technique Trade-Off Analysis for Proposed HIT Diffusion Study
For the proposed HIT diffusion study the following modeling needs can be identifi ed
bull Decompose the HIT adoption ecosystem into actors behaviors etc bull Look at the HIT adoption and diffusion process from various perspectives bull Look at the behavior such as relationships and data exchanged between the
actors bull Document the model bull Simulate or forecast over time
3 Methods and Models
58
Table 33 Need vs solution matrix
UML Theories Systems science and system dynamics
Qualitative methods
Understand and model Actors X X Actor behavior X X Relationships X X Flow of info X X Decisions X X Capabilities X X Policy X X Other X X Prototype Structure X X Behavior X X Model X Simulate Scenarios X X X Model X X Decisions X X Policy X Time X Facilitator and barriers X
bull Prototype bull Communicate the model
In each row of Table 33 the needs mentioned above are shown with more detail The columns list the domain or fi eld that would be used to satisfy that need It is effectively a need vs solution matrix which describes for example UML will be used to prototype structure
Table 34 is an exhaustive list of potential modeling techniques methodologies and tools from softwaresystems engineering and technology management relevant to analyzing and simulating models Members of list that were more relevant to the research are described in detail in the following sections and they include soft sys-tem methodology (SSM) structured system analysis and design method (SSADM) business process modeling (BPM) system dynamics system context diagrams (SCD) data fl ow diagrams (DFDs) fl ow charts UML and Systems Modeling Language (SysML) These tools were examined for applicability in detail before deciding to use the combination listed in Table 33
NA Behkami and TU Daim
59
Table 34 List of relevant system modeling techniques
Full name Abbreviation
Soft systems methodology SSM Business process modeling BPM Systems engineering ndash Software engineering ndash Software development methodology ISDM System development methodology ndash Structured systems analysis and design method SSADM Dynamic systems development method DSDM Structured analysis SA Software design SD Soft systems methodology SSM Structured design ndash Yourdon structured method ndash Jackson structured programming ndash Structured analysis ndash WarnierOrr diagram ndash Soft OR ndash System dynamics ndash Systems thinking ndash General-purpose modeling GPM Graphical modeling languages ndash Algebraic modeling language ndash Domain-specifi c modeling language ndash Framework-specifi c modeling language ndash Object modeling languages ndash Virtual reality modeling languages ndash Fundamental modeling concepts FMC Flow chart ndash Object role modeling ndash Unifi ed modeling language UML Model-driven engineering MDE Model-driven architecture MDA Systems modeling language SysML Functional fl ow block diagram FFBD Mathematical model ndash Functional fl ow block diagram (FFBD) FFBD Data fl ow diagram (DFD) DFD n2 (n-squared) chart ndash idef0 diagram ndash Universal systems language function maps and type maps USL The open group architecture framework TOGAF The British Ministry of Defence Architectural Framework MODAF
(continued)
3 Methods and Models
60
371 Soft System Methodology
Developed by academics at the University of Lancaster Systems Department in the late 1960s SSM is a means to organizational process modeling or also known as BPM (van de Water Schinkel amp Rozier 2006 ) In SSM researchers start with a real-world situation and study the situation in a pseudo-unstructured approach Subsequently rough models of the situation are developed SSM develops specifi c perspectives on the situation builds models from these perspectives and iteratively compares it to the real life (Williams 2005 ) SSM is comprised of seven stages that address the real and conceptual world for the situation under study (Finegan 2003 ) SSM is most useful when the situation under analysis contains multiple stakeholder goals assumptions and perspectives and if the problem is extremely entangled
SSM tries to address many perspectives as a whole and this leads to a complex challenge Clarity is best achieved when addressing key perspectives separately and integrating fi nding from multiple perspectives downstream to this end Checkland developed the mnemonic CATWOE to help (Checkland 1999 Checkland amp Scholes 1990 ) The new tool proposed that the starting point of situation analysis is a transformation (T) asking the question that from a given perspective what is actually transformed moving from input to output Once the transformation has been identifi ed research can proceed to identify other elements of the system (Williams 2005 )
bull Customers who (or what) benefi t from this transformation bull Actors who facilitate the transformation to these customers bull Transformation from ldquostartrdquo to ldquofi nishrdquo bull Weltanschauung what gives the transformation some meaning bull Owner to whom the ldquosystemrdquo is answerable andor could cause it not to exist bull Environment that infl uences but does not control the system
Table 34 (continued)
Full name Abbreviation
Zachman framework ndash Performance moderator function (PMF) models ndash Human behavior models ndash System dynamics ndash Ecosystem model ndash Wicked problem ndash Operations research ndash Stock and fl ow diagrams ndash Causal loop diagrams ndash Dynamical system ndash
NA Behkami and TU Daim
61
372 Structured System Analysis and Design Method
SSADM was developed as a systems approach for the Offi ce of Government Commerce of the UK in the 1980s for the analysis and design of information sys-tems (Robinson amp Berrisford 1994 ) SSADM is comprised of three layers for (1) logical data modeling for modeling the system data requirements (2) data fl ow modeling for documenting how data moves around and (3) entity behavior model-ing to identify events that affect each entity ( SSADM Diagram Software Structured Systems Analysis and Design Methodology ) Figure 322 shows a sample DFD drawn using the SSADM style SSADM consists of fi ve stages which include ( SSADM Diagram Software Structured Systems Analysis and Design Methodology )
Feasibility study A high-level analysis of the situation to a business area to under-stand whether developing a system is feasible Data Flow modeling and (high- level) logical data modeling techniques are used during this stage
Requirement analysis Requirements are identifi ed and the environment is mod-eled Alternative solutions are proposed and a particular option is selected to be further refi ned Data fl ow modeling and logical data modeling technique are used during this stage
Requirement specifi cation Functional and nonfunctional requirements are described
Logical system specifi cation The development and implementation environment is described
Physical design The logical system specs and technical specs are used to create and design a program
373 Business Process Modeling
In systems and software engineering BPM is the activity of describing the enter-prise processes for analysis BPM is often performed to improve process effi -ciency and quality and often involves information technology Newly arriving applications from large-platform vendors make some inroads for allowing BPM models to become executable and capable of use for simulations (Smart Maddern amp Maull 2008 )
374 System Dynamics (SD)
Created during the mid-1950s by Professor Jay Forrester of the Massachusetts Institute of Technology system dynamics is a modeling tool that allows us to build formal computer simulation of complex problem Examples of system dynamics application include studying corporate growth diffusion of new technologies and policy forecasting System dynamics helps us understand better in what ways the
3 Methods and Models
62
fi rmrsquos performance is related to its internal structure (Hendrickson et al 1993 ) SD roots are in control theory and the modern theory of nonlinear dynamics System dynamics is the preferred choice for studying systems at a high level of abstraction where agent-based simulation is better suited for studying phenomena at the level of individuals or other micro levels (Wakeland et al 2004 ) The main components of a system dynamic model include a causal loop diagram (CLD) stock and fl ow dia-gram and its mathematical equations
3741 Causal Loop Diagram
A CLD is a visual illustration of the feedback structures in a system A CLD shows variables connected with arrows illustrating causal infl uences among them CLD can be used for quickly capturing a hypothesis about dynamics of the situation capturing mental models of stakeholders and communicating important feedback that are responsible for the problem being studied CLDs do not show accumulation of resource or rates of change in system that will be in stock and fl ows An example CLD is shown in Fig 313 (Behkami 2009 )
3742 Stock and Flow Diagram
In system dynamics after creating a CLD the next step is to create a stock and fl ow diagram Stocks are accumulations (they characterize the state of the system) and fl ows are rate of accumulation or depletion over time Stocks can create delays by accumulat-ing difference in infl ow versus outfl ow Figure 314 shows a stock and fl ow diagram for a Bass diffusion model Figure 315 shows a sample output for adoption rates from the stock and fl ow diagram in Fig 314 And Fig 316 is a snippet of the differential equi-tations (the behind the scene parts) of the same system dynamics model
375 System Context Diagram and Data Flow Diagrams and Flow Charts
SCD are used to represent external objects or actors that interact with a system (Kossiakoff amp Sweet 2003 ) An SCD illustrates a macro view of a system under investigation showing the whole system with its inputs and outputs related to exter-nal objects This type of diagram is system centric with no details of its interior
LargePotentialAdaptors
SmallPotentialAdaptors
Adaptors Fig 313 Adopter population
NA Behkami and TU Daim
63
PotentialAdopters
P
Total LargePractice Population
N
AdoptionFraction
i
Contact Ratec
MarketSaturation
AdvertisingEffectiveness
a
Adoption fromAdvertising inConferences
B
B
R
MarketSaturation
Adoption RateAR
Word ofMouth
AdoptersA
Adoption fromInstitutional word of
Mouth
+
+
+
+ +
+
-
+
+
Fig 314 Bass diffusion model with system dynamics
20
10100
00
0 10 20 30 40 50
Time (Month)Adoption from Advertising in Conferences CurrentAdoption from Institutional word of Month Current
60 70 80 90 100
200 Fig 315 Sample system dynamics output graph
structure but bounded by interactions and an external environment (Kossiakoff amp Sweet 2003 ) SCD are related to DFD they both show interactions among systems and actors They are often used in the initial phases of problem analysis in order to build consent between stakeholders The building blocks of context diagrams include labeled box and relationships
To describe fl ow of data in a graphical representation DFD is used (Stevens Myers amp Constantine 1979 ) DFDs donrsquot provide information about sequence of operations or timing DFDs are different from fl ow charts since the latter describe fl ow of control in a situation However unlike DFDs fl ow charts donrsquot show the details of data that is fl owing in the situation (Stevens et al 1979 ) On a DFD data items fl ow from an external data source or an internal data store to an internal data store or an external data sink via an internal process
3 Methods and Models
64
Fig 316 System dynamics sample code
376 Unifi ed Modeling Language
UML is a general-purpose modeling language that is a widely accepted industry standard created and managed by the Object Management Group for Software Engineering problems ( UML 20 ) UML is comprised of a set of graphical notation
NA Behkami and TU Daim
65
techniques to create model of software systems UML offers a standard means to illustrate structural and behavior components of system artifacts including actors process components activities database schemas and more UML builds on the notations of the Booch method object modeling technique (OMT) and object- oriented software engineering (OOSE) and effectively combines 1-dimensional tra-ditional workfl ow and datafl ow diagrams into much richer yet condensed and concrete graphical diagrams and models Although UML is a widely accepted stan-dard it has been criticized for standard bloat and being diffi cult to learn and linguis-tically incoherent (Henderson-Sellers amp Gonzalez-Perez 2006 Meyer 1997 )
Using UML two different views of a situation can be represented using static and behavioral types of diagrams Static (or structural) views describe the fi xed struc-ture of the system using objects attributes operations and relationships Dynamic (or behavioral) views describe the fl uid and changing behavior of the situation by documenting collaborations among objects and changes to their internal states
3761 Structural Diagrams
The set of diagrams listed here describe the elements that are in the system being modeled ( Unifi ed Modeling LanguagemdashWikipedia the free encyclopedia )
bull Class diagram describes the structure of a system by showing the systemrsquos classes their attributes and the relationships among the classes
bull Component diagram depicts how a software system is split up into compo-nents and shows the dependencies among these components
bull Composite structure diagram describes the internal structure of a class and the collaborations that this structure makes possible
bull Deployment diagram serves to model the hardware used in system implemen-tations and the execution environments and artifacts deployed on the hardware
bull Object diagram shows a complete or partial view of the structure of a modeled system at a specifi c time
bull Package diagram depicts how a system is split up into logical groupings by showing the dependencies among these groupings
bull Profi le diagram operates at the metamodel level to show stereotypes as classes with the ltltstereotypegtgt stereotype and profi les as packages with the ltltpro-fi legtgt stereotype The extension relation (solid line with closed fi lled arrow-head) indicates what metamodel element a given stereotype is extending
3762 Behavioral Diagrams
These sets of diagrams listed here illustrate the things that happen in the system thatrsquos being modeled ( Unifi ed Modeling LanguagemdashWikipedia the free encyclopedia )
bull Activity diagram represents the business and operational step-by-step workfl ows of components in a system An activity diagram shows the overall fl ow of control
3 Methods and Models
66
bull State machine diagram standardized notation to describe many systems from computer programs to business processes
bull Use case diagram shows the functionality provided by a system in terms of actors their goals represented as use cases and any dependencies among those use cases
bull Communication diagram shows the interactions between objects or parts in terms of sequenced messages They represent a combination of information taken from class sequence and use case diagrams describing both the static structure and dynamic behavior of a system
bull Interaction overview diagram is a type of activity diagram in which the nodes represent interaction diagrams
bull Sequence diagram shows how objects communicate with each other in terms of a sequence of messages Also indicates the life-spans of objects relative to those messages
bull Timing diagrams are specifi c types of interaction diagram where the focus is on timing constraints
377 SysML
For modeling system engineering application SysML is a general-purpose model-ing language It can be used for specifi cation analysis design verifi cation and vali-dation of a variety of systems SysML is developed as an extension of the UML
The main standard for SysML is maintained by the OMG group which also man-ages the UML standard ( OMG SysML ) Figure 338 shows the four pillars of SysML Several modeling tool vendors offer SysML support Improvements over UML that are of importance to system engineers include the following ( SysML ForummdashSysML FAQ ) SysML is a smaller language that is easier to learn and use SysML model management components support views (compliant with IEEE-Std- 1471-2000 Recommended Practice for Architectural Description of Software Intensive Systems) and SysML semantics are more fl exible and less software centric as the ones in UML
38 Conclusions for Modeling Methodologies to Use
After reviewing the candidate methodologies as described in the previous sections the matrix in Fig 317 was generated This matrix shows the needs for modeling as rows and lists the candidate methodologies across the top The intersections of a need and methodology (each cell) are then rated for usefulness (fi t for modeling purpose) In conclusion the only method that was capable of mathematical simulation was system dynamics And the only method capable of adequately separating and model-ing the dynamic and static aspects of the problem was UML
NA Behkami and TU Daim
67
39 Qualitative Research Grounded Theory and UML
391 An Overview of Qualitative Research
The difference between qualitative and quantitative research is man selecting the appropriate methodology depends on the objectives and preferences of the researcher Largely selecting qualitative or quantitative depends on the variables of available time familiarity with research topic access to interview subjects and data research data consumer preference and relationship of researcher to study subjects (Hancock amp Algozzine 2006 )
Quantitative methods can be appropriate when resources and time are limited Since these methods use instruments such as surveys to quickly gather specifi c vari-ables from large groups of people for example political preferences these instru-ments can produce meaningful data in a short amount of time even for small investments However for collecting data qualitative methods require individual interviews observations or focus groups which require a considerable investment in time and resources to adequately represent the domain being studied
In case little is known about a situation qualitative research is a good starting methodology since it attempts to investigate a large number of factors that may be infl uencing a situation However quantitative methods typically investigate the impact of just a few variables For example often a holistic qualitative approach can investigate an array of variables about a problem and later serve as a starting point for a comparative quantitative study
Quantitative research can often be performed with minimal involvement from participants In case access to study subject is diffi cult a quantitative approach is pre-ferred In distinction diffi culties of delays in access to participants for observations or focus group and types of qualitative research could slow down the researcher efforts
Fig 317 Methodology selection matrix
3 Methods and Models
68
Another important factor in considering qualitative or quantitative method is the preference of the consumer of the research results If the potential consumers of research fi nding prefer words and themes to numbers and graphs a qualitative approach would be better suited On the other hand for example a policy setting committee may need and prefer quantifi able data about a community rather than feelings and explain for general policy setting purposes
Finally in qualitative study itrsquos the goal to understand the situation from the insider perspective (the participants) and not from the researcher perspective However in qualitative researcher to maintain objectivity often it is sought to remain blind to the experimental conditions to avoid infl uences of variables being investigated
We can conclude from the reasoning about qualitative and quantitative approaches that they differ in many ways They are each appropriate for certain situation and nei-ther is right or wrong even in some cases researchers combine the activities of both qualitative and quantitative in their research efforts (Hancock amp Algozzine 2006 )
Since this proposal for HIT diffusion is proposing a mainly qualitative method apparent from the reasons above and nature of the problem being studied the rest of the discussion will focus on the qualitative methods There are various fl avors of qualitative research and while they share common characteristics differences among them exist (Creswell 2006 ) Table 35 presents a comparison of general research traditions and fi ve of these major types are important to highlight (Hancock amp Algozzine 2006 )
392 Grounded Theory and Case Study Method Defi nitions
Grounded theory (GT) and case study method are often used independently or together to study social and technological systems In order to select the appropriate methodology and especially for this proposed HIT diffusion research itrsquos important to understand the defi nition of GT and case study They both have been used in conjunction with UML to study information systems among others
Case study method can be used to study one or more cases in detail and its fundamental research question is the following ldquoWhat are the characteristics of this single case or of these comparison casesrdquo (Johnson amp Christensen 2004 ) A case study is often bounded by a person a group or an activity and is interdisciplinary Once classifi cation of case study types includes the following (Stake 1995 )
1 Intrinsic case studymdashonly to understand a particular case 2 Instrumental case studymdashto understand something at a more general level than
the case 3 Collective case studymdashstudying and comparing multiple cases in a single
research study
In a case study approach for data collection multiple methods such as interviews and observations can be used The fi nal output of a case study is a rich and compre-hensive description of the case and its environment
NA Behkami and TU Daim
69
Where case study is detailed account and analysis of one or more cases grounded theory is developed inductively and bottom-up GTrsquos fundamental research question is the following ldquoWhat theory or explanation emerges from an analysis of the data collected about this phenomenonrdquo (Johnson amp Christensen 2004 ) Grounded the-ory is usually used to generate theory and it can also be used to evaluate previously grounded theories The following are important characteristics of a grounded theory (Johnson amp Christensen 2004 )
bull Fit (ie Does the theory correspond to real-world data) bull Understanding (ie Is the theory clear and understandable) bull Generality (ie Is the theory abstract enough to move beyond the specifi cs in the
original research study) bull Control (ie Can the theory be applied to produce real-world results)
Table 35 Research methodology summary (Hancock amp Algozzine 2006 )
Quantitative studies Qualitative studies Case studies
Researcher identifi es topic or question(s) of interest and selects participants and arranges procedures that provide answers that are accepted with predetermined degree of confi dence research questions are often stated in hypotheses that are accepted or rejected using statistical test and analyses
Researcher identifi es topic or question(s) of interest collects information from a variety of sources often as a participant observer and accepts the analytical task as one of discovering answers that emerge from information that is available as a result of the study
Research identifi es topic or question(s) of interest determines appropriate unit to represent it and defi nes what is known based on careful analysis of multiple sources of information of the ldquocaserdquo
Research process may vary greatly from context being investigated (eg survey of how principals spend their time) or appropriately refl ect it (eg observation of how principals spend their time)
Research process is designed to refl ect as much as possible the natural ongoing context being investigated information is often gathered by participant observers (individuals actively engaged immersed or involved in the information collection setting or activity)
Research process is defi ned by systematic series of steps designed to provide careful analysis of the case
Information collection may last a few hours or a few days but generally is of short-term duration using carefully constructed measures designed specifi cally to generate valid and reliable information under the conditions of the study
Information collection may last a few months or as long as it takes for an adequate answer to emerge the time frame for the study is often not defi ned at the time the research is undertaken
Information collection may last a few hours a few days a few months or as long as is necessary to adequately ldquodefi nerdquo the case
Report of the outcomes of the process is generally expository consisting of a series of statistical answers to questions under investigation
Report of outcomes of the process is generally narrative consisting of a series of ldquopages to the storyrdquo or ldquochapters to the bookrdquo
Report of outcomes of the process is generally narrative in nature consisting of a series of illustrative descriptions of key aspects of the case
3 Methods and Models
70
In grounded theory data analysis includes three steps
1 Open coding read transcripts and code themes emerging from data 2 Axial coding organize discovered themes into groupings 3 Selective coding focus on main themes and story development
In a grounded theory approach when no more new themes emerge from data theoretical saturation has been achieved and the fi nal report will include a detailed description of the grounded theory
393 Using Grounded Theory and Case Study Together
Grounded theory is a general method of analysis that can accept quantitative quali-tative or hybrid data (Glaser 1978 ) however it has mainly been used for qualitative researcher (Glaser 2001 ) When using grounded theory and case study together care has to be taken as principles of case study research do not interfere with the emergence of theory in grounded theory (Glaser 1998 ) As Hart ( 2005 ) points out Yin ( 1994 ) states ldquotheory development prior to the collection of any case study data is an essential step in doing case studiesrdquo While Yinrsquos statement is valid for some types of case study research it violates the key principle of open-mindedness (no theory before start) that is in grounded theory Therefore when combining grounded theory and case study the researcher has to explicitly mention which method is driv-ing the investigative research
Supporting the close relationship of GT and case study Hart ( 2005 ) in his own research found that reasons for using grounded theory were consistent with reasons for using case study research set forth (Benbasat Goldstein amp Mead 1987 Hart 2005 )
1) the research can study IS in a natural setting learn the state of the art and generate theo-ries from practice
2) The researcher can answer the questions that lead to an understanding of the nature and complexity of the processes taking place
3) It is an appropriate way to research a previously little studied area
Various researchers have identifi ed generated theory grounded in case study data as a preferred method (Eisenhardt 1989 Lehmann 2001 Maznevski amp Chudoba 2000 Orlikowski 1993 Urquhart 2001 ) Cheryl Chi calls combing grounded the-ory and case studies a ldquotheory building case studyrdquo ( Chi Method-Case Study vs Grounded Theory ) and Eisenhardt ( 1989 ) identifi es the following strength for using case data to build grounded theories
1 Theory building from case studies is likely to produce novel theory this is so because ldquocreative insight often arises from juxtaposition of contradictory or par-adoxical evidencerdquo (p 546) The process of reconciling these accounts using the constant comparative method forces the analyst to a new gestalt unfreezing thinking and producing ldquotheory with less researcher bias than theory built from incremental studies or armchair axiomatic deductionrdquo (p 546)
NA Behkami and TU Daim
71
2 The emergent theory ldquois likely to be testable with constructs that can be readily measured and hypotheses that can be proven falserdquo (p 547) Due to the close connection between theory and data it is likely that the theory can be further tested and expanded by subsequent studies
3 The ldquoresultant theory is likely to be empirically validrdquo (p 547) This is so because a level of validation is performed implicitly by constant comparison questioning the data from the start of the process ldquoThis closeness can lead to an intimate sense of thingsrsquo that lsquooften produces theory which closely mirrors realityrdquo (p 547) [4]
394 Grounded Theory in Information Systems (IS) and Systems Thinking Research
While application of grounded theory in information science (IS) is relatively recent scientists in social science have been using grounded theory method (GTM) for about 40 years The growth of GT in IS while being successful however has miscon-ceptions and misunderstanding associated with it A paper by Orlikowski which was the winner of the MIS Quarterly Best Paper Award for 1993 is a seminal example of grounded theory in information systems (Orlikowski 1993 ) Grounded theory enabled Orlikowski to focus on actions and important stakeholders associated with organizational change Others have published research using grounded theory in IS (Baskerville amp Pries-Heje 1999 Lehmann 2001 Maznevski amp Chudoba 2000 Trauth amp Jessup 2000 Urquhart et al 2001 Zenobia 2008 ) but the appliers still remain in the minority (Lehmann 2001 ) While adoption of grounded theory increases there remains a shortage on how to apply it correctly in IS and one paper tried to contribute as shown in the next fi gure (Lehmann 2001 ) and highlighted the following for GT and IS that need more guidance ldquo(a) describing the use of the grounded theory method with case study data (b) presenting a research model (c) discussing the critical characteristics of the grounded theory method (d) discussing why grounded theory is appropriate for studies seeking both rigor and relevance and (e) highlighting some risks and demands intrinsic to the methodrdquo
In IS research grounded theory has been used to investigate infl uence of systems thinking on the practice of information system practitioners (Goede amp Villiers 2003 ) As discussed by Strauss and Corbin (Strauss amp Corbin 1998 ) qualitative research can be seen as an interpretive research Using the proposed seven princi-ples of interpretive fi eld research summarized (Klein amp Myers 1999 ) one IS study used ldquoGrounded Theory as proposed in this study is used to fulfi ll the fourth of the seven principles The aim is to develop a theory on how IS practitioners unknow-ingly use systems thinking techniques in their work that can be generalized in simi-lar situations Other techniques to fulfi ll this principle include Actor Network Theory and the Hermeneutic processrdquo (Goede amp Villiers 2003 )
Another study examined applying GTM to derive enterprise system require-ments (Chakraborty amp Dehlinger 2009 ) This application was driven by the need for initial design and system architecture to be aligned The paper proposed using
3 Methods and Models
72
grounded theory to extract functional and nonfunctional enterprise requirements from system description They stated that a qualitative data analysis technique GTM could be used to interpret requirements for a software system Their use of GTM generated enterprise requirements and resulted in system model in UML The use of GTM in that study had the following contributions
bull Presents a structured qualitative analysis method to identify enterprise requirements
bull Provides a basis to verify enterprise requirements via high-level EA objectives bull Allows for the representation of business strategy in a requirements engineering
context bull Enables the traceability of EA objectives in the requirements engineering and
design phases
Yet another study analyzed Object-Oriented Analysis amp Design (OOAampD) as a representative of information systems development methodologies (ISDMs) and grounded theory (GT) as a representative of research methods ( What Could OOAampD Benefi t From Gounded Theory ) where ldquoThe basic assumption is that both the research and systems development process are knowledge acquisition processes where methods are used which guide the work of acquiring knowledgerdquo The reason for the study was because the researchers felt that there were both similarities and dissimilarities between the OOAampD and GT and wanted to see how one could ben-efi t from using them together An example of dissimilarity is that GT focuses on describing people and their actions while OOAampD focuses on how IS is used to support people with information Another difference is that OOAampD has a design (of a system) purpose where GT is for understanding and theory building ldquoOn a basic level both research methods and ISDMs are support for asking good questions and presenting good answers in order to acquire knowledgerdquo
395 Criticisms of Grounded Theory
Various researchers have criticized grounded theory The earliest riff is a contro-versy that developed among the originators Strauss has further developed GT (Strauss amp Corbin 1998 ) while Glaser ( 1992 ) criticized this version for violating basic principles Others have proposed a newer multi-GTM that would integrate empirical grounding theoretical grounding and internal grounding (Goldkuhl amp Cronholm 2003 )
Other problems with GT include how to deal with large amounts of data since there is no explicit support for where to start the analysis (Goldkuhl amp Cronholm 2003 ) The open-mindedness in the data collection phase can lead to meaninglessly diverging amount of data (Goldkuhl amp Cronholm 2003 ) Another is that GT practi-tioners are advised to discard pre-assumptions they hold so the real nature of the study fi eld comes out GT researchers are encouraged to avoid reading literature until the completion of the study (Rennie Phillips amp Quartaro 1988 ) Ignoring
NA Behkami and TU Daim
73
existing theory can lead to duplicating effort for theories or constructs already discovered elsewhere (Goldkuhl amp Cronholm 2003 ) Lack of adequate illustration technique is yet another weakness of GT (Goldkuhl amp Cronholm 2003 )
396 Current State of UML as a Research Tool and Criticisms
Current issues in UML research concern with the extent and nature of UML use and UML usability One study found that the use of UML by practitioners varies and non-IT professionals are involved in the development of UML diagrams (Dobing amp Parsons 2005 ) The study concluded that the variation in use was contrary to the idea that UML is a ldquounifi edrdquo language
Another study while acknowledging the popularity that UML has gained in sys-tem engineering felt ldquoit is not fulfi lling its promiserdquo (Batra 2009 ) Others have stated that UML is too big and complicated (Siau amp Cao 2001 ) suffers from vague semantics (Evermann amp Wand 2006 ) and steep learning curve (Siau amp Loo 2006 ) and doesnrsquot allow for easy interchange between diagrams and models At a higher level some have highlighted that it is diffi cult to model a correct and reliable appli-cation using UML and to understand such a specifi cation (Peleg amp Dori 2000 ) Others have claimed that UML is low in usability because it requires multiple models to completely specify a system (Dori 2002 ) and have proposed another methodology namely the object process methodology (OPM) (Dori 2001 )
397 To UML or Not to UML
The emergence of UML has provided an accessible visualization of models which facilitates communication of ideas But as one research study found out UML lacks formal precise semantics and they used the B Language to supplement UML for their need (Snook amp Butler 2006 ) The B language is a state model-based formal specifi cation notation (Abrial 1996 ) But when the clients of the research study found the B Language artifacts hard to understand they asked the research team ldquocouldnrsquot you use UMLrdquo (Amey 2999 )
398 An Actual Example of Using Grounded Theory in Conjunction with UML
A study used the hierarchical coding procedure offered by GTM with UML to create the requirements for an organizationrsquos enterprise application Figure 318 summa-rizes the coding procedures of GTM that were incorporated into the requirements
3 Methods and Models
74
engineering process for the enterprise application (Chakraborty amp Dehlinger 2009 ) For this example the study chose a ldquohigh-level description for a university support system comprising of a student record management system (SRMS) a laboratory management system a course submission system and an admission management sys-temrdquo (Sommerville 2000 ) Recall from earlier sections that grounded theory coding processes are done in three steps of open coding axial coding and selective coding
3981 Open Coding
In this step the transcript of interview or case is read line by line The text is broken down into concepts Concepts are any part of textual description that the researchers believe are descriptive of the system being studied Table 36 shows the concepts extracted after this study applied GTM to a subsystem of the university support system (SRMS) The preliminary concepts are highlighted in bold The open coding led to the identifi cation of other supporting information as expressed in UML shown in Fig 319
3982 Axial Coding
The goal of this step is to organize the concepts identifi ed during open coding into a hierarchical relationship First the higher order categories are sorted out and later sub-categories add more descriptive information The process is continued until all
Fig 318 Categories for SRMS (Chakraborty amp Dehlinger 2009 )
NA Behkami and TU Daim
75
Subsystem
-Student record
Management system
System functionality
-usabilityrequirements
Querying Mechanism Summary reports
Users
-Computational Skill
Student
-Personal Details-course grade
Classescourses
-Courses Name-
Data Item
-Name-Type
Implementation technique
-Database language
-VisualBasic
User Interfaces
Fig 319 Axial coding-description of the SRMS (Chakraborty amp Dehlinger 2009 )
Table 36 Concept extraction (Chakraborty amp Dehlinger 2009 )
Subsystem descriptionmdashStudent record system
The aim of this project is to maintain a student record system maintaining student records within a university or college department The system should allow personal details to be recorded as well as classes taken grades etc It shall provide summary facilities giving information about groups of students to be retrieved Assume that the system is intended for use by departmental administrative staff with no computing background This project may be implemented in a database language or in a language such as Visual Basic
categories have been associated Figure 320 shows the result of this process expressed in UML
3983 Selective Coding
The pervious step of axial coding has provided description for each of the subsys-tems present in the problem space Selective coding integrates the categories and descriptions from the individual subsystems into an overall description of the sys-tem Figure 41 shows this fi nal description derived from grounded theory and pre-sented with UML
3 Methods and Models
76
References
ldquoBasic Flow Chart Samplerdquo ldquoNDE Project Managementrdquo ldquoOMG SysMLrdquo ldquoSysML ForummdashSysML FAQrdquo ldquoUML 20rdquo
Fig 320 System description after selective coding (Chakraborty amp Dehlinger 2009 )
NA Behkami and TU Daim
77
ldquoWhat Could OOAampD Benefi t From Gounded Theoryrdquo ldquoData fl ow diagrammdashWikipedia the free encyclopediardquo ldquoUnifi ed Modeling LanguagemdashWikipedia the free encyclopediardquo ldquoUnifi ed Modeling LanguagemdashWikipedia the free encyclopediardquo Abrahamson E (1991) Managerial fads and fashions The diffusion and refection of innovations
Academy of Management Review 16 586ndash612 Abrahamson E (1996) Management fashion Academy of Management Review 21 254ndash285 Abrahamson E amp Fombrun C J (1994) Macrocultures Determinants and consequences
Academy of Management Review 19 728ndash755 Abrahamson E amp Rosenkopf L (1993) Institutional and competitive bandwagons Using math-
ematical modeling as a tool to explore innovation diffusion Academy of Management Review 18 487ndash517
Abrial J R (1996) The B-book assigning programs to meanings Cambridge Univ Press Adams D A Nelson R R amp Todd P A (1992) Perceived usefulness ease of use and usage
of information technology A replication MIS Quarterly 16 227ndash247 Ajzen I (1985) ldquoFrom intentions to actions A theory of planned behavior SSSP Springer Series
in Social Psychology (pp 11ndash39) New York NY Springer Ajzen I (1991) The theory of planned behavior Organizational Behavior and Human Decision
Processes 50 179ndash211 Ajzen I amp Fishbein M (1973) Attitudinal and normative variables as predictors of specifi c
behaviors Journal of Personality and Social Psychology 27 41ndash57 Ambler S W (2004) The object primer Agile model-driven development with UML 20
Cambridge University Press Amey P Dear sir Yours faithfully An everyday story of formality Proc 12th Safety-Critical
Systems Symposium pp 3ndash18 Amit R amp Schoemaker P J (1993) Strategic assets and organizational rent Strategic
Management Journal 14 33ndash46 Bain J S (1956) Barriers to new competition Cambridge Harvard Univ Press Barney J B (1986) Strategic factor markets Expectations luck and business strategy
Management Science 32 1231ndash1241 Barney J (1991) Special theory forum The resource-based model of the fi rm Origins implica-
tions and prospects Journal of Management 17 97ndash98 Barney J B amp Clark D N (2007) Resource-based theory Creating and sustaining competitive
advantage Oxford Oxford University Press Baskerville R amp Pries-Heje J (1999) Grounded action research A method for understanding IT
in practice Accounting Management and Information Technologies 9 1ndash23 Batra D (2009) Unifi ed modeling language (UML) topics Cognitive issues in UML research
Journal of Database Management Behkami N A (2009) Diffusion of Innovation (Healthcare IT)--System Dynamics Portland State
University Department of Engineering amp Technology Management Working Paper Series Benbasat I Goldstein D K amp Mead M (1987) The case research strategy in studies of infor-
mation systems MIS quarterly 369ndash386 Chakraborty S amp Dehlinger J (2009) Applying the Grounded Theory Method to Derive
Enterprise System Requirements Software Engineering Artifi cial Intelligence Networking and ParallelDistributed Computing ACIS International Conference on Los Alamitos CA USA IEEE Computer Society 2009 pp 333ndash338
Checkland P (1999) Systems thinking systems practice Includes a 30-year retrospective Wiley Checkland P Scholes J (1990) Soft systems methodology in action John Wiley amp Sons Ltd
(Import) Chi C Method-Case Study vs Grounded Theory Chuttur M (2009) Overview of the technology acceptance model Origins developments and
future directions
3 Methods and Models
78
Collan M Teacutetard F (2007) Lazy user theory of solution selection Proceedings or the CELDA 2007 conference pp 7ndash9
Collis D J amp Montgomery C A (1995) Competing on resources Strategy in the 1990s Knowledge and Strategy 25ndash40
Cooper R B amp Zmud R W (1990) Information technology implementation research A tech-nological diffusion approach Management Science 36 123ndash139
Cousins J B amp Simon M (1996) The nature and impact of policy-induced partnerships between research and practice communities Educational Evaluation and Policy Analysis 18 (Autumn) 199ndash218
Creswell J W (2006) Qualitative inquiry and research design Choosing among fi ve approaches Sage Publications Inc
Damanpour F (1988) Innovation type radicalness and the adoption process Communication Research 15 545ndash567
Damanpour F (1991) Organizational innovation A meta-analysis of effects of determinants and moderators Academy of Management Journal 34 555ndash590
Damanpour F amp Evan W M (1984) Organizational innovation and performance The problem of ldquoorganizational lagrdquo Administrative Science Quarterly 29 392ndash409
Data Flow DiagrammdashSSADM DiagramsmdashSmartDraw Tutorials Davis F D (1985) A technology acceptance model for empirically testing new end-user informa-
tion systems Theory and results Cambridge MA Massachusetts Institute of Technology Sloan School of Management
Davis F D (1989) Perceived usefulness perceived ease of use and user acceptance of informa-tion technology MIS Quarterly 13 319ndash340
Davis F D Bagozzi R P amp Warshaw P R (1989) User acceptance of computer technology A comparison of two theoretical models Management Science 35 982ndash1003
Demsetz H (1973) Industry structure market rivalry and public policy Journal of Law and eco-nomics 16 1ndash9
Dierickx I amp Cool K (1989) Asset stock accumulation and sustainability of competitive advan-tage Management Science 1504ndash1511
Dobing B amp Parsons J (2005) Current practices in the use of UML Perspectives in Conceptual Modeling 2ndash11
Dori D (2001) Object-process methodology applied to modeling credit card transactions Journal of Database Management 12 4ndash14
Dori D (2002) Why signifi cant UML change is unlikely Eagly A H amp Chaiken S (1993) The psychology of attitudes Fort Worth TX Harcourt Brace
Jovanovich College Publishers Fort Worth Eisenhardt K M (1989) Building theories from case study research Academy of Management
Review 532ndash550 Erdil N amp Emerson C R (2008) Modeling the dynamics of electronic health records adoption
in the us healthcare system Proceedings of the 26th international conference of the system dynamics society 2008
Evermann J amp Wand Y (2006) Ontological modeling rules for UML An empirical assessment Journal of Computer Information Systems 46 14
Finegan A D (2003) Wicked problems organizational complexity and knowledge manage-mentndasha systems approach The International Journal of Knowledge Culture and Change Management 3
Fishbein M (1967) Attitude and the prediction of behavior Readings in attitude theory and mea-surement 477ndash492
Fishbein M Ajzen I (1975) Belief attitude intention and behavior An introduction to theory and research
Forrester J W (1994) System dynamics systems thinking and soft OR System Glaser B G (1978) Theoretical sensitivity Advances in the methodology of grounded theory
Sociology Press
NA Behkami and TU Daim
79
Glaser B G (1992) Basics of grounded theory analysis Emergence vs forcing Mill Valley CA Sociology Press
Glaser B G (1998) Doing grounded theory Issues and discussions Mill Valley CA Sociology Press
Glaser B G (2001) The grounded theory perspective Conceptualization contrasted with descrip-tion Sociology Press
Goede R amp Villiers C D (2003) The applicability of grounded theory as research methodology in studies on the use of methodologies in IS practices Proceedings of the 2003 annual research conference of the South African institute of computer scientists and information technologists on Enablement through technology South African Institute for Computer Scientists and Information Technologists 2003 pp 208ndash217
Goldkuhl G amp Cronholm S (2003) Multi-grounded theoryndashAdding theoretical grounding to grounded theory European conference on research methodology for business and management studies p 177
Grant R M (1991) The resource-based theory of competitive advantage Implications for strat-egy formulation California Management Review 33 114ndash35
Grant R M (1996) Toward a knowledge-based theory of the fi rm Strategic Management Journal 17 109ndash122
Hall R (1992) The strategic analysis of intangible resources Strategic Management Journal 135ndash144
Hamel G amp Prahalad C K (1990) The core competence of the corporation Harvard Business Review 68 79ndash91
Hancock D R amp Algozzine R (2006) Doing case study research A practical guide for begin-ning researchers Teachers College Press
Hart D N (2005) Information systems foundations ANU E Press Henderson-Sellers B amp Gonzalez-Perez C (2006) Uses and Abuses of the Stereotype
Mechanism in UML 1x and 20 Model Driven Engineering Languages and Systems 16ndash26 Hendrickson A R Massey P D amp Cronan T P (1993) On the test-retest reliability of per-
ceived usefulness and perceived ease of use scales MIS Quarterly 17 227ndash230 Hitt M A Hoskisson R E amp Kim H (1997) International diversifi cation Effects on innova-
tion and fi rm performance in product-diversifi ed fi rms Academy of Management Journal 767ndash798
Hitt M A amp Ireland R D (1985a) Strategy contextual factors and performance Human Relations 38 793
Hitt M A amp Ireland R D (1985b) Corporate distinctive competence strategy industry and performance Strategic Management Journal 6 273ndash293
Hitt M A amp Ireland R D (1986) Relationships among corporate level distinctive competen-cies diversifi cation strategy corporate structure and performance Journal of Management Studies 23 0022ndash2380
Hrebiniak L G amp Snow C C (1982) Top-management agreement and organizational perfor-mance Human Relations 35 1139
Itami H amp Roehl T (1987) Mobilizing intangible assets Cambridge MA Johnson B amp Christensen L B (2004) Educational research Quantitative qualitative and
mixed approaches Research Edition Second Edition Allyn amp Bacon Klein H K amp Myers M D (1999) A set of principles for conducting and evaluating interpretive
fi eld studies in information systems MIS Quarterly 67ndash93 Kossiakoff A amp Sweet W N (2003) Systems engineering Wiley-IEEE Learned E Christensen C Andrews K amp Guth W (1969) Business policy Text and casesrsquo
Homewood IL Richard D Irwin Inc Lehmann H (2001) Using grounded theory with technology cases Distilling critical theory from
a multinational information systems development project Journal of Global Information Technology Management 4 45ndash60
3 Methods and Models
80
Liebeskind J P (1996) Knowledge strategy and the theory of the fi rm Strategic Management Journal 17 93ndash107
Maznevski M L amp Chudoba K M (2000) Bridging space over time Global virtual team dynamics and effectiveness Organization Science 473ndash492
Meyer B (1997) UML The positive spin Cutter IT Journal x Nelson R R amp Winter S G (1982) An evolutionary theory of economic change Belknap Press Nutley S amp Davies H T O (2000) Making a reality of evidence-based practice some lessons
from the diffusion of innovations Public Money amp Management 20 35 ONeill H M Pouder R W amp Buchholtz A K (1998) Patterns in the diffusion of strategies
across organizations Insights from the innovation diffusion literature Academy of Management Review 23 98ndash114
Orlikowski W J (1993) CASE tools as organizational change Investigating incremental and radical changes in systems development MIS Quarterly 309ndash340
Osborne S P (1998) Naming the beast Defi ning and classifying service innovations in social policy Human Relations 51 1133ndash1154
Otto P amp Simon M (2009) Coordinating quality care A policy model to simulate adoption of EHR Proceedings of the 26th international system dynamics conference Albuquerque 2009
Peleg M amp Dori D (2000) The model multiplicity problem Experimenting with real-time specifi cation methods IEEE Transactions on Software Engineering 26 742ndash759
Penrose E (1959) The theory of the growth of the fi rm New York NY Wiley Porter M E (1981) The contributions of industrial organization to strategic management The
Academy of Management Review 6 609ndash620 Porter M E (1985) Competitive advantage Competitive advantage Creating and sustaining
superior performance New York NY Porter Michael E (1979) How competitive forces shape strategy Harvard Business Review 57
137ndash145 Prahalad C K amp Bettis R A (1986) The dominant logic A new linkage between diversity and
performance Strategic Management Journal 485ndash501 Rennie D L Phillips J R amp Quartaro G K (1988) Grounded theory A promising approach
to conceptualization in psychology Canadian Psychology 29 139ndash150 Ricardo D (1817) The principles of political economy and taxation (1817) The Works and
Correspondence of David Ricardo hrsg v Sraffa Piero Bd I Cambridge Robinson K Berrisford G (1994) Object oriented SSADM Prentice Hall PTR Rumelt R P amp Lamb R (1984) Competitive strategic management Toward a Strategic Theory
of the Firm 556ndash570 Scherer M J (2002) Assistive technology Matching device and consumer for successful rehabili-
tation Washington DC APA Books Segars A H amp Grover V (1993) Re-examining perceived ease of use and usefulness A confi r-
matory factor analysis MIS Quarterly 17 517ndash525 Siau K amp Cao Q (2001) Unifi ed modeling language A complexity analysis Journal of
Database Management 12 26ndash34 Siau K amp Loo P P (2006) Identifying diffi culties in learning UML Information Systems
Management 23 43ndash51 Smart P A Maddern H amp Maull R S (2008) Understanding business process management
Implications for theory and practice Snook C amp Butler M (2006) UML-B Formal modeling and design aided by UML ACM
Transactions on Software Engineering and Methodology (TOSEM) 15 122 Sommerville I (2000) Software engineering Addison Wesley Spender J C amp Grant R M (1996) Knowledge and the fi rm Overview Strategic Management
Journal 17 5ndash9 SSADM Diagram SoftwaremdashStructured Systems Analysis and Design Methodology Stake D R E (1995) The art of case study research Sage Publications Inc Stalk G Evans P amp Shulman L E (1992) Competing on capabilities The new rules of corpo-
rate strategy Harvard Business Review
NA Behkami and TU Daim
81
Stevens W Myers G amp Constantine L (1979) Structured design Classics in software engi-neering Yourdon Press 205ndash232
Strauss A L Corbin J M (1998) Basics of qualitative research Techniques and procedures for developing grounded theory Sage Pubns
Subramanian G H (1994) A replication of perceived usefulness and perceived ease of use mea-surement Decision Sciences 25 863ndash863
Szajna B (1994) Software evaluation and choice Predictive validation of the technology accep-tance instrument MIS Quarterly 18 319ndash324
Teece D J (1980) Economy of scope and the scope of the enterprise Journal of Economic Behavior and Organization 1 223ndash247
Teece D J Pisano G amp Shuen A (1997) Dynamic capabilities and strategic management Strategic Management Journal 18 509ndash533
Tetard F amp Collan M (1899) Lazy user theory A dynamic model to understand user selection of products and services HICSS (pp 1ndash9) Big Island HI IEEE
Theories Used in IS Research Wiki York University Thompson A A amp Strickland A J (1983) Strategy formulation and implementation Tasks of
the general manager Business Publications Tornatzky L G amp Fleischer M (1990) Processes of technological innovation New York The
Free Press Trauth E M amp Jessup L M (2000) Understanding computer-mediated discussions Positivist
and interpretive analyses of group support system use MIS Quarterly 24 43ndash79 Urquhart C (2001) An encounter with grounded theory Tackling the practical and philosophical
issues Qualitative Research in IS Issues and Trends 104ndash140 van de Water H Schinkel M amp Rozier R (2006) Fields of application of SSM A categoriza-
tion of publications Journal of the Operational Research Society 58 271ndash287 Vandeven A H amp Rogers E M (1988) Innovations and organizations Critical perspectives
Communication Research 15 632ndash651 Venkatesh V Morris M G Davis G B Davis F D DeLone W H McLean E R et al
(2003) User acceptance of information technology Toward a unifi ed view Inform Management 27 425ndash478
Viswanath V Morris M G Davis G B amp Davis F D (2003) User acceptance of information technology Toward a unifi ed view MIS Quarterly 27 425ndash478
WW Wakeland EJ Gallaher LM Macovsky and CA Aktipis ldquoA Comparison of System Dynamics and Agent-Based Simulation Applied to the Study of Cellular Receptor Dynamicsrdquo Proceedings of the Proceedings of the 37th Annual Hawaii International Conference on System Sciences (HICSSrsquo04) mdash Track 3 mdash Volume 3 IEEE Computer Society 2004 p 300862
Wernerfelt B (1984) A resource-based view of the fi rm Strategic Management Journal 171ndash180
Williams B (2005) Soft systems methodology Williamson O E (1975) Markets and hierarchies analysis and antitrust implications Wixom B H amp Todd P A (2005) A theoretical integration of user satisfaction and technology
acceptance Information Systems Research 16 85ndash102 Wolfe R A (1994) Organizational innovation Review critique and suggested research Journal
of Management Studies 31 405ndash431 Yin R K (1994) Case study research Design and methods Sage Publications Inc Zenobia B (2008) A grounded agent model of the consumer technology adoption process
Portland State University
3 Methods and Models
83copy Springer International Publishing Switzerland 2016 TU Daim et al Healthcare Technology Innovation Adoption Innovation Technology and Knowledge Management DOI 101007978-3-319-17975-9_4
Chapter 4 Field Test
Nima A Behkami and Tugrul U Daim
41 Introduction and Objective
The purpose of this section is to demonstrate the feasibility of the research proposal and its corresponding components on a small scale The general objec-tives of the feasibility study include demonstrating the larger research objectives and demonstrating that the right mix of theories and methodologies has been con-sidered The small fi eld study was conducted at Oregon Health amp Science University (OHSU) with the Care Management Plus (CMP) Team CMP is a proven health information technology (HIT) application for older adults and chronically ill patients with multiple conditions and the innovation includes soft-ware clinic processes and training
Use of qualitative research-based case study with application of diffusion theory and dynamic capabilities using the Unifi ed Modeling Language (UML) notation is demonstrated in this fi eld study In the following sections data collection analysis results conclusions and limitations of research along with propositions for future research are discussed
N A Behkami Merck Research Laboratories Boston MA USA
T U Daim () Portland State University Portland OR USA e-mail tugruludaimpdxedu
84
42 Background Care Management Plus
421 Signifi cance of the National Healthcare Problem
Today care for patients with complex healthcare needs is in a state of crisis in the USA The aging population lifestyle shifts and environmental factors have led to rapid increases in numbers of patients who suffer from complex illnesses while the healthcare system struggles to adapt Treatment for patients with complex needs succeeds when their needs are known their care is well coordinated and their healthcare team is able to make clinical decisions based on the systematically avail-able evidence Tools such as better health IT systems and robust fi nancial incen-tives can facilitate improved quality of care
Patients suffering from chronic illnesses account for approximately 75 of the nationrsquos healthcare-related expenditures However these patients only receive the appropriate treatment about 50 of the time Inadequacy of care is even more of a problem for patients with multiple chronic illnesses For example a patient on Medicare with fi ve or more illnesses will visit 13 different outpatient physicians and fi ll 50 prescriptions per year (Friedman Jiang Elixhauser amp Segal 2006 ) As the number of a patientsrsquo conditions increases the risk of hospitalizations grows exponentially (Wolff Starfi eld amp Anderson 2002 ) While the transitions between providers and settings increase so does the risk of harm from inadequate informa-tion transfer and reconciliation of treatment plans Such risks are a large part of the reason patients like this account for 40 of all Medicare costs Wolff estimates that a third of these costs may be due to inappropriate variation and failure to coor-dinate and manage care (Wolff et al 2002 ) As costs continue to rise the delivery of care must change to meet these costs Components identifi ed as important include better planning on the part of providers and patientsfamilies both in visits and over time better coordination and communication and increased self-manage-ment of conditions by patients and caregivers (Bodenheimer Wagner amp Grumbach 2002a 2002b )
Two changes to healthcare teams that can provide this systematic approach are nurse-based care management and health information technology (Dorr Wilcox et al 2006 Shojania amp Grimshaw 2005 Shojania et al 2006 ) A meta-analysis for redesign for patients with diabetes showed that nurse care managers and team reorganization were the most successful quality improvement techniques infor-mation technology alone was only moderately successful (Shojania et al 2006 ) A care management model for depression in older adults (who tend to have more complicated depression and concurrent illnesses) demonstrated broad success (Steffens et al 2006 Rubenstein et al 2002 ) Patients with schizophrenia bene-fi tted from care management with HIT using the Medical Informatics Network Tool (Young Mintz Cohen amp Chinman 2004 ) The CMP team and others have shown that reduction in hospitalization visits can occur in models focused on older adults with complex needs (Dorr Brunker Wilcox amp Burns 2006 Counsell et al 2007 )
NA Behkami and TU Daim
85
422 Preliminary CMP Studies at OHSU
The CMP model for primary care developed by researchers at Intermountain Healthcare through funding from the John A Hartford Foundation uses specially trained care managers and tracking software to help clinics better care for patients with complex chronic illness
The model helps the clinical team prioritize healthcare needs and prevent com-plications through structured protocols and it provides tools to assist patients and caregivers to self-manage chronic diseases Specialized information technology includes the care manager tracking database patient summary sheet and messaging systems to help clinicians access care plans receive reminders about best practices and facilitate communication between the healthcare team The initial data from implementing CMP was highly positive and demonstrated improved clinical and economic outcomes The initial seven sites for testing CMP were urban practices comprising six to ten clinicians each These clinics employed full-time nurse care managers who each worked with a panel of around 350 active patients
CMP focuses on two primary areas well-trained care managers embedded in the clinic and IT technology to help them manage patients with chronic illnesses Figure 41 describes the primary aspects of the CMP program Physicians refer patients with complex needs (about 3ndash5 of the population in primary care clinics) into the program The care manager then co-creates a care plan with the patient acts as a guide to help the patient and family meet their goals and facilitates access to necessary resources when the patient or family needs navigation ( OHSU )
CMP couples an ambulatory care team with HIT For seniors with complex needs CMP demonstrated a 20 reduction in mortality a 24 reduction in hospi-talizations and a 15ndash25 reduction in complications from diabetes (Dorr Wilcox et al 2006 Dorr Wilcox Donnelly Burns amp Clayton 2005 ) CMP facilitates use of HIT to establish and track care plans and specifi c patient goals to teach and encourage self-management to measure and improve quality and to manage the complex and interleaving tasks as patients and teams prioritize needs Figure 42 shows the system components of CMP (Behkami 2009a ) Experience from the
Care Management
Care Manager
Technology
Referral-For any condition or need-Focus on certainconditions
-Assess amp Plan-Catalyst-Structure
-Access -Best Practices-Communication
Evaluation-Ongoing with feedback-Based on key process and outcome measures
Fig 41 Components of the care management plus program ( OHSU )
4 Field Test
86
dissemination of CMP in more than 75 clinics across the country has led to a deep understanding of the barriers and benefi ts of such HIT Barriers include the need to integrate systems diffi culty communicating with the entire team and representa-tion of workfl ow
43 Research Design
431 Overview
The chart below shows the steps used in conducting the fi eld study Using a litera-ture review a preliminary framework and model were produced Next data was collected using mix methods and various tools were used for analysis and later validation (Fig 43 )
432 Objectives
Objective 1 Identify some dynamic capabilities needed for successful implementa-tion of HIT ( CMP OHSU ) This is the application area that we will derive cases from to develop the dynamic capabilities based on diffusion framework
Fig 42 CMP system view
NA Behkami and TU Daim
87
Objective 2 Demonstrate that dynamic capabilities theory can be used and how to meaningfully extend diffusion of innovation theory
Objective 3 Use software and system engineering methods including 4 + 1 view for perspectives and UML to demonstrate documentation and analysis
Objective 4 Build and run a small simulation of the DOI theory extension using system dynamics The simulation will be used to demonstrate the validity of the new diffusion framework
433 Methodology and Data Collection
The methodology used for the research design is an exploratory case study The case study method is chosen because the proposed research needs to know ldquohowrdquo and ldquowhyrdquo HIT adoptiondiffusion program has worked (or not) Such questions deal with operational links needing to be traced over time rather than mere frequencies or incidences The next three subsections describe the data collection tools used and the last explains the sampling for the fi eld study
Fig 43 Field study research process overview
4 Field Test
88
4331 Site Readiness Questionnaire
The Site Readiness questionnaire is a custom-built structured questionnaire created by the CMP team at OHSU which is sent to sites (clinics) considering adopting CMP The questionnaire attempts to capture the multiple perspectives of the physi-cian nurse care manger as well as IT professionals Each site that participated in the CMP project founded by the John A Hartford Foundation over the last few years was required to fi le out one of these to be eligible The questionnaire is broken into multiple sections that include clinic goals and barriers for adoption current staffpatients current services offered information technology landscape quality measures used to gage services and other
4332 Expert Discussion Guide (Interview)
To understand the perspective of physicians and care managers a CMP interview guide was used A discussion guide is a semi-structured interview guide that is meant to be fl exible to provide room for discovery of new items while still providing some structure to data collection
Overall interview objectives
bull Understand the usersrsquo daily activities attitudes and values bull Determine physician and nurse use patterns with current care management and
HIT productsprocesses (if any) bull Identify the functional and emotional benefi ts that the user is seeking from a care
management (HIT) product bull Learn about how the usage environment impacts the use and perception of the
product
4333 Survey Instrument IT and Administrative Users Questionnaire
To understand the perspective of IT and administrative users a structured question-naire was used
Overall interview objectives
bull Understand the strategic role of IT in the clinic bull Determine past success or failure of IT implementation at the clinic bull Identify systems and IT implementation capabilities of the clinic bull Learn about how IT can enhance or challenge adoption of a new care manage-
ment product at that clinic
NA Behkami and TU Daim
89
4334 Study Sampling
Readiness Assessment
For the Readiness Assessment sample data from four sites in Oregon and one in California who currently participate in the OHSU CMP trail were reviewed This section provides a brief description of each location and its affi liated organizations
The Oregon clinics are members of the Oregon Rural Practice and Research Network (ORPRN) which is a statewide network of primary care clinicians com-munity partners and academicians dedicated to research into delivery of healthcare to rural residents and research to reduce rural health disparities ORPRN includes 42 rural primary practices which care for over 166000 patients ( ORPRN ) The fol-lowing individual clinics participated in providing data Lincoln City Medical Center Eastern Oregon Medical Associates OHSU Scappoose Family Health Center and Klamath Open Door Family Medicine
The fi fth study participant is HealthCare Partners (HCP) LLC a management service organization that manages and operates medical groups and independent physician networks nationally The organization serves more than 500000 patients of whom more than 100000 are older adults HealthCare Partners Medical Group (HCPMG) has been recognized by health plans and business groups for its medical leadership the high quality of medical care delivered operational effectiveness and high rates of patient satisfaction HCPMG employs 500+ primary care and specialty physicians who care for patients in Los Angeles County and north Orange County California through 40 neighborhood offi ces fi ve urgent care centers two medical spas an ambulatory surgery center and an employer on-site offi ce ( Health Care Partners Medical Group )
Physician Discussion Guide and IT Questionnaire
See Table 41
Table 41 Sampling
Subject Clinic Clinic size
EHR- adoption level
Experience with care management Role at the clinic
Interview 1 Oregon Health amp Science University
Large High High Physician principal investigator
Interview 2 Oregon Health amp Science University
Large High Medium Care management plus program director
Interview 3 Oregon Health amp Science University
Large High Medium Nurse care manager
4 Field Test
90
434 Analysis
Using open coding and focused methods of Thematic Analysis the author created themes from the data (Bailey 2006 ) including recurring patterns topics theories viewpoints and concepts Rogersrsquo diffusion of innovation theory and dynamic capability theory and TAM and adoption barriers and infl uences were used to guide the coding Figure 44 shows the workfl ow used for analysis Figure 45 shows a sample of the coding artifacts created
Fig 44 Analysis workfl ow
NA Behkami and TU Daim
91
435 Results and Discussion
After iterating over the themes that emerged from the collected data I was able to group them into eight categories that affected the HIT diffusion process for CMP They included
bull Needs and drivers bull Barriers bull Outcome measures bull Infl uences bull Capabilities bull Adoption decision bull Adoption success criteria bull Awareness of innovation versus actual adoption timeline
Fig 45 Sample fi eld notes
4 Field Test
92
Based on the extracted constructs a process of the adoption from the clinic per-spective was created as shown in Fig 46 The innovation process seems to start for the clinics based on ldquo Drivers rdquo or ldquo Needs rdquo A driver for example is something like the need to more effi ciently manage clinic workfl ow Eventually these needs drive the clinic to adopt the HIT innovation in this case CMP offered by OHSU Then there are ldquo Barriers rdquo and ldquo Infl uences rdquo which are negative and positive reinforce-ments respectively Barriers can discourage both the ldquo Drivers rdquo and the ldquo Adoption Decision rdquo in a negative way For example lack of funding at the clinic for buying an expensive software system can be an example of a barrier Infl uence reinforces both the ldquo Drivers rdquo and the ldquo Adoption Decision rdquo and itrsquos a positive force For example government reimbursement for using HIT in the form of extra revenue for clinic seems to be an example of a positive infl uence on the HIT adoption process
Another theme that emerged from the data which is directly fed related to the adoption decision is ldquo Adoption Success Criteria rdquo This is how a clinic defi nes whether adopting CMP was successful or not These criteria were either mecha-nisms created by the clinic itself or government- or payer-supported ldquo Outcome Measures rdquo that described adoption goals and the progress towards them In time these ldquo Outcome Measures rdquo can either become barriers or infl uences either for the same adopter or future adopters this is similar to the ldquoconfi rmationrdquo stage that Rogers defi ned in Diffusion of Innovation
In all based on the data collected it was clear that the clinics didnrsquot adopt as soon as they became aware of CMP and once they decided to adopt often they didnrsquot know what to do and how to go about adopting it This is where the theme of ldquo Capabilities rdquo comes to light in the adoption process For example having a nurse that was properly trained and skilled in care management to oversee the program was a capability needed and recommended by OHSU for successful adoption As evident from Fig 46 needing ldquo Capabilities rdquo directly became a factor in the
Fig 46 Clinic workfl ow
NA Behkami and TU Daim
93
ldquo Adoption Decision rdquo and indirectly acted as a ldquo Infl uence rdquo or ldquo Barrier rdquo depending on if the clinic had it (or could get it) or didnrsquot have it (or couldnrsquot get it) And fi nally some combination of identifi able barriers infl uences and capabilities leads to the remaining theme discovering that awareness and actual adoption happen over time ldquo Awareness of Innovation versus Actual Adoption Timeline rdquo
4351 Structural Aspects
CMP Adoption Class Diagram
Based on the interviews I was able to build a structural diagram of the stakeholders and actors involved in the CMP diffusion ecosystem as shown in Fig 49 The nota-tion used for the diagram is a UML class diagram that shows the static aspects of the important objects in the system As seen in Fig 47 each object is represented as a rectangle box In the top section of each rectangle is the name of the object and in the second subsection is the attributes of that object A stakeholder or actor is con-sidered to be a type of an object The arrows between object boxes as in Fig 48 show the relationships among objects Itrsquos worth mentioning that these links donrsquot represent behavior which will be shown using dynamic types of UML diagrams in later sections of this document The lines with an arrow at the end show a general-ization relationship meaning for example as in Fig 48 a physician is a type of provider and so are nurses and institutional providers (clinic) This notation allows us to analyze these objects as part of the whole while keeping their specializations in mind The dotted lines between objects represent a link and not a hierarchical relationship like the other line types (Fig 49 )
Physician-Education-Comfort with Technology-Specialization-Role
Fig 47 Physician object
Provider
Physician
NurseInstituational Provider
-Size-Location-Technology
-Education-Comfort with Technology-Specialization-Role
Fig 48 Provider parent class
4 Field Test
94
CMP Ecosystem Package Diagram
The ecosystem is made up of fi ve major packages of objects as shown in the top part of Fig 410 as a UML component diagram These packages include the provider government innovation supplier care seeker and payer packages Being able to identify and correctly group these objects is useful in studying the diffusionadop-tion process This eventual categorization will be one of the benefi ts and unique contributions of the proposed research HIT diffusion research
4352 Behavioral Aspects
There are a range of activities that occur at the clinic for adoption of CMP which require analysis These include adoption rejection dissemination developing capabilities implementation usage reconfi rmation developing capabilities and
Fig 49 Field study class diagram
NA Behkami and TU Daim
95
Fig 410 Field study packages
4 Field Test
96
managing capabilities In Fig 411 these are expressed in a UML use case diagram notation Within the scope of the fi eld test subset of these activities including the knowledge stage and developing capabilities stage are evaluated in more detail in the following sections
Knowledge Stage for CMP
The UML sequence diagram in Fig 412 was created and shows the stakeholders and sequence of actions that shape the ldquo Knowledge Stage rdquo of Rogersrsquo diffusion process The ldquo HIT Innovation Supplier rdquo (in this case OHSU for CMP) attends a ldquo Conference rdquo such as the Annual AGA Conference (American Geriatrics Association) where a ldquo Physician rdquo comes to their presentation and becomes aware of the innovation (CMP) at the conference If the ldquo Physician rdquo decides that CMP may be useful for their clinic they go back and inform the ldquo Clinic rdquo that they work at about CMP including the ldquo Nurses rdquo ldquo CEO rdquo (or other administrative decision maker) and other ldquo Physician ( s )rdquo The interactions of these multiple stakeholders over time forms the ldquoKnowledge Stagerdquo of Rogersrsquo Diffusion Theory Having this model with such level of detail allows us to examine the precise participants and decision points and examine the time elements of CMP adoption and diffusion processes
Dynamic Capability Development Stage
The UML sequence diagram in Fig 413 was created from data collected and shows the stakeholders and sequence of actions that shape the ldquo Dynamic Capability Development Stage rdquo for adoption of CMP Once a potential adopter gains knowl-edge of an innovation and later decided to adopt the innovation it goes into the loop
Government
Supplier
Care Seeker
Adoption
Rejection
Dissemenation
DevelopCapabilities
Manage Capabilities
Reconfirmation
Usage
Implementation
Provider
Payer
Fig 411 Field study use case diagram
NA Behkami and TU Daim
97
of acquiring the dynamic capabilities necessary to successfully adopt the innova-tion Figure 413 shows the dynamic capabilities needing to be in place to adopt CMP which include (1) having CMP software (2) nurse care manager and (3) get-ting reimbursed from the government for using HIT The sequence diagram here only shows the positive path meaning that it assumes that the adopter was able to acquire the capabilities and adopt CMP
Fig 412 Sequence diagram ldquoknowledge stagerdquo
Fig 413 Sequence diagram ldquodynamic capability development stagerdquo
4 Field Test
98
Overall Adoption Decision State Chart
What the sequence diagram in the previous section couldnrsquot show about alternative paths for decisions can be illustrated in Fig 414 using a UML activity diagram The happy path is down the middle of the diagram where when the clinic decides to adopt CMP it already has the three needed capabilities (CMP software a nurse care man-ager and a way to get paid by payers) In that case it can quickly move down the middle and adopt CMP and therefore is less likely it would reject the innovation (CMP) However whatrsquos more interesting about this graph based on the interviews with experts and users is the alternate paths the scenario can take If some of the three needed capabilities are not in place the adoption has to wait until those remaining capabilities are either built or bought before true adoption happens This supports the objective of the proposed research that awareness alone is not enough as described in Rogers to move to next step of adoption Meaning after knowledge of innovation capabilities need to be developed or bought to truly adopt an innovation
4353 Classifi cation of Capabilities
Recall from earlier sections of this document that various researchers have attempted to classify capabilities or competencies necessary for competitive advantage namely Barney Figure 415 and Itami Figure 416 Similar to their works based on the data collected from my feasibility study a classifi cation of dynamic capabilities for HIT adoption (CMP) can be generated (Fig 417 )
4354 Limitations
While the purposed model is fl exible and could accommodate studying various types of organizations (hospitals) patients or providers the following are some of the limitations
bull The proposed model is a qualitative-based descriptive case study What it tries to do is to understand and bound the problem for one case Therefore the fi ndings
NoNo
No
Develop or BuyCapability
(CMP Software)
Develop or BuyCapability
(Receive Payments)
Develop or BuyCapability
(Nurse Care Manager)
Decides to AdoptInnovation
AdoptInnovation
RejectInnovation
already haveCapability
already haveCapability
already haveCapability
Yes Yes Yes
Fig 414 Field study state chart for adoption decision
NA Behkami and TU Daim
99
cannot be immediately generalized to a whole population of clinics with wide varying capabilities However it does set the foundations for a second-phase qualitative research studies in the future For example the results can be used in a qualitative study to measure the prevalence of certain type of capabilities across a group of fi rms (clinics)
bull Different fi rms (clinics) that adopt an innovation (CMP) may implement capa-bilities in various ways with varying implementation qualities The quality of capability implementation and its effect on the adoption and diffusion process are not directly captured in this model and are a good future research topic
bull Capabilities that are needed in the context of adoption of one HIT innovation (eg CMP) often exist alongside capabilities used in other hospital systems at the clinic The current research doesnrsquot specifi cally look at the relationship
Fig 415 Barneyrsquos classifi cation of capabilities
Fig 416 Itamirsquos classifi cation of assets
4 Field Test
100
between CMP capabilities (unless directly interfacing with CMP) and other hos-pital systems for example billing electronic health record disease registry etc
bull This research does not look at the internals of the process required for acquiring capabilities itrsquos treated as a black box Existence of (or lack of) these capabili-ties interfacing with them and their timing are of most importance to the proposal
bull Although due to its sophistication the CMP product at OHSU in many ways is a perfect HIT innovation to study but it mostly targets older adults and extremely sick patients A healthier target population such as professional workers less than 40 years of age may have unique infl uences on the HIT adoption and diffusion process that may not be highlighted in this choice of application to study
bull Similar to using multi-perspective to represent stakeholder and views in classi-fi cation of capabilities for HIT innovation (CMP) it could be benefi cial to use levels For example a small clinic may need a subset of capabilities that a larger hospital would need for adoption Using multi-levels would be a constructive endeavor for future research
436 Simulation A System Dynamics Model for HIT Adoption
Adoption of healthcare IT (HIT) is a critical factor in addressing quality and cost of patient care The assessment and diffusion of health IT have been the subjects of numerous studies Through this model factors infl uencing the adoption process and the relationships between them are examined As highlighted in the previous sec-tions healthcare systems are complex systems Their highly fragmented structure
HIT AdoptionCapabilities (CMP)
Technology
Work Flow
CMPSoftware
EHRIntegration
ReimbursementPayment Processing
Training
Nurse CareManager Training
PhysicianTraining
Patient LearningCommunity
Patient PanelManagement
Skilled Worker(Nurse Care Manager)
Fig 417 Field study taxonomy of capabilities
NA Behkami and TU Daim
101
makes it diffi cult to clearly understand healthcare problems Without a clear understanding evaluating response strategies becomes a diffi cult endeavor One methodology that can take us closer to a solution is system dynamics This report uses a system dynamics (SD) approach to evaluate a part of the problem (Behkami 2009b ) SD allows exploration of policy options through simulation The main objective of this study is to uncover the basic adoption process in the US healthcare system and evaluate each source of adoption
4361 Reference Behavior Pattern
Actual behavior of the real-world model for this report is based on two theories and two examples
bull Diffusion of innovation theory by Rogers bull Bass diffusion model with modeling disease epidemics example (Sterman amp
Sterman 2000 ) bull Bass diffusion model with cable TV penetration in US households (Sterman amp
Sterman 2000 )
ldquo Diffusion is the process in which an innovation is communicated through certain channels over time among the members of a social system rdquo (Rogers amp Rogers 2003 ) This special type of communication is concerned with new ideas It is through this process that stakeholders create and share information together in order to reach a shared understanding Some researchers use the term ldquodissemina-tionrdquo for diffusion that are directed and planned In his classic work (Rogers amp Rogers 2003 ) Rogers identifi es four main elements in the diffusion process that are virtually present in all diffusion research (1) an innovation (2) communication channels (3) over time and (4) social systems
The diffusion and adoption of new ideas and new products often follows S-shaped growth patterns Adoption of new technologies spreads as those who have adopted them come into contact with those who havenrsquot and persuade them to adopt the new system The new believers in turn then persuade others An example of the Bass diffusion model for adoption of cable TV (Sterman amp Sterman 2000 ) by house-holds can be used as a reference for health IT model The example identifi ed the following important factors in a householdrsquos decision to subscribe to cable TV
bull Favorable word of mouth from existing subscribers bull Positive experience viewing cable at the homes of friends and family bull Keeping up with the Joneses bull Feeling hip because of consuming on cable only knowledge
Similarly adoptions of HIT applications depend on favorable word of mouth from hospitals or clinics that currently use the HIT product Also positive empirical and fi nancial evidence through industry publications shows that the HIT application improved patient care and fi nancials of the clinic
4 Field Test
102
4362 Model Development
In this Bass style model as seen in Fig 418 potential adopters were broken down into large and small practices Small practice is enticed by large government reim-bursement to adopt and is assumed not to be affected by word of mouth or advertis-ing for adoption Itrsquos important to mention that word of mouth may affect the choice of HIT vendor for adoption in a small clinic but nonetheless act of adoption is for certain and itrsquos this part that is of interest to this report
The model in this report captures some of the important variables that have been identifi ed through a literature review and interviewing a physician The model includes three stocks
bull Small Practice Potential Adopters ldquoSPrdquo represents the number of small clinic that have not adopted health IT
bull Large Practice Potential Adopters ldquo LP rdquo represents the number of large clinics that have not adopted health IT
bull Adopters ldquo A rdquo represents the number of small and large clinics that have adopted health IT
In this model potential adopters are grouped into small and large practice The small practices will be receiving a $40000 reimbursement check from the OBAMA stimulus package for adopting health IT Large practices will not receive any stimu-lus and they will continue adopting health IT per their business and strategic plans Adoption rates ldquoLARrdquo and ldquoSARrdquo represent number of clinics adopting per time for large and small practices respectively
1 LAR = Adoption from advertising + adoption from word of mouth
(a) Adoption from advertising = a times SP (b) Adoption from word of mouth = c times i times LP times AN
2 SAR = Adoption from government stimulus = j times LP
Adoption for large clinic can occur from two sources
3 Adoption_from_Advertising = Large_Potential_Adopters times Advertising_ Effectiveness
4 Adoption_from_word_mouth = contact_rate times adoption_fraction_i times (adopterstotal_population)
Adoption for small clinics can happen only because of
5 Adoption_from_Government_ incentive = Small_Potential_Adopters times Adoption_ fraction_j
Total adopters
6 Adopters ldquoArdquo = SAR + LAR
LargePotentialAdaptors
SmallPotentialAdaptors
Adaptors Fig 418 Small and large clinic adaptors
NA Behkami and TU Daim
103
4363 Assumptions
bull Model refers to health IT as a set of defi nable features that would be benefi cial to use for the clinics and patients For the purposes of this model it is not assuming any particular product(s)
bull Model assumes at time = 0 that there are no adopters in existence from small or large practices
bull Model assumes that small clinics are infl uenced by government stimulus for adoption only while large practices are infl uenced by advertising or word of mouth adoption only
bull All clinics (small or large) will at some point adopt the HIT bull Once a clinic adopts it will not reject the HIT and go back to potential adopters
Table 42 lists the other assumptions and parameters for the model
Table 42 Parameters for system dynamics model
Parameter Description Value
HIT adoption carrying capacity
This is the number of clinics or hospitals that exist in the USA that are potential adopters
There are 52 hospitals in Oregon to get a national level number I simply times 50 states rarr N = 2600
Large clinicshospital N Large practice potential adopters 1000 Small clinichospital Small practice potential adopters 1600 Advertising effectiveness ldquo a rdquo
Is a parameter to be estimated statically from the data on adopters According to interviews for one HIT application a presentation is usually made to 20ndash40 attendees at average of 3ndash5 conference per year
Range of contacts is 60ndash200 person contacts per year Based on very rough data about 1ndash10 of these contacts through conference advertising adopt the particular HIT = 0003
Adoption fraction for word of mouth ldquo i rdquo
Not every encounter results in adoption The portion of contacts that are suffi ciently persuasive to induce the potential adopter to adopt the innovation is termed here the adoption fraction and denoted i
Rough estimate = 001
Contact rate from word of mouth
Adopters and potential adopters encounter one another with a frequency determined by contact rate
8
Adoption fraction ldquo j rdquo for small practices
The government stimulus available for a 2-year period If all small clinics take advantage it can be estimated
02
4 Field Test
104
4364 Role of Feedback (Fig 419 )
Loop ldquoadopters from advertisingrdquo
( LP rarr adoption _ from _ advertising rarr LAR rarr A rarr LP ) When the innovation or new product is introduced the adoption rate consists entirely of people who learned about the innovation from external sources of information such as advertising
Loop ldquoadopters from word of mouthrdquo ( LP rarr adoption _ from _ word _ mouth rarr LAR rarr A rarr LP ) As the pool of potential adopt-
ers declines while the adopter population grows the contribution of advertising to the total adoption rate falls while the contribution of word of mouth rises Soon word of mouth dominates and the diffusion process plays out as in the logistic diffusion model
Loop ldquogovernment incentives accelerate adoption by small clinicsrdquo
( SP rarr Government _ Incentive rarr SAR rarr A rarr SP ) When government incentive is intro-duced small practice adoption rate is stimulated
4365 Model Verifi cation
For verifi cation purposes the implemented model is compared to the conceptual model To build confi dence unintentional errors were removed and the model was checked for common errors such as units of measure data-entry errors (parameters
PotentialAdopters
Large PracticeLP
Potential AdoptersSmall Practice
SP
Total LargePractice Population
N
AdoptionFraction
Contact Ratec
MarketSaturation
AdvertisingEffectiveness
a
Adoption fromAdvertising inConferences
B
B
B
R
MarketSaturation
Adoption RateLAR
Word ofMouth
AdoptersA
Adoption fromInstitutional word of
Mouth
Adoption RateSAR
AdoptionFraction
j
Adoption from GovermnetSmall Practice Incentive
$40k
+
+
+
+ +
+
-
+
+
+
i
Fig 419 Vensim model for HIT
NA Behkami and TU Daim
105
initial values etc) and time scale errors Process of isolating errors include doubting frame of mind outside doubters walkthrough and hypothesis testing techniques
Doubting Frame of Mind
The goal of this activity is to fi nd scenarios that cause the model to fail so that we can isolate and correct errors Table 43 shows the scenarios tested for and their results
Outside Doubters
The model was shown to an engineering graduate student The student knew and understood the modeled system and its intended operation but it was not involved in its construction Model passed outside doubter check and future additions were suggested
Walkthroughs
The modeler explained the modelrsquos logic to a small group of individuals who are familiar with the system being modeled they included a physician and a health-care researcher Model passed walkthrough and three items were highlighted (1) the Bass model of diffusion was the correct theory to apply and (2) healthcare systems and policies are much more complicated than the current model however this is an acceptable and promising fi rst pass at modeling heath IT adoption (Table 44 )
Table 43 Doubting frame of mind tests
Test Expected result Actual result or fi x
Advertising_effectiveness = 0 No move from potential adopters to adopters
Pass Adoption_fraction_word_mouth = 0 Adoption_fraction_advertising = 0 Advertising_effectiveness = 3000 Make sure that advertising_
effectiveness is always less than 1 Total population N (used for word_of_mouth_effectivness calculation not matching starting population of potential adopters 1000 versus 2000)
Model still runs but wrong shape to adoption curve
Correct
Starting population lt 0 Model still works but wrong shape to adoption curve
Make sure that starting population is correct each time (initial condition)
4 Field Test
106
Hypothesis Testing
To fully exercise the model hypothesis tests with various conditions were developed
Tornado Diagram
Tornado diagram is used to summarize results of varying model parameters and initial values Each parameter and its initial condition are varied from baseline by plusmn10 (Fig 420 )
Table 44 Hypothesis testing cases
Conditions Performance estimate Run and compare
Large_Potential_Adopters = 1000 Advertising will dominate word_of_mouth adoption in the fi rst months Government_adoption will be fastest
Pass Small_Potential_Adoptors = 1600 Adopters = 0 Advertising_Effectiveness = 003 Word_of_mouth_adoption_fraction = 001 Contact_Rate = 8 Government_adoption_fraction = 002 Large_Potential_Adopters = 1000 No adopters at all Pass Small_Potential_Adoptors = 1600 Adopters = 0 Advertising_Effectiveness = 0 Word_of_mouth_adoption_fraction = 0 Contact_Rate = 0 Government_adoption_fraction = 0 Large_Potential_Adopters = 1000 Adopters from government_
incentive only Pass
Small_Potential_Adoptors = 1600 Adopters = 0 Advertising_Effectiveness = 0 Word_of_mouth_adoption_fraction = 0 Contact_Rate = 0 Government_adoption_fraction = 002 Large_Potential_Adopters = 1000 Adopters from large
practices only Pass
Small_Potential_Adoptors = 1600 Adopters = 0 Advertising_Effectiveness = 003 Word_of_mouth_adoption_fraction = 001 Contact_Rate = 8 Government_adoption_fraction = 0
NA Behkami and TU Daim
107
4366 Model Validation
Having verifi ed the model it is validated against reference behavior pattern (RBP) comparing the conceptual model to reality In validating the health IT adoption model the two validation ldquoparadigmsrdquo of rational and practical are suitable fi ts The model fi ts the rational (conceptual) paradigm by being believable and one is able to reason about its structureassumptionslogic The model fi ts the practical paradigm because it meets its intended goal to understand how quickly hospitals may adopt HIT (under optimistic conditions) The learning realized from the model justifi es its development cost
Earlier in this report in the RBP we identifi ed two theories of diffusion with two real-world examples of innovation adoption Using a multi-perspective approach (of modeler technical evaluator and user) based on the models conceptual validity operation validity and believability were able to validate that the correct model has been built
Conceptual Validity
The created model exhibits the concepts identifi ed by Rogersrsquo classical theory on Diffusion of Innovation (Rogers amp Rogers 2003 ) Theory states that Diffusion of Innovation includes communicating messages This communication requires chan-nels by which messages move from one individual or unit to another The context of the information sharing determines the experience of the communication and whether ultimately the receivers adopt the innovation According to Rogers adoption evaluations can be objective or subjective However they are often subjective based on information reaching the individual through other communication channels
Communication can occur between hemophilic or heterophilic individuals Homophily refers to how similar two interacting individuals are based on their beliefs education etc Heterophily is the opposite and refers to how different from each other interacting individuals are
Two individuals that are homophilous are able to create more meaningful com-munications One of the barriers in innovation of diffusion is that participants are very heterophilous For example an inventor with an engineering background often has diffi culty communicating merits of his or her innovation to investors or poten-tial nontechnical users
+ndash10AdoptorsLarge_Potential_AdoptersSmall_Potential_AdoptorsAdvertising_EffectinvessWord_of_mouth_adoption_fractionContact_RateGovernment_adoption_fraction
ndash20 ndash10 +10 +20260010001600003001
802
Base
Fig 420 Tornado diagram
4 Field Test
108
Time is involved in three stages (1) the time that passes between fi rst knowledge and adoption or rejection of an innovation (2) the earliness or lateness that an individual adopts compared to the group (3) innovation rate of adoption which is the number of people that adopt it during a particular period of time
Operational Validity
Looking and comparing the model-generated behavioral data is characteristic of other real-world system behavioral data In this regard the Bass diffusion model (Sterman amp Sterman 2000 ) has showed that when the innovation or new product is introduced the adoption rate consists entirely of people who learned about the inno-vation from external sources of information such as advertising As the pool of potential adopters declines while the adopter population grows the contribution of advertising to the total adoption rate falls while the contribution of word of mouth rises Soon word of mouth dominates and the diffusion process plays out as in the logistic diffusion model The Bass model solves the start-up problem of the logistic innovation diffusion model because the adoption rate from advertising does not depend on the adopter population
The developed model is further validated by the Bass model used for modeling epidemics in section 92 of Shermanrsquos Business Dynamics book
Believability
Sterman introduced an S-shaped growth discussing the adoption of cable TV view-ing in households in the 1960s This model is widely accepted and verifi ed in aca-demics and industry Additionally the concept of adoption of cable TV is a concept that many individuals can easily comprehend today Therefore using cable TV adop-tion as an analogy the developed model is rendered believable to majority of indi-viduals Cable TV adoptions and HIT share many of the same diffusion dynamics
4367 Results and Discussion
When an innovation is introduced and the adopter population is zero the only source of adoption will be external infl uences such as advertising The advertising effect will be largest at the state of the diffusion process and steadily diminish as the pool of potential adopters is depleted Figure 421 shows the behavior of the Bass model for CMP The total population N is assumed 2600 hospitals Advertising effectiveness a and the number of contacts resulting in adoption from word of mouth ci were estimated to be 0005 per year and 016 per year respectively The contribution of adoption from advertising is small in general and on a decline after the fi rst year as seen in Figs 422 and 423 Adoption through word of mouth peeks after the second year
NA Behkami and TU Daim
109
Adopters A
4000
4000
1000
00 6 12 18 24 30
Time (Month)Adopters A Current
36 42 48 54 60
3000
Fig 421 Adopters
Adoption Rates
40
2020040
000
0 6 12 18 24 30
Time (Month)
Adoption from Advertising in Conferences Current
Adoption from Government Small Practice Incentive $40k Current
Adoption from Institutional word of Mouth Current
36 42 48 54 60
80400
Fig 422 Adoption rates
Selected Variables
4000
2000500
1000
000
0 6 12 18 24 30
Time (Month)Adopters A Current
Potential Adopters P Current
Small Practice Potential Adopters S Current
36 42 48 54 60
20001000
Fig 423 Other model variables
4 Field Test
110
This report presented an SD model to study the HIT adoption process in the US healthcare system Using a system dynamics view brings a fresh and much-needed means for studying the adoption process The overview of the model does not show an unexpected dominant loop and more work remains to be done to benefi t more comprehensive conclusions
4368 Limitations
The presented model includes several limitations that should be addressed in future work in order to improve the representation of the system For example the model does not explicitly refl ect the interests of patients payers the high-tech industry etc The proposed model is valuable in providing a common ground for interested research parties and presenting an overall view of the system By expanding the model a simulation for evaluating policies and strategies can be obtained which is a main objective of developing system dynamics theory
References
Bailey D C A (2006) A guide to qualitative fi eld research Thousand Oak CA Pine Forge Press Behkami N A (2009a) Qualitative research interview design for a health IT application
Portland Department of Engineering amp Technology Management Portland State University Working Paper Series
Behkami N A (2009b) A system dynamics model for adoption of healthcare information tech-nology Portland Department of Engineering amp Technology Management Portland State University Working Paper Series
Bodenheimer T Wagner E amp Grumbach K (2002a) Improving primary care for patients with chronic illness Journal of the American Medical Association 288 (14) 1775ndash1779
Bodenheimer T Wagner E amp Grumbach K (2002b) Improving primary care for patients with chronic illness The chronic care model Journal of the American Medical Association 288 (15) 1909ndash1914
Counsell S Callahan C Clark D Tu W Buttar A Stump T et al (2007) Geriatric care management for low-income seniors A randomized controlled trial Journal of the American Medical Association 298 (22) 2623ndash2633
Dorr D Brunker C Wilcox A amp Burns L (2006) Implementing protocols is not enough The need for fl exible broad based care management in primary care
Dorr D Wilcox A Burns L Brunker C Narus S amp Clayton P (2006) Implementing a multidisease chronic care model in primary care using people and technology Disease Management 9 (1) 1ndash15
Dorr D Wilcox A Donnelly S Burns L amp Clayton P (2005) Impact of generalist care man-agers on patients with diabetes Health Services Research 40 (5) 1400ndash1421
Friedman B Jiang H Elixhauser A amp Segal A (2006) Hospital inpatient costs for adults with multiple chronic conditions Medical Care Research and Review 63 327ndash346
Health Care Partners Medical Group ldquoAbout HealthCare Partnersrdquo OHSU ldquoCare Management Plus Program Websiterdquo ORPRN ldquoOregon Rural Practice-based Research Network Websiterdquo Rogers E amp Rogers E (2003) Diffusion of innovations (5th ed) New York Free Press
NA Behkami and TU Daim
111
Rubenstein L Parker L Meredith L Altschuler A dePillis E Hernandez J et al (2002) Understanding team-based quality improvement for depression in primary care Health Services Research 37 (4) 1009ndash1029
Shojania K amp Grimshaw J (2005) Evidence-based quality improvement The state of the sci-ence Health Affairs (Millwood) 24 (1) 138ndash150
Shojania K Ranji S McDonald K Grimshaw J Sundaram V Rushakoff R et al (2006) Effects of quality improvement strategies for type 2 diabetes on glycemic control A meta- regression analysis Journal of the American Medical Informatics Association 296 (4) 427ndash440
Steffens D Snowden M Fan M Hendrie H Katon W amp Unutzer J (2006) Cognitive impairment and depression outcomes in the IMPACT study The American Journal of Geriatric Psychiatry 14 (5) 401ndash409
Sterman J amp Sterman J D (2000) Business dynamics Systems thinking and modeling for a complex world with CD-ROM Irwin McGraw-Hill
Wolff J Starfi eld B amp Anderson P G (2002) Expenditures and complications of multiple chronic conditions in the elderly Archives of Internal Medicine 162 (20) 2269ndash2276
Young A Mintz J Cohen A amp Chinman M (2004) A network-based system to improve care for schizophrenia The Medical Informatics Network Tool (MINT) Journal of the American Medical Informatics Association 11 (5) 358ndash367
4 Field Test
113copy Springer International Publishing Switzerland 2016 TU Daim et al Healthcare Technology Innovation Adoption Innovation Technology and Knowledge Management DOI 101007978-3-319-17975-9_5
Chapter 5 Conclusions
Tugrul U Daim and Nima A Behkami
51 Overview and Theoretical Contributions
Despite the fact that diffusion theory was introduced several decades earlier we still donrsquot seem to truly understand how the phenomenon impacts our society In recent years many researchers including Rogers the father of diffusion theory have called for renewed interest in diffusion research One domain as discussed in this proposal which can benefi t from better understanding of diffusion is the fi eld of healthcare specifi cally improvements in understanding adoption and diffusion process for health information technology (HIT) Due to various factors including changing demographics the US healthcare delivery system is facing a crisis and having real-ized this government and private entities are pouring support into advocating HIT adoption-related research amongst other initiatives
One such research that would help with this agenda is the research proposed in this study This study has shown that indeed an extension of Rogersrsquo diffusion the-ory using the extension of dynamics capabilities can help further our understanding of what it takes for successful innovations to diffuse in the US Healthcare industry This report started by proposing a dynamic capability extension to diffusion theory Then it was reasoned for why diffusion theory rather than other adoption theory due to its macro-level property rather than micro is the appropriate theory for the pro-posed study It was also shown that how dynamic capabilities as a one manifestation of ldquofactors of productionrdquo originating from the strategic management fi eld can be used to further characterize the adoptiondiffusion decision and its life cycles
T U Daim () Portland State University Portland OR USA e-mail tugruludaimpdxedu
N A Behkami Merck Research Laboratories Boston MA USA
114
This study also shows that use of a case study or grounded theory types of quali-tative research is necessary to do an exploratory study of the problem Itrsquos through this type of research that we hope to gain in-depth understanding of situation and meaning for those involved In future research the results of such mostly qualitative- based research can be inputs for hybrid or purely quantitative method research on the same topics and in the same fi eld after the problem and whatrsquos really going on have been structured a little more with qualitative methods Additionally in this report various system modeling tools were compared and contrasted for purposes of analysis documentation and communication of research fi ndings It was shown that for this research the use of the Unifi ed Modeling Languages (UML) is a productive fi t UML benefi ts from having constructs for both showing static and dynamics aspects of the system UML also supports multi-perspective views of the problem which was also shown here to be essential for understanding HIT diffusion innovation
In addition to comparing and discussing various methodologies theories and aspects of the problem in this document the proposed research was accompanied and verifi ed for demonstrability and validity by conducting a fi eld study at Oregon Health amp Science University with its Care Management Plus team CMP a HIT- based innovation is an ambulatory care model for older adults and people with multiple conditions components of CMP include software clinic business pro-cesses and training The fi eld study was conducted using site readiness survey and expert interviews The data collected was analyzed using thematic analysis includ-ing open and focus coding Models were created using diffusion and dynamic capa-bility theory and they were documented using multi-perspectives and the UMLrsquos structural and behavior diagrams A system dynamics model based on Bass diffu-sion model was also created and demonstrated And in conclusion conducting the fi eld study was able to demonstrate that the research objectives (generally for pro-posal and specifi cally for fi eld study) were met
Objectives 1 and 2 were about showing that DOI and dynamic capabilities can be combined in a meaningful manner
Objective 2 Demonstrate that dynamic capability theory can be used and how to meaningfully extend diffusion of innovation theory
This objective was demonstrated based on the model constructed from site data collection as described in Fig 51 where itrsquos the clinic need(s) that drives them to consider adopting an innovation And this need and decision have barriers andor infl uences that can affect them in a negative or positive way Additionally as that same fi gure shows whether a clinic has the needed capabilities to adopt or not becomes a pressure point as either an positive infl uence (in case they already have the capabilities) or a barrier (in case clinic doesnrsquot have the needed capability yet)
In further support of the Objective 2 Fig 52 a depiction of the ldquodynamic capa-bility development stagerdquo shows the sequence and time frame of acquiring capabili-ties prior to truly adopting an innovation These two points mentioned indeed validate and support the second objective which helps in drawing the picture in Fig 53 that demonstrates how dynamic capabilities can be used to meaningfully extend diffusion of innovation theory
TU Daim and NA Behkami
115
Objective 1 Identify some dynamic capabilities needed for successful implementa-tion of HIT (Care Management Plus OHSU)
In supporting Objective 1 data collection and analysis from OHSU CMP adop-tion verifi ed that indeed dynamic capabilities needed for successful implementation of HIT can be defi ned Compliant with classifi cations from prior work namely Fig 54 Barneyrsquos classifi cation of factors of production (aka capabilities compe-tences) from Resource Based Theory and Fig 55 Itamirsquos classifi cation of assets for competitive advantage a classifi cation of capabilities for CMP adoption was devel-oped and the taxonomy is shown in Fig 56
Fig 51 Clinic workfl ow
Fig 52 Sequence diagram ldquodynamic capability development stagerdquo
5 Conclusions
116
Fig 53 New extensions to Rogersrsquo DOI theory
Fig 54 Barneyrsquos classifi cation of capabilities
Fig 55 Itamirsquos classifi cation of assets
TU Daim and NA Behkami
117
Objective 3 Use Software and system engineering methods including ldquo4 + 1 viewrdquo for perspectives and UML to demonstrate documentation and analysis
Support for Objective 3 in the fi eld study was demonstrated by the choice of qualitative data collection methodology The data collection was analyzed using standard qualitative thematic analysis similar to grounded theory with fi rst open coding and then focused coding Then the analysis model was built and documented using UML and later analyzed (in the form of discussing results) using static and behavioral aspects of the system Examples of software engineering artifacts pro-duced in the study included the static UML diagrams of Fig 57 fi eld study class diagram Fig 58 fi eld study package diagram the behavioral UML diagrams of Fig 59 fi eld study use case and the sequence diagrams of Fig 510 ldquoknowledge stagerdquo Fig 52 ldquodynamic capability development stagerdquo and the UML state chart Fig 511 fi eld study start chart for adoption decision The scenarios and use cases used in building the behavioral UML artifacts just mentioned are compliant with the ldquo4 + 1 viewrdquo model for describing system architectures
Generation of these UML diagrams verifi es that indeed software engineering thinking and tools were successfully applied to the research These UML artifacts and the multi-perspective analysis in this document support Osterweilrsquos hypothesis that process is software in spite of domain (Osterweil 1987 1997 ) and demon-strates that software principles also hold for social and organizational processes
Objective 4 Build and run a small simulation of the DOI theory extension using system dynamics
A complete system dynamics model was developed for the fi eld study and docu-mented in this report The model was based on Rogersrsquo diffusion theory and Bass diffusion model In the model adoptiondiffusion rates for CMP at OHSU were
HIT AdoptionCapabilities (CMP)
Technology
Work Flow
CMPSoftware
EHRIntegration
ReimbursementPayment Processing
Training
Nurse CareManager Training
PhysicianTraining
Patient LearningCommunity
Patient PanelManagement
Skilled Worker(Nurse Care Manager)
Fig 56 Field study taxonomy of capabilities
5 Conclusions
118
modeled using word of mouth and advertising A complete set of system dynamics components were developed including causal loop diagram (CLD) (Fig 512 ) and stock and fl ow system dynamic model in Vensim software (Fig 513 ) The model was extensively validated and verifi ed using popular methods Verifi cation was per-formed with the techniques of doubting frame of mind outside doubter walk-through hypothesis testing and tornado diagram testing Model was validated using conceptual validity operational validity and the believability test Figure 514 an S-curve of adopter population along with Figs 515 and 516 growth curves showing adoption rates were outputted by the model The generate model and its outputs show that itrsquos possible to effectively model the HIT adoption and diffusion process in a good enough way so that we can experiment with scenarios and forecasting In future research this model can be extended to integrate dynamic capabilities
Fig 57 Field study class diagram
TU Daim and NA Behkami
119
Fig 58 Field study packages
5 Conclusions
120
In conclusion all objectives of the research proposal were met and demonstrated through preparation of this document Along with the results of the included feasi-bility fi eld study itrsquos verifi ed that indeed there is a need for extension of Rogersrsquo theory Dynamic capabilities are a good fi t candidate integrating with Rogersrsquo diffu-sion theory and extending it Additionally the combination of the presented theories and methods in this document can assist healthcare stakeholders understand their problems and solution more effi ciently as they set new policies and investment for their support
Government
Supplier
Care Seeker
Adoption
Rejection
Dissemenation
DevelopCapabilities
Manage Capabilities
Reconfirmation
Usage
Implementation
Provider
Payer
Fig 59 Field study use case diagram
Fig 510 Sequence diagram ldquoknowledge stagerdquo
TU Daim and NA Behkami
121
NoNo
No
Develop or BuyCapability
(CMP Software)
Develop or BuyCapability
(Receive Payments)
Develop or BuyCapability
(Nurse Care Manager)
Decides to AdoptInnovation
AdoptInnovation
RejectInnovation
already haveCapability
already haveCapability
already haveCapability
Yes Yes Yes
Fig 511 Field study state chart for adoption decision
LargePotentialAdaptors
SmallPotentialAdaptors
Adaptors Fig 512 Small and large clinic adaptors
PotentialAdopters
Large PracticeLP
Potential AdoptersSmall Practice
SP
Total LargePractice Population
N
AdoptionFraction
Contact Ratec
MarketSaturation
AdvertisingEffectiveness
a
Adoption fromAdvertising inConferences
B
B
B
R
MarketSaturation
Adoption RateLAR
Word ofMouth
AdoptersA
Adoption fromInstitutional word of
Mouth
Adoption RateSAR
AdoptionFraction
j
Adoption from GovermnetSmall Practice Incentive
$40k
+
+
+
+ +
+
-
+
+
+
i
Fig 513 Vensim model for HIT
5 Conclusions
122
Adopters A
4000
4000
1000
00 6 12 18 24 30
Time (Month)Adopters A Current
36 42 48 54 60
3000
Fig 514 Adopters
Adoption Rates
40
2020040
000
0 6 12 18 24 30
Time (Month)
Adoption from Advertising in Conferences Current
Adoption from Government Small Practice Incentive $40k Current
Adoption from Institutional word of Mouth Current
36 42 48 54 60
80400
Fig 515 Adoption rates
Selected Variables
4000
2000500
1000
000
0 6 12 18 24 30
Time (Month)Adopters A Current
Potential Adopters P Current
Small Practice Potential Adopters S Current
36 42 48 54 60
20001000
Fig 516 Other model variables
TU Daim and NA Behkami
123
52 Recommended Proposition for Future Research
The following research propositions are formulated in the context of information discussed in the previous sections
Proposition 1 Even though the clinics obtain knowledge of a new innovation and decide to adopt it it is actually the acquirement of the needed minimum set of capabilities (for meaningfully using the innovation) which strongly infl uences successful adoption
Proposition 2 Only meaningful adoption can be considered the ldquoreal adoptionrdquo and should be the main type used in planning and management Meaningful is using the adopted innovation according to defi ned set of criteria that has some type of agreed on or expected benefi t (eg the recent HIT meaningful use intuitive and measures sponsored by the US Health and Human Services [HHS] department)
Proposition 3 Acquiring capabilities that need to be implemented and using an innovation (part of adoption) will take time The velocity by which a potential adopter can acquire the needed capabilities will strongly infl uence adoption rates and overall diffusion
Proposition 4 Taking inventory and tracking of capabilities across a similar or competing group of fi rms regions or situations can act as a scoreboarddashboard of sorts for better analysis decision making and overall general stra-tegic management
Proposition 5 Investment in acceleration of acquiring of capabilities (for successful adoption) rather than the classical and hard-to-track general fi nancial invest-ments (or the likes) by sponsors can strongly infl uence diffusion rates
Proposition 6 Classical diffusion theory needs to be extended to account for the period in time and effort that fi rms (in this example clinics) expand to contem-plate or acquire capabilities
Proposition 7 When an adopter (clinic) decides to adopt an innovation either it suc-cessfully acquires the needed capabilities and the conditions to use the innova-tion or the adoption eventually fails
Proposition 8 The Software Engineering techniques of Object-Oriented Analysis and Design (OOAD) in conjunction with UML can be used to study social and organizational processes in new and more effective ways
References
Osterweil L J (1987) Software processes are software too In Proceedings of the 9th International Conference on Software Engineering (p 13)
Osterweil L J (1997) Software processes are software too revisited an invited talk on the most infl uential paper of ICSE 9rsquo paper presented to the International Conference on Software Engineering In Proceedings of the 19th International Conference on Software Engineering Boston
5 Conclusions
Part II Evaluating Electronic Health Record Technology Models and Approaches
Liliya Hogaboam and Tugrul U Daim
This part reviews electronic health records and considers technology assessment scenarios for multiple purposes These are the following
(a) The adoption of EHR with focus on barriers and enablers (b) The selection of EHR with focus on different alternatives (c) The use of EHR with focus on impacts
The exploration will assume that the adoption selection and use of EHR relate to the ambulatory EHR accepted in small practices
The fi rst section will highlight the gaps each scenario will address and list match-ing research goals and research questions
The second section will describe a research project matching each objective above In each case we will explain the methodology of choice describe other methods that may also be considered and list the reasons to justify the methodology we are choosing We will develop a preliminary model for each research and list the theories behind
The third section will explain what kind of data we will need and how we will acquire it We will consider the following in this section
(a) The required data size in terms of number of data points respondents or experts
(b) Data access issues such as sample size or access to experts
The fourth section will explain the types of analyses to be done for each scenario We will consider the following in this section
(a) Types of metrics used to measure accuracy (b) Validity and reliability in each case
127copy Springer International Publishing Switzerland 2016 TU Daim et al Healthcare Technology Innovation Adoption Innovation Technology and Knowledge Management DOI 101007978-3-319-17975-9_6
Chapter 6 Review of Factors Impacting Decisions Regarding Electronic Records
Liliya Hogaboam and Tugrul U Daim
L Hogaboam bull T U Daim () Portland State University Portland OR USA e-mail tugruludaimpdxedu
61 The Adoption of EHR with Focus on Barriers and Enablers
Letrsquos explore the gaps found in the literature that relate to adoption of EHR with focus on enablers and barriers
bull The impact and signifi cance of implementation barriers and enablers (fi nancial technical social personal and interpersonal) have not been satisfactorily studied
bull Signifi cance of the relationship of factors of perceived usefulness perceived ease of use and perceived benefi ts on attitude toward using EHR in ambulatory set-tings has not been adequately shown with global studies
bull Lack of studies in the USA involving TAM models and research on a global scale
bull Lack of quantitative studies in EHR adoption toward small ambulatory settings
Palacio Harrison and Garets ( 2009 ) provided a research that documented an increased adoption of EHR in the US hospitals through the period of 2005ndash2007 The authors also indicate potential barriers of HIT implementation as cost lack of fi nancial incentives for providers and the need for interoperable systems
A systematic literature review on perceived barriers to electronic medical record (EMR) adoption identifi ed eight categories (fi nancial technical time psychologi-cal social legal organizational and change process Boonstra amp Broekhuis 2010 ) The study is bibliographical and explorative in nature and the barriers are not tested
128
for signifi cance rather interpreted as guidelines for EMR adopters and policy mak-ers and as a foundation for future research
Taxonomy of the primary and secondary barriers is listed in Table 61 below (Boonstra amp Broekhuis 2010 )
Boonstra and Broekhuis ( 2010 ) also noted that barriers in primary categories vary signifi cantly between small and large practices since small practices face greater diffi culties overcoming those barriers Those differences may greatly impact the focus and the effort needed to overcome fi nancial technical and time barriers
Table 61 Taxonomy of the primary and secondary barriers (Boonstra amp Broekhuis 2010 )
Primary category Primary barriers
Secondary category Associated barriers
Financial bull High start-up costs bull High ongoing costs bull Uncertainty about return
on investment (ROI) bull Lack of fi nancial
resources
Psychological bull Lack of belief in EMRs bull Need for control
Technical bull Lack of computer skills of the physicians andor the staff
bull Lack of technical training and support
bull Complexity of the system bull Limitation of the
system bull Lack of customizability bull Lack of reliability bull Interconnectivity
standardization bull Lack of computers
hardware
Social bull Uncertainty about the vendor
bull Lack of support from other external parties
bull Interference with doctor-patient relationship
bull Lack of support from other colleagues
bull Lack of support from the management level
Time bull Time to select purchase and implement the system
bull Time to learn the system bull Time to enter data bull More time per patient bull Time to convert the
records
Legal bull Privacy or security concerns
Organizational bull Organizational size bull Organizational type
Change process bull Lack of support from organizational culture
bull Lack of incentives bull Lack of participation bull Lack of leadership
L Hogaboam and TU Daim
129
While the study by Lorenzi et al ( 2009 ) reviews the benefi ts and the barriers of EHR in ambulatory settings it does not address EHR models or the barriers associ-ated with interconnectivity of EHR The authors indicate that more research is needed in those fi elds
A group of Canadian researchers (McGinn et al 2011 ) conducted a systematic literature review of EHR barriers and facilitators The review categorized the stud-ies based on the user groups (physicians healthcare professionals managers and patients) while the differences of clinic size and type of setting and the factors that are particular to each type were not discussed The study though is interesting in the sense of general ranking of the factors and commonalities in studies of those factors Technical issues are at the top of the list while organizational factors are not that common (McGinn et al 2011 ) The ranking (from most to least common) is shown in Table 62
The three studies mentioned in McGinn et al ( 2011 ) related to ambulatory care were exploratory andor qualitative in nature
Table of categories of studies examined through literature review is shown in Table 63
Electronic health records have been a topic of research in various countries throughout the world some with high rates of adoption and implementation and others with low ones While researching and working on my independent studies I have found a number of studies in foreign countries (Bates et al 2003 Rosemann et al 2010 Were et al 2010 ) High transition to EHR technology was reported in Australia New Zealand and England through fi nancial support and incentives evidence-based decision support standardization and strategic framework (Bates et al 2003 )
Those studies give a possibility to engage a similar research or test a certain framework here in the USA while studying adoption of EHR by small ambulatory clinics In Table 64 I have summarized some of those important studies
The US research in EHR adoption lacks rich involvement of TAM with structural equation modeling especially in ambulatory care While researching EHR adoption
Table 62 Common EHR implementation factors ranked by the number of studies
Common EHR implementation factors
Number of studies
Design or technical concerns 22 Privacy and security concerns 21 Cost issues 19 Lack of time and workload 17 Motivation to use EHR 16 Productivity 14 Perceived ease of use 13 Patient and health professional interaction 12 Interoperability 10 Familiarity ability with EHR 9
6 Review of Factors Impacting Decisions Regarding Electronic Records
130
in my independent studies projects and performing thorough literature reviews there were some interesting studies on EHR adoption in hospitals that deserve atten-tion Thus the researchers in New York built extended and modifi ed TAM with external variables (age specialty position in hospital attitudes toward HIT cluster ownership) and latent variables of pre- and post-adoption (Vishwanath Brodsky amp Shaha 2009 ) The signifi cant links of the external variable impacts were as follows age rarr perceived usefulness attitudes toward HIT rarr perceived usefulness as well as ease of use and position in hospital and cluster ownership rarr perceived ease of use (Vishwanath et al 2009 ) A study of physicianrsquos adoption of electronic detailing proposed the model that included innovation characteristics (perceived relative advantage compatibility complexity trialability observability) communication channels (peer infl uence) social system (academic affi liation presence of restric-tive policy urban vs rural) and physician characteristics (specialty years in prac-tice attitudes toward the information usefulness)
Some statistical studies related to EHR barriers have been performed For exam-ple a study by Valdes et al had one of the main objectives of the characterization of user and non-users of EHREMR software and identifi ed potential barriers to EHR proliferation (Valdes et al 2004 ) They performed a secondary analysis of member survey data collected by the American Academy of Family Physicians (AAFP) as well as the number of different software vendors reported by users of EHREMR The researchers reported at least of 264 different EHREMR software
Table 63 Categories of related studies examined in preparation to the exam
Type of study Research works
Qualitative or empirical evaluation of TAM or other acceptance models
Chiasson et al ( 2007 ) Dillon and Morris ( 1996 ) Im Kim amp Han ( 2008 ) Premkumar and Bhattacherjee ( 2008 ) Tsiknakis et al ( 2002 ) Szajna ( 1996 ) Scott and Briggs ( 2009 ) Yang ( 2004 ) Yusof et al ( 2008 )
Exploration of particular aspects of the HIT adoption
Burton-Jones and Hubona ( 2006 ) Cresswell and Sheikh ( 2012 ) Degoulet Jean and Safran ( 1995 ) Haron Hamida and Talib ( 2012 ) Janczewski and Shi ( 2002 ) Jeng and Tzeng ( 2012 ) Folland ( 2006 ) Hagger et al ( 2007 ) Karahanna and Straub ( 1999 ) Kim and Malhotra ( 2005 ) Lee and Xia ( 2011 ) Malhotra ( 1999 ) Martich and Cervenak ( 2007 ) McFarland and Hamilton ( 2006 ) Melone ( 1990 ) Shin ( 2010 ) Storey and Buchanan ( 2008 ) Viswanathan ( 2005 )
Applications of TAM and its derivatives in other countries
Jimoh et al ( 2012 ) Maumlenpaumlauml et al ( 2009 ) Polančič Heričko and Rozman ( 2010 ) Ortega Egea and Romaacuten Gonzaacutelez ( 2011 ) Yu Li and Gagnon ( 2009 )
Frameworks of IT adoption in healthcare that differed greatly from TAM
Davidson and Heineke ( 2007 ) Hatton et al ( 2012 )
Frameworks of IT adoption experimental in nature
Andreacute et al ( 2008 ) Ayatollahi Bath and Goodacre ( 2009 ) Becker et al ( 2011 )
L Hogaboam and TU Daim
131
Table 64 Summary of studies and a variety of methodologies and analyses used
Authors Country Study
Ludwick and Doucette ( 2009 )
Canada Lessons-learned study from EHR implementation in seven countries Concluded that systemsrsquo graphical user interface design quality feature functionality project management procurement and user experience affect implementation outcomes Stated that quality of care patient safety and provider-patient relations were not impacted by system implementation
Aggelidis and Chatzoglou ( 2009 )
Greece Examined the use of health information technology acceptance with the use of modifi ed and extended TAM Facilitating conditions (new computers support during information system usage and fi nancial rewards) was the main factor that positively impacted behavioral intention Perceived usefulness and ease of use were the most important factors of direct infl uence on behavioral intention Anxiety during system use shown to be reduced by facilitating conditions perceived usefulness and self-effi cacy
Melas et al ( 2011 )
Greece Researchers implemented confi rmatory factor analysis (CFA) structural equations modeling (SEM) and multi-group analysis of structural invariance (MASI) in a study of examining the intention to use clinical information systems in Greek hospitals The results showed direct effect of perceived ease of use on behavioral intention to use
Chen and Hsiao ( 2012 )
Taiwan Modifi ed TAM was used for IT acceptance research Confi rmatory factor analysis for reliability and validity of the model and SEM for causal model estimation were used According to the results of the study top management support had signifi cant impact on perceived usefulness while project team competency and system quality signifi cantly impact perceived use
Hung Ku and Chien ( 2012 )
Taiwan Modifi ed TBP was used and results indicated that physiciansrsquo intention to use IT was signifi cantly impacted by attitude subjective norm and perceived behavior control Studied impactful factors included interpersonal infl uence personal innovativeness in IT and self-effi cacy
Cheng ( 2012 ) Taiwan The researchers looked at IT adoption by nurses in two regional hospitals with extended TAM where the other factors impacting intention to use consisted of learner-system interaction instructor-learner interaction learner-learner interaction and fl ow
Pareacute and Sicotte ( 2001 )
Canada The study concluded that IT sophistication and perceived usefulness of clinical applications are moderately to highly correlated while no relationship was found between the level of sophistication and perceived usefulness of administrative applications
Moores ( 2012 )
France The researchers found that there are differences in signifi cant impacts depending on the experience of the users while applying extended and modifi ed TAM in studying adoption of clinical management system by hospital workers
(continued)
6 Review of Factors Impacting Decisions Regarding Electronic Records
132
Table 64 (continued)
Authors Country Study
Handy Hunter and Whiddett ( 2001 )
New Zealand
Conducted longitudinal study into primary care practitionersrsquo adoption of electronic medical record system for maternity patients in a large urban hospital applying TAM with additional variables like individual characteristics system characteristics organizational characteristics and system acceptability They concluded that technical aspects of information system should not be considered in isolation from organizational and social context
Van Schaik et al ( 2004 )
The UK The researchers outlined the need to consider the balance of benefi ts (perceived advantages) and costs (disadvantages) of a new system in technology acceptance modeling
Chow Chan et al ( 2012 ) Chow Herold et al ( 2012 )
Hong Kong
Included external variable for TAMmdashcomputer self-effi cacy in study of the factors impacting the intention to use clinical imaging portal
Pai and Huang ( 2011 )
Taiwan Study of HIT adoption by district nurses head directors and other related personnel where TAM was used with external variables (information quality service quality and system quality)
Duumlnnebeil et al ( 2012 )
Germany SEM model with six external variables (intensity of IT utilization importance of data security importance of documentation eHealth knowledge importance of standardization process orientation) was used to study physicianrsquos acceptance of e-health in ambulatory care The researchers stated that the diversities of public systems throughout the world should be integrated into TAM research in order to correctly explain the drivers Perceived importance of standardization and perceived importance of current IT utilization were the most signifi cant
programs in use which indicates highly fragmented market which authors note as a barrier to proliferation Statistical analysis involving demographic data was per-formed and linear regression was utilized to analyze the variance in EHREMR interest and the amount of willingness to pay (Valdes et al 2004 )
One important study was done to assess intensive care unit (ICU) nursesrsquo accep-tance of EHR technology and examine the relationship between EHR design imple-mentation factors and user acceptance (Carayon et al 2011 ) This study was regional (northeastern USA) and local to the medical center and nurses working in four ICUs It tested only two major components of TAM usability (ease of use) and usefulness Three functionalities of EHR (computerized provider order entry (CPOE) the electronic medication administration record (eMAR) and nursing doc-umentation fl ow chart) were studied using multivariate hierarchical modeling The results showed that EHR usability and CPOE usefulness predicted EHR acceptance while looking at the periods of 3 and 12 months after implementation (Carayon et al 2011 )
L Hogaboam and TU Daim
133
One study of an outpatient primary care practice at the Western Pennsylvania hospital was conducted for research of social interactionsrsquo infl uence on physician adoption of EHR system (Zheng et al 2010 ) This empirical study involved 55 physiciansmdasha small sample size (most of them graduating or completing the residency program) The researchers used two SNA measures (ldquodensityrdquomdashldquothe number of social relations identifi ed divided by the total number of relations that could possibly be presentrdquo and ldquoFreemanrsquos degree centralityrdquomdashldquothe degree to which a social network is organized around its well-connected central networksrdquo) (Zheng et al 2010 ) Correlation method was used to capture the similarity between interaction patterns of pairs while quadratic assignment procedure (QAP) was used to test network correlations Network effects model (NEM) was used to evaluate the impact of social network structures on the measurements of the physicianrsquos utiliza-tion rates of the EHR system
The use of social contagion lens was engaged in a study of EHR adoption in US hospitals (Angst et al 2010 ) The researchers used the data from a nationwide annual survey of care delivery organizations in the USA (conducted by HIMSS Analytics) and applied the heterogeneous diffusion model technique for their hypothesis testing (Angst et al 2010 )
62 The Selection of EHR with Focus on Different Alternatives
In the study of EHR selection based on different alternatives certain gaps emerge from the body of literature
bull A comprehensive decision-making model of EHR selection in small ambulatory settings has not been successfully introduced andor implemented
bull Combination of elements of human criteria (perceived usefulness and ease of use) fi nancial technical organizational personal and interpersonal criteria in one decision-making model has not been performed
bull There is a lack of large-scale studies in the USA using HDM for EHR selection for small ambulatory setting
Ash and Bates ( 2005 ) indicate that comprehensive national surveys with a high response rate are not available and data in their study comes from the industry resources that may have some vested interests in EHR usage or selection The authors also indicate that small practices are less likely to adopt comparing to larger ones with various adoption gaps between the types of practices (pediatric internal medicine etc) Another interesting aspect provided by the authors is that there is a considerable amount of international experience (for example Sweden the Netherlands and Australia) that the USA can gain insights from (Ash amp Bates 2005 )
6 Review of Factors Impacting Decisions Regarding Electronic Records
134
In the selection of EHR the decision makers should consider factors that are environmental (fi nancial and safety social and behavioral) organizational per-sonal and technical (for example ability of systems to interoperate with each other) in nature (Ash amp Bates 2005 )
Study by Lorenzi et al stresses the need for fl exible change management strategy for EHR introduction in a small practice environment while detailing the EHR implementation through stages of decision selection pre-implementation imple-mentation and post-implementation (Lorenzi et al 2009 )
One important study about the attitudes of physicians toward EHR implementa-tion was performed by Morton and Wiedenbeck using the framework grounded in diffusion of innovations theory and TAM while being conducted at the University of Mississippi Medical Center (UMMC) (Morton amp Wiedenbeck 2009 ) The research-ers acknowledged that their fi ndings might not be generalized to other physicianrsquos offi ces since the study was limited to one large healthcare system however they revealed an overwhelming need for customizable and fl exible EHR products (Morton amp Wiedenbeck 2009 )
One important observational study on selection of EHR software discussed chal-lenges considerations and recommendations for identifying solutions mainly tar-geted toward small practices and presented fi ndings on installation training and use of EHR software as well as a detailed industry analysis of over 200 vendors and their offerings (Piliouras et al 2011 ) According to their analysis successful EHR system implementation has certain aspects (Piliouras et al 2011 )
bull The American Recovery and Reinvestment Act (ARRA) government mandates knowledge and conformance
bull Application of techniques in operations management systems analysis and change management
bull Learning EHR software bull Secure information technology infrastructure installation and maintenance bull Establishment of backup and disaster recovery procedures and processes
Piliouras et al (2011) also describe major challenges and recommendations
1 Conforming to ARRA mandates 2 Adherence to industry best practices 3 Installation and maintenance of secure IT infrastructure 4 Learning complex software
(a) Availability and quality of training (b) Quality software design
EHR systems could be either of a ldquoclient-serverrdquo or a ldquoservice-in-a-cloudrdquo infra-structure with the latter one with data maintained on dedicated vendor facilities and accessed over the Internet having capability of reducing capital outlay for computer and network infrastructure and associated upgrades and allowing expenditures to be
L Hogaboam and TU Daim
135
monetized as a fi xed monthly expense (Piliouras et al 2011) At the same time the practice needs to make sure that the vendor could satisfy the following criteria
bull Access privileges bull Regulatory compliance bull Data location bull Data segregation bull Data recovery bull Monitoring and reporting bull Vendor viability
The key differences between the two types of EHR software infrastructure taken from small practicersquos offi ce viewinterest are described in Table 65
Cloud computing in healthcare IT particularly for EHR also should not be con-sidered as a single concept with the same privacy and security concerns Zhang and
Table 65 Two types of EHR software infrastructure (Piliouras et al 2011)
Feature
Infrastructure type
Service-in-a-cloud Client-server
Location of system code and execution
Remote (mainly at vendorrsquos premise)
Local (mainly at doctorrsquos offi ce)
System data control Less More Same vendor system migrationextension
Easier Harder and more complex
Security More Less Hardware requirements Fewer More Response time Depends on the Internet
service provider (ISP) network provisioning and EHR vendor
Depends on the system maintenance and confi guration
Reliability Depends on the Internet service provider (ISP) network provisioning and EHR vendor
Depends on the system maintenance and confi guration backup and recovery process
Remote access via the Internet
Easy Possible with extra security measures
Maintenance Easier Harder Data synchronization for clinic with multiple offi ces
Easier Harder
Data backup and disaster recovery
Easier and cheaper Requires extra expense and technical support
Initial cost Lower Higher Total life cycle cost (3ndash5 years)
Lower Higher
6 Review of Factors Impacting Decisions Regarding Electronic Records
136
Table 66 Taxonomy of healthcare clouds (Zhang amp Liu 2010)
Healthcare cloud product layer
Explanation of capability for consumers
Control from the consumerrsquos side Security and privacy
Applications in the cloud (Software as a ServicemdashSaaS)
Can use the providerrsquos applications running on a cloud infrastructure
None Provided as an integral part of the system
Platforms in the cloud (Platform as a ServicemdashPaaS)
Can deploy consumer-created or -acquired applications written using supported programming languages and tools
No control over cloud infrastructure (network servers operating systems storage) control over the deployed applicationshosting environment confi gurations
Lower system levelmdashbasic security mechanisms (end-to-end encryption authentication and authorization) Higher system levelmdashthe consumers defi ne application- dependent access control policies authenticity requirements etc
Infrastructure in the cloud (Infrastructure as a ServicemdashIaaS)
Can provision processing storage networks and other fundamental computing sources to deploy and run arbitrary software operating systems and applications
No control over cloud infrastructure control over operating systems storage deployed applications possibly limited control of select networking components (host fi rewalls)
The healthcare application developers hold full responsibility
Liu ( 2010 ) provide taxonomy of healthcare clouds stressing those issues of privacy and security (Table 66 )
A very recent qualitative phenomenological study (ten interviews with physi-cians) in south-central Indiana looked into physicianrsquos view and perceptions of EHR which could help in the study of EHR selection (Hatton Schmidt amp Jelen 2012 ) Most reported and fi ltered challenges and benefi ts (Hatton et al 2012 ) are shown in Table 67
Roth et al ( 2009 ) also studied EHR use and stated that many EHR users may not always use EHR fully but only a fraction of EHR capabilities Some of the features and possibilities for documentation or structured recording of information may be ignored opted out or dismissed at the beginning of setup and use and the data may not be easily accessible through the automated extraction schemes when needed Free text fi elds (commonly used for patientsrsquo complaints) require natural language processing software While a lot has been accomplished in the area of natural lan-guage parsing and identifi cation many challenges still remain in the area of detec-tion of targeted clinical events from free text documents (Roth et al 2009 ) Through
L Hogaboam and TU Daim
137
the focus groups participating in the study the researchers learned that providers want EHR that requires less complexitymdasha minimum of keystrokes mouse clicks scrolling window changes etc While the fl exibility that accommodates various data entry styles has been built in it could complicate data extracting accuracy and effi ciency (Roth et al 2009 )
63 The Use of EHR with Focus on Impacts
Below are the gaps found through an extensive literature review of EHR impacts
bull The use of EHR in ambulatory settings and impact on quality of healthcare have not been adequately studied
bull The magnitude of the impacts from EHR use in the small ambulatory setting has not been adequately studied
bull The effects of user satisfaction and quality impacts in ambulatory settings are not adequately analyzed with quantitative measures
Table 67 Challenges and benefi ts of EHR (Hatton et al 2012 )
Challenges Benefi ts
Loss of control (major)
1 Procedural or workfl ow challenges 2 The EMR causing them to work slowly 3 The pace of technology obsolescence 4 Too much information is available to
patients or needs to be gathered from patients
5 The cognitive distraction during physicianrsquos use of the computer in the examination room
Supporting physician decisions (major) (particularly useful in noting drug allergies and drug-to-drug interactions)
Attitude of providers
1 Sense that paper charts were easier than electronic records
2 Technical ability of the physician or lack of it
3 Physicianrsquos age
Physician access to information (major) (structured and retrievable format integrating patient data so that demographic fi nancial and medical information could be accessed transmitted and stored in a digital format)
Financial negatives
1 Cost of the software 2 Cost of maintenance 3 Cost of the support personnel
Financial improvements (major) (sense that EMR makes them cost effective and more effi cient being proactive with patients increases patient loads getting government incentives opportunities for data mining)
Continuity of care (referrals and care coordination)
Time improvements (improved communication with staff though the EMR messaging capability) Patient access to information (better informed patients could provide opportunities for improved care which could also lead to healthier outcomes)
6 Review of Factors Impacting Decisions Regarding Electronic Records
138
bull There is a lack of large-scale studies in the USA using HDM for EHR impacts in small ambulatory setting
While the attention of greater quality of care always persists with research focus on how providers patients and policies could affect factors that infl uence the quality of care despite high investments (over 17 trillion annually) and increased healthcare spending the USA ranks lower compared to other countries on several health measures (Jung 2006 Girosi Meili amp Scoville 2005 ) Jung listed specifi c benefi ts of HIT in regard to quality of care
bull Medical error reduction (improved communication and access to information through information systems could have a great impact in this area)
bull Adherence support (the decision support functions embedded in EHR can show the effect of HIT on adherence to guideline-based care and enhancing preventive healthcare delivery (Dexter et al 2004 Overhage 1996 Jung 2006 )
bull Effective disease management (potential to improving the health outcomes of patients with specifi c diseases)
Jung ( 2006 ) also explained that while effi ciency is a complex concept some effi ciency savings have been reported by researchers as a result of HIT adoption as reduction in administrative time (Wong 2003 Jung 2006 ) and hospital stays Positive effects on cost were documented as
bull Improved productivity bull Paper reduction bull Reduced transcription costs bull Drug utilization bull Improved laboratory tests
Additional benefi ts reported by several (Bates et al 1998 Agarwal 2002 Jung 2006 ) were as follows
bull Improved patient safety (from safety alerts and medication reminders of EHR system)
bull Improved regulatory compliance (record keeping and reporting compliance with federal regulations including Health Insurance Portability and Accountability Act (HIPAA))
Increased emphasis on preventive measures and early detection of diseases primary care intermittent healthcare services and continuity of care are prevalent in our ever-changing healthcare domain (Tsiknakis Katehakis amp Orphanoudakis 2002 ) Information and communication technologies are taking lead in this dynamic environment with the need for improved quality of healthcare services and cost control (Tsiknakis et al 2002 ) Another important trend in the healthcare system is movement toward shared and integrated care (integrated electronic health recordmdashiEHR) growth of home care through sophisticated telemedicine services (facili-tated by intelligent sensors handheld technologies monitoring devices wireless technologies and the Internet) which pushes the need for EHR that supports qual-ity and continuity of care (Tsiknakis et al 2002 ) While the researchers enlisted a
L Hogaboam and TU Daim
139
number of valuable benefi ts they would need to be examined and the relationships of EHR impacts and their signifi cance would need to be studied further The envi-sioned benefi ts are listed in Fig 61 and Table 68
A systematic review by Goldzweig lists only a few studies of commercial health IT system use with reported results and experiences of the impacts of EHR imple-mentation (Goldzweig et al 2009 ) In one of the studies described in their publica-tion authors concluded that EHR implementation (EpiCare at Kaiser Northwest) had no negative impact on quality of care measures of quality like immunizations and cancer screening did not change (Goldzweig et al 2009 ) In the second study of implementation of a commercial EHR in a rural family practice in New York the authors report various fi nancial impacts (average monthly revenue increase due to better billing practices) clinical practice satisfaction as well as the support of the core mission of providing care
Agency for Healthcare Research Quality defi ned quality healthcare as ldquodoing the right thing at the right time in the right way to the right person and having the best pos-sible resultsrdquo (Agency for Healthcare Research Quality 2004 in Kazley amp Ozcan 2008 )
One important retrospective study in the USA by Kazley and Ozcan looked at EMR impacts on quality performance in acute care hospitals (Kazley amp Ozcan 2008 ) Retrospective cross-sectional format with linear regression is used in order to assess the relationship between hospital EMR use and quality performance (Kazley amp Ozcan 2008 ) The authors concluded that there is a limited evidence of the relationship between EMR use and quality There are some interesting observa-tions made by the authors toward measuring quality and they describe it as a multi-
Vital health informaon is
available 24 hrs a day 7 days a
week regardless of the paents
locaon
Healthcare praccioners are able to view paents relevant medical historybullmore effecve
and efficient treatment
bullmore quality me spent with the paent
Access to informaon of previous lab results or medical procedurebullreduce the
number of redundant procedure
bullresults in greater cost savings
Enhanced ability of health planners and administrators to develop relevant healthcare policies with EHR informaonbullinformaon for
researchersbullpopulon health
stascsbullimproved quality
of care
Access to individuals own personal health recordsbullindividuals can
make informed choices about opons available
bullopportunity to excercise greater control over their health
Fig 61 Envisioned EHR benefi ts
6 Review of Factors Impacting Decisions Regarding Electronic Records
140
faceted and complex construct which may grow and change Ten process indicators related to three clinical conditions acute myocardial infarction congestive heart failure and pneumonia are used to measure quality performance based on their validity (Kazley amp Ozcan 2008 ) The authors noted that they didnrsquot measure such elements of quality as patient satisfaction and long-term outcomes and that EMR implementation and practice should be further explored
Leu et al ( 2008 ) performed a qualitative study with in-depth semi-structured inter-views to describe how health IT functions within a clinical context Six clinical domains were identifi ed by the researchers result management intra-clinic communication patient education and outreach inter-clinic coordination medical management and provider education and feedback Created clinical process diagrams could provide clinicians IT and industry with a common structure of reference while discussing health IT systems through various time frames (Leu et al 2008 )
Table 68 Potential benefi ts and their related features
Potential benefi t Related EHR features
Dissemination and distribution of essential patientclient information
Open communication standards over transparent platforms
Improved protection of personal data Encryption and authentication mechanisms for secure access to sensitive personal information auditing capabilities for tracking purposes
Informed decision making resulting in improved quality of care
Semantic unifi cation and multimedia support for a more concise and complete view of medical history
Prompt and appropriate treatment Fast response times through transparent networks and open interfaces
Risk reduction (access to a wider patientclient knowledge base)
Appropriate usable human-computer interfaces through awareness of contextual factors
Facilitation of cooperation between health professionals of different levels of health social care organization
Role-based access mechanisms and access privileges
Reduction in duplicate recordingquestioning of relevant patient information
A robust and scalable interface (HII) that could extend from corporatehospital to regional and national level
More focused and appropriate use of resources due to shared information of assessment and care plan
Access to all diagnostic information through adaptive user interfaces
Improved communication between professionals
Multimedia information is in the best format by clinical information system for communication without loss of quality
Security and guarantee of continuity of care Permanent access and control of interventions Identifi cation of a single patient across multiple systems
Mechanism for identifying a single client record and associated data that may have been stored on various source systems
Consistent shared language (between professionals)
Mapping tool to display information in a generic format to bridge the gap in terminology and semantic differences
L Hogaboam and TU Daim
141
Results of 2003 and 2004 National Ambulatory Medical Care Survey indicated that electronic health records were used in 18 of estimated 18 billion ambulatory visits in the USA for years 2003 and 2004 (Linder et al 2007 ) The researchers stated that despite the large number of patient records the sample size was small for some of the used quality indicators The study didnrsquot identify the implementation barriers for such low computerized registry use but outlined 17 ambulatory quality indicators and while some quality indicators showed signifi cance for quality of care the researchers didnrsquot fi nd consistent association between EHR and the quality of ambulatory care The main categories (Linder et al 2007 ) of researched indica-tors were the following
bull Medical management of common diseases (EHR had positive effect on aspirin use for coronary artery disease (CAD) but worse effect on antithrombotic ther-apy for atrial fi brillation (AF))
bull Recommended antibiotic use bull Preventive counseling bull Screening tests bull Avoiding potentially inappropriate prescribing in elderly patients
While it would be expected that EHR-extracted data would allow quality assess-ment and other impact assessment without expensive and time-consuming process-ing of medical documentation some researchers (Roth et al 2009 ) conclude that only about a third of indicators of the quality assessment tools system would be readily available through EHR with some concerns that only components of quality would be measured perhaps to the detriment of other important measures of healthcare quality The researchers provided a table of accessibility of quality indicators (clinical variables) which have been narrated in Table 69
A group of researchers looked into the problem of improving patient safety in ambulatory settings and throughout this qualitative study developed a tool kit of best practices and a collaborative to enhance medication-related practices and patient safety standards (Schauberger amp Larson 2006 ) The list of best practices for the inpatient setting was the following with 6 10 and 3 being the top three pro-cess improvements on best practices
1 Maintaining accurate and complete medication list 2 Ensuring medication allergy documentation 3 Standardizing prescription writing 4 Removing all IV potassium chloride from all locations 5 Emphasizing non-punitive error reporting 6 Educating about look-alike sound-alike drugs 7 Improving verbal orders 8 Ensuring safety and security of sample drugs 9 Following protocols for hazardous drug use 10 Partnering with patients 11 Notifying patients of laboratory results
Figures 62 63 and 64 summarize this chapter
6 Review of Factors Impacting Decisions Regarding Electronic Records
142
Table 69 Accessibility of quality indicators
Accessible indicators (most to least) Hard-to-access indicators (most to least)
Demographics Disease-specifi c history Diagnosis Care site Prescription Physical exam Past medical history Refusal Procedure date Patient education Lab date Social history Problemchief complaint Treatment Vital signweightheight Diagnostic test result Allergy Imaging result Lab result Contraindication Medication history Pathology Diagnostic test date Family history Imaging date EKG result Medications current X-ray result Vaccination X-ray date EKG date
1
Research Gaps Research Goals Research Questions
The impact and significance of implementation barriers and enablers has not been satisfactorily studied
Significance of the relationship of factors of perceived usefulness perceived ease of use and perceived benefits on attitude toward using EHR in ambulatory settings has not been adequately shown with global studies
Lack of large-scale studies in the United States withTAM models application for small ambulatory setting
Lack of quantitative studies engaging SEM on a large scale for small clinics
Define a research framework for impact of EHR barriers and enablers on adoption of EHR system in small ambulatory settings
Assess the impact of barriers and enablers on framework components of EHR adoption in small ambulatory settings
What factors impact perceived ease of use perceived usefulness and perceived benefits in small clinics
Do interpersonal factors have any direct or indirect impacts
Do factors of perceived usefulness ease of use and benefits significantly impact EHR use in small ambulatory settings
Do subjective norms and attitudes impact intention to use EHR
Does perceived ease of use have a significant impact on perceived usefulness in small clinics
What is the impact significance of intention to use EHR into EHR use
Fig 62 Research gaps goals and questions for the adoption of EHR with focus on barriers and enables
L Hogaboam and TU Daim
Research Gaps Research Goals Research QuestionsA comprehensive decision-making model of EHR selection in small ambulatory settings has not been successfully introduced andor implemented
Combination of elements of human criteria (perceived usefulness and ease of use) financial technical organizational personal and interpersonal criteria in one decision-making model has not been performed
There is a lack of large-scale studies in the United States using HDM for EHR selection for small ambulatory setting
Define a research framework for EHRselection in small ambulatory settings
Assess the importance of criteria and subcriteria and the lower level of HDM through expert judgment quantification
Do criteria of perceived usefulness and ease of use play a significant role in EHR selection
Do interpersonal factors matter in selection of EHR software
Do financial factors impact the decision-making of EHR software in a significant way
Do organizational factors strongly influence decision-making in EHR selection process
Do personal factors of productivity and privacy play an important role in selection of EHR software
Fig 63 Research gaps goals and questions for the selection of EHR with focus on different alternatives
Research Gaps Research Goals Research QuestionsThe use of EHR in ambulatory settings andimpact on quality of healthcare has not been adequately studied
The magnitude of the impacts from EHR use in the small ambulatory setting has not been adequately studied
The effects of user satisfaction and quality impacts in ambulatory settings are not adequately analyzed with quantitative measures
Define a research framework relating EHR use in small ambulatory settings with comprehensive impacts hierarchy including quality criteria
Assess the impact of criteria and subcriteria of the model as a result of EHR use in ambulatory settings from physicianrsquos point of view
Which quality measures (system information or service) have higher importance from physicianrsquos point of view
Does EHR use greatly impacts organizational criteria of structure and environment
From physicianrsquos point of view does EHR use improve clinical outcomes andor save costs
There is a lack of large-scale studies in the United States using HDM for EHR impacts in small ambulatory setting
Fig 64 Research gaps goals and questions for the use of EHR with focus on impacts
144
References
Agarwal A (2002) Return on investment analysis for a computer-based patient record in the outpatient clinic setting Journal of the Association for Academic Minority Physicians 13 (3) 61
Aggelidis V P amp Chatzoglou P D (2009) Using a modifi ed technology acceptance model in hospitals International Journal of Medical Informatics 78 (2) 115ndash126 Retrieved October 29 2012 from httpwwwncbinlmnihgovpubmed18675583
Andreacute B et al (2008) Experiences with the implementation of computerized tools in health care units A review article International Journal of Human-Computer Interaction 24 (8) 753ndash775 Retrieved November 12 2012 from httpwwwtandfonlinecomdoiabs10108010447310802205768
Angst C M et al (2010) Social contagion and information technology diffusion The adoption of electronic medical records in US hospitals Management Science 56 (8) 1219ndash1241 Retrieved November 12 2012 from httpmanscijournalinformsorgcgidoi101287mnsc11001183
Ash J amp Bates D (2005) Factors and forces affecting EHR system adoption Report of a 2004 ACMI discussion Journal of the American Medical Informatics 12 8ndash13 Retrieved May 15 2012 from httpwwwsciencedirectcomsciencearticlepiiS1067502704001495
Ayatollahi H Bath P A amp Goodacre S (2009) Paper-based versus computer-based records in the emergency department Staff preferences expectations and concerns Health Informatics Journal 15 (3) 199ndash211 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed19713395
Bates D W et al (1998) Effect of computerized physician order entry and a team intervention on prevention of serious medication errors The Journal of the American Medical Association 280 (15) 1311ndash1316 httpwwwncbinlmnihgovpubmed9794308
Bates D W et al (2003) A proposal for electronic medical records in US primary care Journal of American Informatics Association 10 (1) 1ndash10
Becker A et al (2011) A new computer-based counselling system for the promotion of physical activity in patients with chronic diseasesmdashResults from a pilot study Patient Education and Counseling 83 (2) 195ndash202 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed20573467
Boonstra A amp Broekhuis M (2010) Barriers to the acceptance of electronic medical records by physicians from systematic review to taxonomy and interventions BMC Health Services Research 10 231 httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2924334amptool=pmcentrezamprendertype=abstract
Burton-Jones A amp Hubona G S (2006) The mediation of external variables in the technology acceptance model Information and Management 43 (6) 706ndash717 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0378720606000504
Carayon P et al (2011) ICU nursesrsquo acceptance of electronic health records Journal of the American Medical Informatics Association 18 (6) 812ndash819 Retrieved November 8 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=3197984amptool=pmcentrezamprendertype=abstract
Chen R-F amp Hsiao J-L (2012) An investigation on physiciansrsquo acceptance of hospital infor-mation systems A case study International Journal of Medical Informatics (60) 1ndash11 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed22652011
Cheng Y-M (2012) Exploring the roles of interaction and fl ow in explaining nursesrsquo e-learning acceptance Nurse Education Today Retrieved November 10 2012 from httpwwwncbinlmnihgovpubmed22405340
Chiasson M et al (2007) Expanding multi-disciplinary approaches to healthcare information technologies What does information systems offer medical informatics International Journal of Medical Informatics 76 Suppl 1 S89ndashS97 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed16769245
L Hogaboam and TU Daim
145
Chow M Chan L et al (2012) Exploring the intention to use a clinical imaging portal for enhancing healthcare education Nurse Education Today 1ndash8 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed22336478
Chow M Herold D K et al (2012) Extending the technology acceptance model to explore the intention to use Second Life for enhancing healthcare education Computers and Education 59 (4) 1136ndash1144 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0360131512001327
Cresswell K amp Sheikh A (2012) Organizational issues in the implementation and adoption of health information technology innovations An interpretative review International Journal of Medical Informatics Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS1386505612001992
Davidson S amp Heineke J (2007) Toward an effective strategy for the diffusion and use of clini-cal information systems Journal of the American Medical Informatics Association 14 (3) 361ndash367 Retrieved November 12 2012 from http17167114118content143361abstract
Degoulet P Jean F C amp Safran C (1995) The health care professional multimedia worksta-tion Development and integration issues International Journal of Bio-medical Computing 39 (1) 119ndash125 httpwwwncbinlmnihgovpubmed7601524
Dexter P R et al (2004) Inpatient computer-based standing orders vs physician reminders to increase infl uenza and pneumococcal vaccination rates A randomized trial The Journal of the American Medical Association 292 (19) 2366ndash2371 httpwwwncbinlmnihgovpubmed15547164
Dillon A amp Morris M G (1996) User acceptance of new information technologymdashTheories and models In M Williams (Ed) Annual review of information science and technology (Vol 31 pp 3ndash32) Medford NJ Information Today
Duumlnnebeil S et al (2012) Determinants of physiciansrsquo technology acceptance for e-health in ambulatory care International Journal of Medical Informatics 81 (11) 746ndash760 Retrieved November 6 2012 from httpwwwncbinlmnihgovpubmed22397989
Folland S (2006) Health care in small areas of three command economies What do the data tell us Eastern European Economics 43 (6) 31ndash52 httpmesharpemetapresscomopenurlaspgenre=articleampid=doi102753EEE0012-8755430602
Girosi F Meili R amp Scoville R (2005) Extrapolating evidence of health information technol-ogy savings and costs pub no MG-410 Santa Monica CA
Goldzweig C L et al (2009) Costs and benefi ts of health information technology New trends from the literature Health Affairs (Project Hope) 28 (2) w282ndashw293 Retrieved March 29 2012 from httpwwwncbinlmnihgovpubmed19174390
Hagger M S et al (2007) Aspects of identity and their infl uence on intentional behavior Comparing effects for three health behaviors Personality and Individual Differences 42 (2) 355ndash367 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0191886906002881
Handy J Hunter I amp Whiddett R (2001) User acceptance of inter-organizational electronic medical records Health Informatics Journal 7 (2) 103ndash107 Retrieved November 12 2012 httpjhisagepubcomcgidoi101177146045820100700208
Haron S N Hamida M Y amp Talib A (2012) Towards healthcare service quality An under-standing of the usability concept in healthcare design ProcediamdashSocial and Behavioral Sciences 42 (July 2010) 63ndash73 Retrieved November 12 2012 httplinkinghubelseviercomretrievepiiS187704281201049X
Hatton J D Schmidt T M amp Jelen J (2012) Adoption of electronic health care records Physician heuristics and hesitancy Procedia Technology 5 706ndash715 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS2212017312005099
Hung S-Y Ku Y-C amp Chien J-C (2012) Understanding physiciansrsquo acceptance of the Medline system for practicing evidence-based medicine A decomposed TPB model International Journal of Medical Informatics 81 (2) 130ndash142 Retrieved November 5 2012 from httpwwwncbinlmnihgovpubmed22047627
6 Review of Factors Impacting Decisions Regarding Electronic Records
146
Im I Kim Y amp Han H-J (2008) The effects of perceived risk and technology type on usersrsquo acceptance of technologies Information and Management 45 (1) 1ndash9 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0378720607000468
Janczewski L amp Shi F X (2002) Development of information security baselines for health-care information systems in New Zealand Computers and Security 21 (2) 172ndash192 Retrieved November 12 2012 from httpwwwsciencedirectcomsciencearticlepiiS0167404802002122
Jeng D J-F amp Tzeng G-H (2012) Social infl uence on the use of Clinical Decision Support Systems Revisiting the unifi ed theory of acceptance and use of technology by the fuzzy DEMATEL technique Computers and Industrial Engineering 62 (3) 819ndash828 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0360835211003895
Jimoh L et al (2012) A model for the adoption of ICT by health workers in Africa International Journal of Medical Informatics 81 (11) 773ndash781 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed22986218
Jung S (2006) The perceived benefi ts of healthcare information technology adoption Construct and survey development Retrieved March 22 2013 from httpetdlsuedudocsavailableetd-11162006-125102
Karahanna E amp Straub D W (1999) The psychological origins of perceived usefulness and ease-of-use Information and Management 35 (4) 237ndash250 httplinkinghubelseviercomretrievepiiS0378720698000962
Kazley A S amp Ozcan Y A (2008) Do hospitals with electronic medical records (EMRs) pro-vide higher quality care An examination of three clinical conditions Medical Care Research and Review 65 (4) 496ndash513 Retrieved May 14 2012 from httpwwwncbinlmnihgovpubmed18276963
Kim S amp Malhotra N (2005) A longitudinal model of continued IS use An integrative view of four mechanisms underlying postadoption phenomena Management Science 51 (5) 741ndash755 Retrieved November 12 2012 from httpmanscijournalinformsorgcontent515741short
Lee G amp Xia W (2011) A longitudinal experimental study on the interaction effects of persua-sion quality user training and fi rst-hand use on user perceptions of new information technol-ogy Information and Management 48 (7) 288ndash295 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0378720611000772
Leu M G et al (2008) Centers speak up The clinical context for health information technology in the ambulatory care setting Journal of General Internal Medicine 23 (4) 372ndash378 Retrieved March 1 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2359517amptool=pmcentrezamprendertype=abstract
Linder J A et al (2007) Electronic health record use and the quality of ambulatory care in the United States Archives of Internal Medicine 167 (13) 1400ndash1405 httpwwwncbinlmnihgovpubmed17620534
Lorenzi N M et al (2009) How to successfully select and implement electronic health records (EHR) in small ambulatory practice settings BMC Medical Informatics and Decision Making 9 (15) 1ndash13 Retrieved May 14 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2662829amptool=pmcentrezamprendertype=abstract
Ludwick D A amp Doucette J (2009) Adopting electronic medical records in primary care Lessons learned from health information systems implementation experience in seven coun-tries International Journal of Medical Informatics 78 (1) 22ndash31 Retrieved February 29 2012 from httpwwwncbinlmnihgovpubmed18644745
Maumlenpaumlauml T et al (2009) The outcomes of regional healthcare information systems in health care A review of the research literature International Journal of Medical Informatics 78 (11) 757ndash771 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed19656719
Malhotra Y (1999) Bringing the adopter back into the adoption process A personal construction framework of information technology adoption The Journal of High Technology Management Research 10 (1) 79ndash104 httplinkinghubelseviercomretrievepiiS1047831099800042
L Hogaboam and TU Daim
147
Martich G amp Cervenak J (2007) Eyes wide shut The ldquohiddenrdquo costs of deploying health infor-mation technology Journal of Critical Care 7ndash8 Retrieved November 12 2012 from httpwwwjournalselsevierhealthcomperiodicalsyjcrcarticleS0883-9441(06)00217-6abstract
McFarland D J amp Hamilton D (2006) Adding contextual specifi city to the technology accep-tance model Computers in Human Behavior 22 (3) 427ndash447 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS074756320400130X
McGinn C A et al (2011) Comparison of user groupsrsquo perspectives of barriers and facilitators to implementing electronic health records A systematic review BMC Medicine 9 (46) 1ndash10 httpwwwpubmedcentralnihgovarticlerenderfcgiartid=3103434amptool=pmcentrezamprendertype=abstract
Melas C D et al (2011) Modeling the acceptance of clinical information systems among hospi-tal medical staff An extended TAM model Journal of Biomedical Informatics 44 (4) 553ndash564 Retrieved November 7 2012 from httpwwwncbinlmnihgovpubmed21292029
Melone N (1990) A theoretical assessment of the user-satisfaction construct in information sys-tems research Management Science 36 (1) 76ndash91 Retrieved November 12 2012 from httpmanscijournalinformsorgcontent36176short
Moores T T (2012) Towards an integrated model of IT acceptance in healthcare Decision Support Systems 53 (3) 507ndash516 Retrieved November 12 2012 from httplinkinghubelse-viercomretrievepiiS0167923612001108
Morton M E amp Wiedenbeck S (2009) A framework for predicting EHR adoption attitudes A physician survey Perspectives in Health Information ManagementAHIMA American Health Information Management Association 6 1 httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2804456amptool=pmcentrezamprendertype=abstract
Ortega Egea J M amp Romaacuten Gonzaacutelez M V (2011) Explaining physiciansrsquo acceptance of EHCR systems An extension of TAM with trust and risk factors Computers in Human Behavior 27 (1) 319ndash332 Retrieved November 7 2012 from httplinkinghubelseviercomretrievepiiS0747563210002530
Overhage J M (1996) Computer reminders to implement preventive care guidelines for hospital-ized patients Archives of Internal Medicine 156 (14) 1551
Pai F-Y amp Huang K-I (2011) Applying the Technology Acceptance Model to the introduction of healthcare information systems Technological Forecasting and Social Change 78 (4) 650ndash660 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0040162510002714
Palacio C Harrison J P amp Garets D (2009) Benchmarking electronic medical records initia-tives in the US A conceptual model Journal of Medical Systems 34 (3) 273ndash279 Retrieved May 12 2012 from httpwwwspringerlinkcomindex101007s10916-008-9238-5
Pareacute G amp Sicotte C (2001) Information technology sophistication in health care An instrument validation study among Canadian hospitals International Journal of Medical Informatics 63 (3) 205ndash223 httpwwwncbinlmnihgovpubmed11502433
Piliouras Teresa (Raymond) Yu Pui Lam Huang Housheng Liu Xin Kumar Vijay Siddaramaiah Ajjampur Sultana Nadia Selection of electronic health records software Challenges considerations and recommendations Systems Applications and Technology Conference (LISAT) 2011 IEEE Long Island Issue Date 6ndash6 May 2011
Polančič G Heričko M amp Rozman I (2010) An empirical examination of application frame-works success based on technology acceptance model Journal of Systems and Software 83 (4) 574ndash584 Retrieved October 26 2012 from httplinkinghubelseviercomretrievepiiS0164121209002799
Premkumar G amp Bhattacherjee A (2008) Explaining information technology usage A test of competing models Omega 36 (1) 64ndash75 Retrieved November 5 2012 from httplinkinghubelseviercomretrievepiiS0305048305001702
Rosemann T et al (2010) Utilisation of information technologies in ambulatory care in Switzerland Swiss Medical Weekly 140 (September) w13088 Retrieved April 20 2012 from httpwwwncbinlmnihgovpubmed20853193
6 Review of Factors Impacting Decisions Regarding Electronic Records
148
Roth C P et al (2009) The challenge of measuring quality of care from the electronic health record American Journal of Medical Quality 24 (5) 385ndash394 Retrieved May 14 2012 from httpwwwncbinlmnihgovpubmed19482968
Schauberger C W amp Larson P (2006) Implementing patient safety practices in small ambula-tory care settings Journal on Quality and Patient Safety 32 (8) 419ndash425
Scott P J amp Briggs J S (2009) A pragmatist argument for mixed methodology in medical informatics Journal of Mixed Methods Research 3 (3) 223ndash241 Retrieved November 12 2012 from httpmmrsagepubcomcgidoi1011771558689809334209
Shin D-H (2010) The effects of trust security and privacy in social networking A security- based approach to understand the pattern of adoption Interacting with Computers 22 (5) 428ndash438 Retrieved November 4 2012 from httplinkinghubelseviercomretrievepiiS0953543810000494
Storey J amp Buchanan D (2008) Healthcare governance and organizational barriers to learning from mistakes Journal of Health Organisation and Management 22 (6) 642ndash651 Retrieved November 12 2012 from httpwwwemeraldinsightcom10110814777260810916605
Szajna B (1996) Empirical evaluation of the revised technology acceptance model Management Science 42 (1) 85ndash92 Retrieved November 12 2012 from httpmanscijournalinformsorgcontent42185short
Tsiknakis M Katehakis D G amp Orphanoudakis S C (2002) An open component-based information infrastructure for integrated health information networks International Journal of Medical Informatics 68 (1-3) 3ndash26 httpwwwncbinlmnihgovpubmed12467787
Valdes I et al (2004) Barriers to proliferation of electronic medical records Informatics in Primary Care 12 3ndash9 Retrieved May 15 2012 from httpwwwingentaconnectcomcon-tentrmpipc20040000001200000001art00002
Van Schaik P et al (2004) The acceptance of a computerised decision-support system in primary care A preliminary investigation Behaviour and Information Technology 23 (5) 321ndash326 Retrieved November 12 2012 from httpwwwtandfonlinecomdoiabs1010800144929041000669941
Vishwanath A Brodsky L amp Shaha S (2009) Physician adoption of personal digital assistants (PDA) Testing its determinants within a structural equation model Journal of Health Communication 14 (1) 77ndash95 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed19180373
Viswanathan S (2005) Competing across technology-differentiated channels The impact of net-work externalities and switching costs Management Science 51 (3) 483ndash496 Retrieved November 12 2012 from httpmanscijournalinformsorgcontent513483short
Were M C et al (2010) Evaluating a scalable model for implementing electronic health records in resource-limited settings Journal of the American Medical Informatics Association 17 (3) 237ndash244 Retrieved March 15 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2995711amptool=pmcentrezamprendertype=abstract
Wong D H (2003) Changes in intensive care unit nurse task activity after installation of a third- generation intensive care unit information system Critical Care Medicine 31 (10) 2488
Yang H (2004) Itrsquos all about attitude Revisiting the technology acceptance model Decision Support Systems 38 (1) 19ndash31 Retrieved November 9 2012 from httpportlandstateworld-catorgtitleits-all-about-attitude-revisiting-the-technology-acceptance-modeloclc198488645ampreferer=brief_results
Yu P Li H amp Gagnon M-P (2009) Health IT acceptance factors in long-term care facilities A cross-sectional survey International Journal of Medical Informatics 78 (4) 219ndash229 Retrieved November 7 2012 from httpwwwncbinlmnihgovpubmed18768345
Yusof M M et al (2008) An evaluation framework for Health Information Systems Human organization and technology-fi t factors (HOT-fi t) International Journal of Medical Informatics 77 (6) 386ndash398 Retrieved October 29 2012 from httpwwwncbinlmnihgovpubmed17964851
L Hogaboam and TU Daim
149
Rui Zhang and Ling Liu ldquoSecurity Models and Requirements for Healthcare Application Cloudsrdquo Proceedings of the 3rd IEEE International Conference on Cloud Computing (Cloud 2010) July5ndash10 2010 Miami Florida USA
Zheng K et al (2010) Social networks and physician adoption of electronic health records Insights from an empirical study Journal of the American Medical Informatics Association 17 (3) 328ndash336 Retrieved March 5 2012 from httpwwwpubmedcentralnihgovarticleren-derfcgiartid=2995721amptool=pmcentrezamprendertype=abstract
6 Review of Factors Impacting Decisions Regarding Electronic Records
151copy Springer International Publishing Switzerland 2016 TU Daim et al Healthcare Technology Innovation Adoption Innovation Technology and Knowledge Management DOI 101007978-3-319-17975-9_7
Chapter 7Decision Models Regarding Electronic Health Records
Liliya Hogaboam and Tugrul U Daim
71 The Adoption of EHR with Focus on Barriers and Enables
Modifications to the models and extensions also have roots in theoretical back-ground and have proven to be effective in studying various cases of IT adoption under various conditions Knowledge of specific implementation barriers and their impact and statistical significance on the improvement of EHR use could lead to the creation of guidelines and incentives toward elimination of those barriers in ambula-tory settings Focused incentives training and education in the right direction could speed up the process of adoption and use of computerized registries as well as implementation of more sophisticated IT systems (Miller amp Sim 2004)
711 Theory of Reasoned Action
In their study of perceived behavioral control and goal-oriented behavior Ajzen and Fishbein proposed TRA (Ajzen amp Madden 1986) The fundamental point of TRA is that the immediate precedent of any behavior is the intention to perform behavior in question Stronger intention increases the likelihood of performance of the action according to the theory (Ajzen amp Madden 1986) Two conceptually independent determinants of intention are specified by TRA attitude toward the behavior (the degree to which an individual has favorable evaluation of behavior in mind or oth-erwise) and subjective norm (perceived social pressure whether the behavior should
L Hogaboam bull TU Daim () Department of Engineering and Technology Management Portland State University SW 4th Ave Suite LL-50-02 1900 97201 Portland OR USAe-mail liliyanascentiacom tugruludaimpdxedu
152
be performed or not ie acted upon or not) TRA also states that the behavior is a function of behavioral beliefs and normative beliefs which are relevant to behavior (Ajzen amp Madden 1986)
Atude toward the behavior
Subjec13ve norm
Inten13on Behavior
712 Technology Acceptance Model
In 1985 Fred Davis presented his work that was centered toward improving the understanding of user acceptance process for successful design and implementation of information systems and providing theoretical basis for a practical methodology of ldquouser acceptancerdquo through TAM which could enable implementers and system designers to evaluate proposed systems (Davis 1985) Perceived usefulness and perceived use are outlined to be the main two variables influencing attitude toward using the system Perceived usefulness is ldquothe degree to which individual believes that using a particular system would enhance his or her job performancerdquo Perceived ease of use is ldquothe degree to which an individual believes that using a particular system would be free of physical and mental effortrdquo Davis also shows that per-ceived ease of use has a causal effect on the variable of perceived usefulness (Davis 1985 Davis amp Venkatesh 1996)
Conceptual framework from Davis is shown in Fig 71His proposed model sheds light on the behavioral part of the concept with over-
all attitude of a potential user toward system use being a main determinant of the systemrsquos use On the other hand perceived usefulness and perceived use are out-lined to be the main two variables influencing attitude toward using the system Perceived usefulness is ldquothe degree to which individual believes that using a particu-lar system would enhance his or her job performancerdquo Perceived ease of use is ldquothe degree to which an individual believes that using a particular system would be free of physical and mental effortrdquo He argues that system that is easier to use will result in increased job performance and greater usefulness for the user all else being equal Davis also shows that perceived ease of use has a causal effect on the variable of
L Hogaboam and TU Daim
153
perceived usefulness (Davis 1985 Davis amp Venkatesh 1996) While ease of use is important with a lot of emphasis on user friendliness of the applications that increase usability no amount of ease of use could compensate for the reality of the useful-ness of the system (Davis 1993) Causal relationships in the model are represented by arrows (Fig 72) Attitude toward use is referred to as the degree of evaluative effect that an individual associates with using the target system in hisher job while actual system use is the individualrsquos direct usage of the given system (Davis 1985 Davis amp Venkatesh 1996)
Described mathematically TAM will look like this (Davis 1985)
Perceived easeof use EOU Xi n
i i( ) = +=aring1
b e
(71)
Perceived usefulness USEF iX EOUi n
i n( ) = + +=
+aring1
1
b b e
(72)
Attitude toward using ATT EOU USEF( ) = + +b b e1 2
(73)
Actual useof thesystem USE ATT( ) = +b e1
(74)
System Features and Capabili13es
Users Mo13va13on
to Use System
Actual System Use
S13mulus Organism Response
Fig 71 Conceptual framework for building TAM (Davis 1985)
x1
x2
Perceived Usefulness
Atude Toward Using
Actual System Use
Perceived Ease of Usex3
User Movaon
Design Features
Cognive Response
Affecve Response
Behavioral Response
Fig 72 Technology acceptance model (Davis 1985)
7 Decision Models Regarding Electronic Health Records
154
where
Xi is a design feature I i = 1hellipnβi is a standardized partial regression coefficientε is a random regression term
713 Theory of Planned Behavior
TPB extends TRA by including the concept of behavioral control The importance of control could be observed through the fact that the resources and opportunities available to individuals have to dictate to some extent the likelihood of behavioral achievement (Ajzen amp Madden 1986) According to the TPB a set of beliefs that deals with the presence or absence of requisite resources and opportunities could ultimately determine intention and action The more opportunities and resources individuals think they possess the fewer obstacles they anticipate and the greater their perceived control over behavior should be (Ajzen amp Madden 1986) (Fig 73)
Holden amp Karsh (2010) analyzed studies where TAM was used and compared the percentage of variance explained by this theoretical framework The percentage varies from 30 to 70 but in most cases tested in healthcare the percentage of variance is higher than 40 which means that the model explains at least 40 of phenomenon
The proposed framework for assessing EHR adoption in ambulatory settings has elements of TAM TRA and TPA along with important elements described in the literature that were frequently mentioned showed significant relationships or were expressed in qualitative and quantitative way This framework consists of barriers and enablers since some of those variables might have a positive influence on the system use The concepts of perceived ease of use and perceived usefulness and subjective norm have been explained earlier in this part of the exam The external factors have been constructed through the comprehensive literature review during the independent studies and the short and extended version of external element con-structs is shown in Fig 74
Extended taxonomy is listed in Table 71The summarized taxonomy barriers and enablers are displayed in Fig 75Mathematical description of the proposed model is presented below
Perceived easeof use EOU Xi
i i( )= +=
aring1 5
b e
(75)
Perceived usefulness USEF iX EOUi
i n( ) = + +=
+aring1 5
1
b b e
(76)
Attitude toward using ATT EOU USEF( ) = + +b b e1 2
(77)
L Hogaboam and TU Daim
155
Atude toward the
behavior
Subjec13ve norm
Inten13on Behavior
Perceived
behavioral
control
Fig 73 Theory of planned behavior (Ajzen amp Madden 1986)
Perceived
usefulness
Perceived
ease of use
Atude toward
using EHR
Intention to
use EHR
system
EHR system use
Technical
factors
Financial
factors
Subjective
Norm
Interpersonal
Influence
Social
(organizatio
nal) factors
Personal
factors
Fig 74 Proposed framework for Study 1
7 Decision Models Regarding Electronic Health Records
156
Tabl
e 7
1 E
xten
ded
taxo
nom
y of
ext
erna
l fa
ctor
s
Fin
anci
al
bull S
tart
-up
cost
s(B
oons
tra
amp B
roek
huis
201
0 C
ress
wel
l amp
She
ikh
201
2 F
onky
ch amp
Tay
lor
200
5 M
cGin
n et
al
201
1
Men
ache
mi
amp B
rook
s 2
006
Pal
acio
et
al
2009
S
hoen
amp O
sbor
n 2
006
Sim
on e
t al
20
07
Val
des
et a
l 2
004
Zar
ouki
an
2006
)
bull O
ngoi
ng c
osts
Ash
amp B
ates
200
5 B
oons
tra
amp B
roek
huis
20
10
DeP
hill
ips
200
7 M
arti
ch amp
Cer
vena
k 2
007
Pol
ice
et a
l 2
011
W
itte
r 2
009)
bull F
inan
cial
unc
erta
inti
es (
lack
of
tang
ible
ben
efits
la
ck o
f fi
nanc
ial
retu
rn r
eim
burs
emen
t)
(Blu
men
thal
200
9 C
haud
hry
et a
l 2
006
Gol
dzw
eig
et a
l 2
009
Men
ache
mi
et a
l 2
008)
bull L
ack
of fi
nanc
ial
reso
urce
s (i
n so
me
sour
ces
refe
rred
to
as l
ack
of c
apit
al l
ack
of f
undi
ng e
tc)
(Ash
amp B
ates
200
5 B
oons
tra
amp B
roek
huis
201
0 B
owen
s et
al
201
0 F
onky
ch amp
Tay
lor
200
5 G
orol
l et
al
200
8
Lor
enzi
et
al
2009
P
alac
io e
t al
20
09
Rob
ert
Woo
d Jo
hnso
n F
ound
atio
n 2
010
Shi
elds
et
al
2007
S
imon
et
al
2008
S
hen
amp G
inn
201
2 S
imon
et
al
2007
)
Tec
hnic
al f
acto
rs
bull In
form
atio
n qu
alit
y (a
ccur
acy
co
nten
t f
orm
at t
imel
ines
s)(B
oden
heim
er amp
Gru
mba
ch 2
003
Che
n amp
Hsi
ao
2012
C
ress
wel
l amp
She
ikh
201
2 K
im amp
Cha
ng 2
006
Lia
ng e
t al
20
11
Mor
es 2
012
Wu
et a
l
2007
)
bull In
tens
ity
of I
T u
tili
zati
on
inte
nsit
y of
IT
uti
liza
tion
(da
ta
secu
rity
doc
umen
tati
on te
chni
cal
supp
ort
com
plex
ity
cu
stom
izat
ion
rel
iabi
lity
in
terc
onne
ctiv
ity
int
erop
erab
ilit
y
hard
war
e is
sues
)
(Ang
st e
t al
20
10
Bat
es e
t al
20
03
Blu
men
thal
200
9 B
oons
tra
amp B
roek
huis
20
10
Bow
ens
et a
l
2010
C
hen
et a
l 2
010
Che
n amp
Hsi
ao 2
012
Duumln
nebe
il e
t al
20
12
Gla
ser
et a
l 2
008
Gor
oll
et a
l 2
008
Gre
enha
lgh
et a
l
2009
H
andy
et
al
2001
Ji
an e
t al
20
12
Lor
ence
amp C
hurc
hill
200
5 L
udw
ick
amp D
ouce
tte
200
9 M
enac
hem
i amp
B
rook
s 2
006
Mil
ler
amp S
im 2
004
Ort
ega
Ege
a amp
Rom
aacuten G
onzaacute
lez
201
1
Pal
acio
et
al
2009
P
olic
e et
al
20
11
Rah
impo
ur e
t al
20
08
Rin
d amp
Saf
ran
199
3 R
osem
ann
et a
l 2
010
Rob
ert
Woo
d Jo
hnso
n F
ound
atio
n 2
010
Sim
on
et a
l 2
007
Tsi
knak
is e
t al
20
02
Tyl
er 2
001
Val
des
et a
l 2
004
Ved
vik
et a
l 2
009
Yoo
n-F
lann
ery
et a
l 2
008
Z
hang
amp L
iu 2
010)
Soc
ial
orga
niza
tion
al
bull T
op m
anag
emen
t su
ppor
t(A
ndreacute
et
al
2008
C
hen
amp H
siao
20
12
Kim
amp C
hang
200
6 L
egri
s et
al
200
3 M
orto
n amp
Wie
denb
eck
200
9 Y
usof
et
al
200
8)
bull P
roje
ctt
eam
com
pete
ncy
(Car
ayon
et
al
2011
C
hen
amp H
siao
20
12
Cho
w e
t al
20
12a
201
2b Y
arbr
ough
amp S
mit
h 2
007
Zar
ouki
an
2006
)
bull P
roce
ss o
rien
tati
on(C
hias
son
et a
l 2
007
Duumln
nebe
il e
t al
20
12)
bull S
tand
ardi
zati
on(B
oons
tra
amp B
roek
huis
201
0 C
ress
wel
l amp
She
ikh
201
2 G
lase
r et
al
200
8 G
reen
halg
h et
al
20
09
Hel
ms
amp
Wil
liam
s 2
011
Hol
den
amp K
arsh
201
0 K
azle
y amp
Ozc
an
2008
K
umar
amp A
ldri
ch 2
010
Lan
ham
et
al
2012
L
apin
sky
et a
l 2
008
Leu
et
al
2008
L
oren
zi e
t al
20
09
Lud
wic
k amp
Dou
cett
e 2
009
Mat
ysie
wic
z amp
Sm
ycze
k
2009
R
ande
ree
200
7 T
sikn
akis
et
al
2002
T
yler
200
1 W
agne
r amp
Wei
bel
200
5 Z
arou
kian
200
6)
L Hogaboam and TU Daim
157
bull S
taff
rea
lloc
atio
nem
ploy
men
t(G
reen
halg
h et
al
200
9 P
apat
heod
orou
199
0 J
ancz
ewsk
i amp
Shi
20
02)
bull S
ecur
ity
confi
dent
iali
typ
riva
cy
conc
erns
(Alp
er amp
Ols
on 2
010
Ang
st e
t al
20
10 A
sh amp
Bat
es
2005
B
oons
tra
amp B
roek
huis
20
10
Bow
ens
et a
l 2
010
D
uumlnne
beil
et
al
2012
M
orto
n amp
Wie
denb
eck
200
9 P
ilio
uras
et
al
2011
R
ind
amp S
afra
n 1
993
R
osem
ann
et a
l
2010
T
yler
200
1)
bull In
cent
ives
(Ash
amp B
ates
200
5 B
ates
et
al
2003
B
ecke
tt e
t al
20
11
Boo
nstr
a amp
Bro
ekhu
is
2010
C
ress
wel
l amp
She
ikh
201
2
For
d et
al
200
6 G
oldz
wei
g et
al
200
9 G
reen
halg
h et
al
200
9 K
umar
amp A
ldri
ch 2
010
Ros
eman
n et
al
20
10
Sch
oen
et a
l 2
006)
bull P
olic
y dr
awba
cks
and
supp
orts
(And
reacute e
t al
20
08
C
hen
amp H
siao
20
12
Chu
mbl
er e
t al
20
11
Gor
oll
et a
l 2
008
Mil
ler
amp S
im
2004
S
choe
n et
al
200
6 S
imon
et
al
2008
V
ishw
anat
h et
al
200
9 W
itte
r 2
009)
bull T
rans
ienc
e of
ven
dors
(Bat
es e
t al
20
03
For
d et
al
200
6 R
ande
ree
200
7)
bull W
orkfl
ow r
edes
ign
(Boo
nstr
a amp
Bro
ekhu
is 2
010
Bow
ens
et a
l
2010
C
haud
hry
et a
l 2
006
Dix
on e
t al
20
10
Fur
ukaw
a 2
011
G
orol
l et
al
200
8 L
oren
zi e
t al
20
09
Men
ache
mi
amp B
rook
s 2
006
Mil
ler
amp S
im 2
004
Zan
dieh
et
al
2008
Z
arou
kian
20
06)
Per
sona
l
bull A
ge s
peci
alty
pos
itio
n
fam
ilia
rity
(Ang
st e
t al
20
10
Ber
gman
-Eva
ns e
t al
20
08
Che
n amp
Hsi
ao
2012
E
gea
amp G
onza
lez
201
1 H
andy
et
al
2001
Je
ng amp
Tze
ng 2
012
Kim
amp H
an 2
008
Mil
ler
amp S
im 2
004
Mor
ton
amp W
iede
nbec
k 2
010
Pai
amp H
uang
20
11
Pol
ice
et a
l 2
011
Rah
impo
ur e
t al
20
08
Ros
eman
n et
al
201
0 V
ishw
anat
h et
al
200
9 W
u et
al
20
07)
bull M
otiv
atio
n(B
ecke
tt e
t al
20
11
Cre
ssw
ell
amp S
heik
h 2
012
Dix
on 1
999
Fra
mba
ch amp
Sch
ille
wae
rt 2
002
Gre
enha
lgh
et a
l
2009
P
ilio
uras
et
al
2011
W
u et
al
20
07 Y
arbr
ough
amp S
mit
h 2
007
Yu
et a
l 2
009)
bull P
rodu
ctiv
ity
(Bow
ens
et a
l 2
010
D
eLia
et
al
2004
M
orto
n amp
Wie
denb
eck
200
9 Y
oon-
Fla
nner
y et
al
200
8)
bull P
erso
nal
inno
vati
vene
ss(F
ram
bach
amp S
chil
lew
aert
200
2 H
ung
et a
l 2
012
Jen
g amp
Tze
ng
2012
M
oore
s 2
012
Vis
hwan
ath
et a
l 2
009
Y
i et
al
200
6)
bull S
elf-
effi
cacy
(Cha
u amp
Hu
200
2 C
hen
amp H
siao
20
12
Cho
w e
t al
20
12a
201
2b
Cre
ssw
ell
amp S
heik
h 2
012
Dix
on 1
999
Kuk
afka
et
al
200
3 L
egri
s et
al
200
3 M
cFar
land
amp H
amil
ton
200
6 R
ahim
pour
et
al
2008
W
u et
al
20
07
Wu
et a
l
2009
Yu
et a
l 2
009)
bull A
nxie
ty(A
ggel
idis
amp C
hatz
oglo
u 2
009
Che
ng 2
012
Kuk
afka
et
al
2003
L
udw
ick
amp D
ouch
ette
2
009
Sto
rey
amp B
ucha
nan
20
08
Wu
et a
l 2
007
Yar
brou
gh amp
Sm
ith
200
7)
Inte
rper
sona
l(C
hang
201
2 C
hen
amp H
siao
20
12
Chi
asso
n et
al
200
7 D
uumlnne
beil
et
al
2012
F
ram
bach
amp S
chil
lew
aert
20
02
L
iu a
nd M
a 2
005
Wu
et a
l 2
007
Yan
g 2
004
Yar
brou
gh amp
Sm
ith
200
7 Y
u an
d G
agno
n 2
009
Yus
of e
t al
20
08)
bull D
octo
r-do
ctor
bull D
octo
r-nu
rse
bull D
octo
r-pa
tien
t
7 Decision Models Regarding Electronic Health Records
158
Intention to useEHRsystem INT ATT SN( ) = + +b b e1 2
(78)
Actual useof thesystem USE INT( ) = +b e1
(79)
where
Xi is an enablerbarrier factor I i = 1hellip5SN is subjective normβi is a standardized partial regression coefficientε is a random regression term
Based on the above-presented framework the following hypothesis will be tested
HA n External barriers and enablers impact PEoU and PU in small ambulatory clinics (n is the number of barriers and enablers that will be finalized through expert validation)
HB1-B2 Interpersonal implementation factors influence subjective norm and atti-tude toward EHR use in clinician practices
HC1-C2 PU and PEoU have significant impact on the attitude toward EHR use
Impactfactors
Financial Technical Socialorganizational Personal Interpersonal
doctor-doctor
doctor-nurse
doctor-patient
start-upcosts
ongoingcosts
financialuncertainties
lack offinancial
resources
information
quality
intensit of
IT utilization
data securty
documentation
technical support
complexity
customization
reliability
interconnectivity
interoperability
hardware issues
accuracy
content
format
timeliness
top managementsupport
projectteamcompetency
process orientation
standardization
staff reallocation
employment
securityconfidentialityprivancy concerns
incentives
policy drawbacks andsupports
transience of vendors
workflow redesign
age
specialty
position
familiarity
motivation
productivity
personalinnovative-
ness
self-efficacy
anxiety
Fig 75 Taxonomy of barriers and enablers
L Hogaboam and TU Daim
159
HD1-D2 Intention to use EHR system is impacted by subjective norms and attitude toward using EHR and PU
HE PEoU influences PU of EHR in small ambulatory settingsHF Positive intention to use EHR system translates into EHR use
72 The Selection of EHR with Focus on Different Alternatives
When we are trying to select a product or technology based on a number of alterna-tives we engage in a decision-making process While we make our decisions every day some of them are more complex than the routine kind and require established managerial methodologies created for this purpose Hierarchical decision model (HDM) is used to decompose the problem into hierarchical levels and using pair-wise comparison scales and judgment quantification technique the researcher arrives at the calculated alternative However the process of decision analysis is even more of a value than the answer it brings since it forces systematic assessment of the alternatives (Henriksen 1997) Decision analysis provides information so that managers of technology in this case healthcare information technology spe-cifically EHR can make more informed decisions Some interesting examples of HDM in healthcare were described by Bohanec and others (Bohanec 2000) and were clinical in nature (assessment of breast cancer risk assessment of basic living activities in community nursing risk assessments in diabetic foot care etc) using DEX an expert system shell for multi-attribute decision support
Community-wide implementation of EHR was studied by Goroll et al where Massachusetts eHealth Collaborative (MAeHC) was formed in order to improve patient safety and quality of care through HIT use promotion (Goroll et al 2008) The working group outlined a set of system features that were involved in the selec-tion of vendors Those were (Goroll et al 2008)
bull User friendlinessbull Functionalitybull Clinical decision support capabilitybull Interoperabilitybull Securitybull Reliabilitybull Affordability
The authors also stress that despite the national push of EHR implementation positive encouragements in terms of vendor certification and system standards the current state of standards cannot ensure sufficient specific fit for a routine use by practices interoperability and ease of use therefore considerable technical as well as organizational efforts need to be engaged in the system (Goroll et al 2008)
7 Decision Models Regarding Electronic Health Records
160
Below are some figures depicting the bodies of knowledge surrounding organi-zational issues in HIT innovation (Fig 76) and theoretical approaches that concep-tualize interaction between technology humans and organizations (Cresswell amp Sheikh 2012) (Table 712)
Table 72 is the table of theoretical approaches that conceptualize interaction between technology humans and organizations (Cresswell amp Sheikh 2012)
Table 73 shows some information derived from Table 31 of 2009 Oregon Ambulatory EHR survey (Witter 2009)
The model is shown in Fig 77
721 Criteria
Seven criteria were chosen based on the extensive literature review Perceived use-fulness and perceived ease of use are based on the elements of the TAM Since the above-described research indicates that the acceptance of the technology is based on perceptions of users (physicians of small clinics with decision-making power in this
Organizaonal issues in HIT innovaon
Human factors
ergonomics
Organizational occupational
social psychology
Management amp organizational
change management
Information systems
Fig 76 Bodies of knowledge surrounding organizational issues in HIT innovation
L Hogaboam and TU Daim
161
Table 72 Theoretical approaches of interaction between technology humans and organizations
Name of the theory Explanations and definitions
Diffusion of innovations
Focuses on how innovations spread in and across organization over time
Normalization process Describes the incorporation of complex interventions in healthcare into the day-to-day work of healthcare staff
Sense making Assumes that organizations are not existing entities as such but produced by sense-making activities and vice versa they discover meaning of the status quo often by transforming situations into words and displaying a resulting action as a consequence
Social shaping theory Views technology as being shaped by social processes and highlights the importance of wider macro-environmental factors in influencing technology
Sociotechnical changing
Conceptualizes change as a nonlinear unpredictable and context- dependent process assuming that social and technical dimensions shape each other in a complex and evolving environment over time
Technology acceptance model
Assumes that individualrsquos adoption and usage of the system are shaped by the attitude toward use perceived ease of use and perceived usefulness
The notion of ldquofitrdquo Accentuates that social technological and work process factors should not be considered in isolation but in the appropriate alignment with each other
Table 73 Organizations and clinicians not planning to implement EHR in Oregon in 2009
Percent of organizations and clinicians with no plan to implement an EHREMR All entities
Clinicians all entities
Total organizations and clinicians 626 2313
Barriers
Security and privacy issues 181 112
Confusing number of EMR choices 03 01
Lack of expertise to lead or organize the project 195 166
No currently available EMR product satisfies our [needs] 182 208
Staff would require retraining 260 310
Expense of purchase 802 841
Expense of Implementation 586 684
Inadequate return on investment 361 298
Concern the product will fail 179 156
Staff is satisfied with paper-based records 348 259
Practice is too small 478 257
Plan to retire soon 173 77
Other 147 231
case) those criteria were included in the model It is assumed that EHR systems comply with ARRA mandates and have legal compliance
Those seven criteria and subcriteria will also be reviewed and justified by the experts in the field Experts will be chosen from academia in the field of healthcare and healthcare management and physicians
7 Decision Models Regarding Electronic Health Records
162
Fig 77 Hierarchical model of EHR software selection
7211 Perceived Usefulness
This criteria has its roots in TAM (Davis 1989) and identifies the userrsquos perception of the degree to which using a particular system will improve his or her perfor-mance The psychological origins of the concept are grounded in social presence theory social influence theory and Triandis modifications to the TRA (Karahanna amp Straub 1999) Perceived usefulness has been shown to have a great impact on technology acceptance in healthcare (Chen amp Hsiao 2012 Cheng 2012 Cresswell amp Sheikh 2012 Despont-Gros et al 2005 Kim amp Chang 2006 King amp He 2006 McGinn et al 2011 Melas et al 2011 Morton amp Wiedenbeck 2009 Yusof et al 2008) The concepts of TAM and relative research have been instrumental in explaining how beliefs about systems lead users to have positive attitudes toward systems intentions to use these systems and system use (Karahanna amp Straub 1999)
With the concepts of perceived usefulness the subcriteria that were selected from the literature review included the following
bull Data securityThe concept of data security has been brought up by many researchers as well as the government (Alper amp Olson 2010 Bowens Frye amp Jones 2010 Chen et al 2010 Duumlnnebeil et al 2012 Liu amp Ma 2005 Lorence amp Churchill 2005 Rind amp Safran 1993 Tsiknakis Katehakis amp Orphanoudakis 2002 Vedvik Tjora amp Faxvaag 2009 Yusof et al 2008 Zhang amp Liu 2010) The concept of
L Hogaboam and TU Daim
163
data security encryption and secure storage has been described in the literature review sections above Differences of in-cloud vs remote storage have been discussed as having various security features
bull InteroperabilityThe system should be able to function well with other applications in the net-work local and shared Alper and Olson (2012) note that interoperability is important to improve and coordinate care delivery While in the USA most patients receive care from several providers a lack of interoperability in the network would mean that physicians do not have access to a complete record for a patient and a ldquomaster recordrdquo might not exist or might not be complete at any point in time (Alper amp Olson 2012) Different systems will provide various levels of interoperability and the users may require more or less advanced sys-tems for their clinics A number of researchers stressed the importance of interop-erability of the EHR system as expressed by administrators physicians and other EHR users and the need to invest in improvements in it (Alper amp Olson 2012 Ash amp Bates 2005 Blumenthal 2009 Blumenthal 2010 Box et al 2010 Bufalino et al 2011 Cresswell amp Sheikh 2012 Degoulet Jean amp Safran 1995 DePhillips 2007 Dixon Zafar amp Overhage 2010 Duumlnnebeil et al 2012 Fonkych amp Taylor 2005 Furukawa 2011 Glaser et al 2012 Goldzweig et al 2009 Goroll et al 2008 Jian et al 2012 Jung 2006 Kazley amp Ozcan 2008 Lapinsky et al 2008 Maumlenpaumlauml et al 2009 McGinn et al 2011 Palacio Harrison amp Garets 2009 Tsiknakis et al 2002 Yao amp Kumar 2013 Yoon- Flannery et al 2008 Zaroukian 2006 Zhang amp Liu 2010)
bull CustomizationCustomization is an extremely important concept since various clinics with their unique specializations services provided and clientspatients of various needs have different needs in software customization as far as costs complexities and training required are concerned While some prefer a system that could be tai-lored in a unique way others may prefer a low-cost off-the-shelf product without elaborate customization capabilities (Alper amp Olson 2012) The issue of cus-tomization in EHR selection has been stressed by a number of researchers (Alper amp Olson 2012 Ash et al 2001 Cresswell amp Sheikh 2012 Degoulet et al 1995 Kim amp Chang 2006 Ludwick amp Doucette 2009 Menachemi amp Brooks 2006 Randeree 2007 Roth et al 2009 Witter 2009 Zandieh et al 2008)
bull ReliabilityReliability is a complex issue as well since a certain level of reliability of the system and the vendor must be present for the successful use of the EHR Thus Alper and Olson (2010) stated that the health information network that is able to be aggregated with a reasonable degree of accuracy and reliability would improve the ability to track known epidemics and identify new epidemics or other threats to public health such as bioterrorism or environmental exposures at an early stage Cresswell and Sheikh (2012) look at the lack of reliability of the system from the view of system stabilitymdashsoftware crashes etc Other researchers
7 Decision Models Regarding Electronic Health Records
164
include the concept of reliability when they study healthcare IT and EHR in par-ticular (Alper amp Olson 2010 Box et al 2010 Cresswell amp Sheikh 2012 Degoulet et al 1995 Despont-Gros et al 2005 Goroll et al 2008 Liu amp Ma 2005 Maumlenpaumlauml et al 2009 Moores 2012 Yusof et al 2008 Zaroukian 2006)
bull Product life cycleGenerally product life cycle of software (EHR as well) is short (Goroll et al 2008) therefore the physicians that are planning to acquire those systems should look into the fact of how fast they would need to upgrade and change the system when it will become obsolete and how long could it run and be supported after being installed It is closely tied with concepts of upgradability and system obso-lescence This concept is mentioned by a number of authors (Carayon et al 2011 David amp Jahnke 2005 DePhillips 2007 Goroll et al 2008 Hatton Schmidt amp Jelen 2012 Randeree 2007 Vedvik et al 2009 Witter 2009 Zaroukian 2006 Zhang amp Liu 2010)
7212 Perceived Ease of Use
Just like perceived usefulness the concept of ease of use has been known from Davisrsquos TAM (Davis 1989) and it is the userrsquos perception of the extent to which using a particular system would be free of effort A large body of research has shown that perceived ease of use significantly impacts technology acceptance and influences userrsquos decision-making process (Ayatollahi et al 2009 Carayon et al 2011 Chen amp Hsiao 2012 Cheng 2012 Chow Chan et al 2012a 2012b Chow Herold et al 2012b Cresswell amp Sheikh 2012 Davis amp Venkatesh 1996 Despont- Gros 2005 Dixon 1999 Duumlnnebeil et al 2012 Garcia-Smith amp Effken 2013 Jian et al 2012 Karahanna amp Straub 1999 Kim amp Chang 2006 King amp He 2006 Legris et al 2003 Liu amp Ma 2005 Melas et al 2011 Vishwanath et al 2009 Yusof et al 2008 and others)
The subcriteria for ldquoperceived ease of userdquo are the following
bull Ease of data extractionaccessThe EHR system could be packed with valuable data but if it is not easy for the user to access it (in a timely manner with not a significant amount of effort) the value of that system to the user diminishes greatly Easy access to information facilitates communication and decision making in healthcare (Kim amp Chang 2006) Certain decision support tools could be enabled in EHR software for improving physicianrsquos ease of access to data (Bodenheimer amp Grumbach 2003) The concept of accessibility and data extraction is studied in the context of health-care management IT acceptance and software or application selection (Ayatollahi et al 2009 Chumbler et al 2011 Duumlnnebeil et al 2012 Furukawa 2011 Garcia-Smith amp Effken 2013 Leu et al 2008 Maumlenpaumlauml et al 2009 Millstein amp Darling 2010 Rind amp Safran 1993 Roth 2009 Zhang amp Liu 2010)
L Hogaboam and TU Daim
165
bull Search abilitySystemrsquos user should be able to search the system in a timely effortless manner with acceptable and meaningful results Search capabilities could be one of the most important subcriteria as having a good-quality search engine with quick searching capabilities could greatly benefit a small practice however some phy-sicians may not feel like they need an elaborate searching system and may opt out for software with a modest acceptable searching capabilities Researchers have noted the feature of good data mining or data search (Alper amp Olson 2010 Ayatollahi et al 2009 Palacio et al 2009 Randeree 2007)
bull InterfaceConvenient interface that is easy to use and adjust to is possibly one of the most and first noticeable user-friendly features of the EHR system However the user might not require a fancy interface and may need an interface that fits the need of the clinic A user interface that is poorly designed with fragmented screens and multiple sign-ins can increase computer time and also lead to dissatisfaction (Furukawa 2011) Interface is a discussed topic in research and is often men-tioned in phrases as ldquointerface designrdquo or ldquointerface design qualityrdquo (Alper amp Olson 2010 Ayatollahi et al 2009 Becker et al 2011 Cresswell amp Sheikh 2012 Davis 1989 Degoulet et al 1995 Despont-Gros 2005 Ludwick amp Doucette 2009 Melas et al 2011 Moores 2012 Valdes et al 2004 Yusof et al 2008)
bull ArchivingArchiving and storing of the data is also an important concept since the quality of archiving can impact quality of retrieval of information Also the ease of archiving or the simplicity of it should benefit the physician the patient and the clinic overall The importance of archiving is captured in various research jour-nals and reports (Alper amp Olson 2010 Chen et al 2010 Goldberg 2012 Ludwick amp Doucette 2009 Maumlenpaumlauml et al 2009 Sanchez et al 2013 Vedvik et al 2009 Wu et al 2009 Zhang amp Liu 2010)
7213 Financial Criterion
A financial criterion is well mentioned in the literature as affordability of EHR by small clinics is a large issue Some researchers indicated that facilitating conditions like financial rewards have been main factors to positively affect behavioral inten-tion (Aggelidis amp Chatzoglou 2009) Shen and Ginn (2012) devoted their research to analyzing financial position and adoption of electronic health records through a retrospective longitudinal study Their conclusions stated that financial position indeed relates to EHR adoption in midterm and long-term planning (Shen amp Ginn 2012) Goldzweig et al (2009) have noted that the costs still remain the number one barrier cited by surveys assessing adoption and stressed the need for a better align-ment between ldquowho paysrdquo and ldquowho benefitsrdquo from health IT Miller and Sim (2004)
7 Decision Models Regarding Electronic Health Records
166
indicated that EMR use could be increased through implementation of financial rewards for quality improvement and for public reporting of quality performance measures
Through my independent studies besides the abovementioned articles I have found a large number of researchers studying importance of financial incentives identification of financial barriers and outlining financial attributes that are funda-mental for healthcare IT implementation (Andreacute et al 2008 Ash amp Bates 2005 Blumenthal 2009 Boonstra amp Broekhuis 2010 Cresswell amp Sheikh 2012 Dixon et al 2010 Fonkych amp Taylor 2005 Furukawa 2011 Goldberg 2012 Im et al 2008 Jung 2006 Leu et al 2008 Linder et al 2007 Martich amp Cervenak 2007 McGinn et al 2011 Ortega Egea amp Roman Gonzalez 2011 Randeree 2007 Simon et al 2007 Zandieh et al 2008)
bull Start-up costs (affordability)Major investment in EHR begins with costs required in order to acquire EHR system Small clinics could do it from their own savings investorsrsquo capital financial incentive or loans Researchers have stressed importance of this sub-criterion (Boonstra amp Broekhuis 2010 Cresswell amp Sheikh 2012 Fonkych amp Taylor 2005 McGinn et al 2011 Menachemi amp Brooks 2006 Palacio et al 2009 Shoen amp Osborn 2006 Simon et al 2007 Valdes 2004 Zaroukian 2006)
bull Ongoing and maintenance costsIn addition to initial costs required to obtain a system there are various costs associated with maintaining the system possibly updating it personnel costs associated with system upkeep etc Other researchers also note the importance of these costs (Ash amp Bates 2005 Boonstra amp Broekhuis 2010 DePhillips 2007 Martich amp Cervenak 2007 Police et al 2011 Witter 2009) and it would be interesting to assess physicianrsquos concerns about those costs as well as report about physicianrsquos awareness of those costs during the decision-making process
bull Ease of upgradeJust like with any software with an ongoing innovations and process changes in the industry and shorter life cycles of the products the upgrade may bring techni-cal and financial difficulties Those financial difficulties could be associated with a need to hire additional personnel to compensate for delays in patientrsquos care during the process of upgrade need to updatechangepurchase new computers install new additional programs etc Those costs could be 5ndash10 of providerrsquos current EHR costs (Alper amp Olson 2010) Randeree (2007) also discusses physi-ciansrsquo need to weigh in the costs of creating and supporting their IT structure as well as applications compared to using the external vendors for those services Those additional costs (upgrade coordination monitoring negotiating and governance) may delay the adoption since for small practices a typical EMR soft-ware costs approximately $10000 per physician not including the maintenance costs and costs for hardware and other software (Randeree 2007) Those issues are noted in other papers (Carayon et al 2011 David amp Jahnke 2005 DePhillips 2007 Dixon 1999 Goroll et al 2008 Janczewski amp Shi 2002 Kumar amp
L Hogaboam and TU Daim
167
Aldrich 2010 Martich amp Cervenak 2007 Menachemi amp Brooks 2005 2006 Piliouras et al 2011 Vedvik et al 2009 Witter 2009 Zaroukian 2006)
7214 Technical Criterion
With constant technological advances in the area of information technology and particularly EHR technical aspects are very important to consider but most impor-tant is to assess how well they will fit in within the organizational and social aspect whether those technical capabilities would be a good fit and whether they get a good use under the current circumstances While technical criteria is difficult to keep current because of ever-changing capabilities of the system and the types and brands of software coming out on the market we would ask the experts to closely examine the subcriteria and assess the additional technical aspects based on the selection of software Technical criterion is mentioned extensively in the literature (Angst et al 2010 Bates et al 2003 Blumenthal 2009 Bodenheimer amp Grumbach 2003 Boonstra amp Broekhuis 2010 Bowens et al 2010 Chen et al 2010 Chen amp Hsiao 2012 Cresswell amp Sheikh 2012 Duumlnnebeil et al 2012 Glaser et al 2008 Goroll et al 2008 Greenhalgh et al 2009 Handy et al 2001 Jian et al 2012 Kim amp Chang 2006 Liang et al 2011 Lorence amp Churchill 2005 Ludwick amp Doucette 2009 Menachemi amp Brooks 2006 Miller amp Sim 2004 Mores 2012 Ortega Egea amp Romaacuten Gonzaacutelez 2011 Palacio et al 2009 Police et al 2011 Rahimpour et al 2008 Rind amp Safran 1993 Robert Wood Johnson Foundation 2010 Rosemann et al 2010 Simon et al 2007 Tsiknakis et al 2002 Tyler 2001 Valdes et al 2004 Vedvik et al 2009 Wu et al 2007 Yoon-Flannery et al 2008 Zhang amp Liu 2010)
bull Supporting databasesThis is a subcriteria that has its links to interconnectivity of an EHR system since it may be important for many doctors to have access to certain clinical databases or other medical databases helpful in providing better healthcare since doctors may be able to provide more informed diagnoses may have access to new infor-mation about prescription drugs and their effects and newest clinical trials etc For example McCabe (2006) did some research into available databases for mental health in an effort to promote and study evidence-based practice which is a strategy to incorporate research results into the process of care They found that some sources like Cochrane Database of Systematic Reviews provide high- quality reviews of randomized controlled trials (RCTs) and other sources like the Database of Abstracts of Reviews and Effectiveness and the Agency for Health Care Research and Quality offer structured abstracts and clinical guide-lines for medical treatments (McGabe 2006)
There is some evidence that medication dispensation data obtained from claims databases improves the medication reconciliation and refill process in clinics (Leu et al 2008) Other supporting literature for database support was also found (Chen et al 2010 Degoulet et al 1995 Henrickren 1997 Hung Ku amp Chien 2012 Janczewski amp Shi 2002 Jung 2006 Lorenzi et al 2009
7 Decision Models Regarding Electronic Health Records
168
Pareacute amp Sicotte 2001 Police et al 2011 Randeree 2007 Vishwanath et al 2009 Zaroukian 2006 Zhang amp Liu 2010)
bull CompatibilityEnsuring compatibility of the EHR system with current work practices one of the key beliefs that influence adoptionmdashthe extent to which the system fits or is com-patible with the way the user likes it to work is a necessary component of IT acceptance (Moores 2012) The system must fit the needs of the user however some users may require higher degree of compatibility due to specialization of the practice certain procedures and particular processes in place while others may not perceive it as such a deciding factor in EHR selection Other researchers stressed the importance of the compatibility issue (Aggelidis amp Chatzoglou 2009 Alhateeb et al 2009 Chow et al 2012a 2012b Goroll et al 2008 Helfrich et al 2007 Holden amp Karsh 2010 Hung et al 2012 Kukafka et al 2003 Pynoo et al 2011 Randeree 2007 Shibl et al 2013 Staples et al 2002 Wu et al 2007 Yi et al 2006 Zaroukian 2006) Compatibility also is mentioned in diffu-sion theory as one of the five characteristics of innovation that affect their diffu-sion as innovationrsquos consistency with usersrsquo social practices and norms (Dillon amp Morris 1996) The other four are relative advantage (the extent to which technol-ogy offers improvements over tools that are currently available) complexity (innovationrsquos ease of use or learning) trialability (the opportunity of trying an innovation before committing to use it) and observability (the extent to which the outputs and gains of the new technology are clearly seen) (Dillon amp Morris 1996)
bull Clinical data exchangeClinical data exchange system gives the capability to move clinical information electronically across organization while maintaining the meaning of the informa-tion being exchanged (Li et al 1998) Communication standardization fund-ing and interoperability are some of the main barriers for the global clinical data exchange networks While selecting EHR the importance of clinical data exchange system to the users of the EHR system would be very interesting to assess Other researchers that studied the importance of clinical data exchange or included it as one of the important aspects of EHR use are the following Bowens et al (2006) Dixon et al (2010) Goroll et al (2008) Jian et al (2012) Maumlenpaumlauml et al (2009) Miller and Sim (2004) and Moores (2012)
7215 Organizational Criterion
In addition to the technical and financial aspects of EHR selections it is also impor-tant to consider organizational aspect that plays a crucial role in a decision-making process Box et al (2010) state that throughout health information technology imple-mentation success requires a careful balance of technical clinical and organiza-tional factors Cresswell and Sheikh (2012) dedicate an empirical and interpretative review study on organizational issues in HIT adoption and implementation
L Hogaboam and TU Daim
169
Organizational issues were described by the number of researchers Alper and Olson (2010) Ash and Bates (2005) Boonstra and Broekhuis (2010) Brand et al (2005) Burton-Jones and Hubona (2006) Chen et al (2010) Chumbler et al (2011) Davis (1989) Goldberg et al (2012) Johnson et al (2012) Kim and Chang (2006) Kukafka et al (2003) Lanham et al (2012) McGinn et al (2011) Moores (2012) Morton and Wiedenbeck (2009) Pynoo et al (2011) Weiner et al (2011) Yarbrough and Smith (2007) Yi et al (2006) and Zaroukian (2006)
bull StandardizationConforming to specific standards is an important issue and as various EHR sys-tems exist as well as various standards some systems might be more standardized than others From another perspective some standardization may be required in physicianrsquos practices for implementation of EHR McGinn et al (2012) talk about a lack of uniform standards at all levels (local regional national) which may contribute to physicianrsquos and managerrsquos disorientation when choosing an EHR system Hatton et al (2012) explain that even simple attempts at standard-ization (like ordering common blood chemistry tests) could be challenging for physicians which authors associate with physiciansrsquo challenges with EHR implementation Various perspectives of standardization issue have been men-tioned in the literature (Cresswell amp Sheikh 2012 Duumlnnebeil et al 2012 Kumar amp Aldrich 2010 Lanham et al 2012 Li et al 1998 Ludwick amp Doucette 2009)
bull TrainingWith any new system there will be some time for adjustment from an organiza-tional point of view and some training required Some systems may require more or less training and physicians need to be aware of those variables In addition to the possible financial impact the process of training will require it may also involve hiring more personnel or using vendorsrsquo training human resources The intensity timing and availability of training and support post-implementation affect user experience (Ludwick amp Doucette 2009) The issue of training is an important one to consider and has been mentioned by various researchers (Ayatollahi et al 2009 Chaudhry et al 2006 Kumar amp Aldrich 2010 Lee amp Xia 2011 Ludwick amp Doucette 2009 McGinn et al 2011 Moores 2012 Morton amp Wiedenbeck 2009 Noblin et al 2013 Pilouras et al 2011 Police et al 2011 Yeager et al 2010 Yi et al 2006 and others)
bull Tech SupportThe availability of tech support is important in EHR selection with some that may have straightforward personalized system or online-only system or the vendor might not provide tech support Depending on the IT infrastructure and the in-house capabilities physicians need to carefully examine this aspect to decide how important tech support is for them and how much tech support they will require Tech support or lack of thereof is an issue described by
7 Decision Models Regarding Electronic Health Records
170
researchers with bright examples in qualitative studies (Boonstra amp Broekhuis 2010 Goroll et al 2008 Holden amp Karsh 2010 Lustria et al 2011 Miller amp Sim 2004 Pynoo et al 2011 Valdes et al 2004 Wu et al 2007 Yu et al 2009)
7216 Personal Factors
There is some empirical research that expresses concern about EHR systems infring-ing on physiciansrsquo personal and professional privacy and acting as management control mechanisms (McGinn et al 2011) Boonstra and Broekhuis (2010) also discuss physicianrsquos personal issues about the questionable quality improvement associated with EHR and worry about a loss of professional autonomy Pilouras et al (2011) note that some practitioners use personal references and place high reliance on the experiences of other practices to help them make decision on which package to select
bull Privacy issuesPrivacy concerns have been some of the well-noted issues for physicians while choosing an EHR system
Issues of privacy are mentioned in numerous research articles (Angst et al 2010 Ash amp Bates 2005 Bates et al 2003 Blumenthal 2010 Bufalino et al 2011 Dephillips 2007 Glaser et al 2008 Goroll et al 2008 Handy et al 2001 Kazley amp Ozcan 2007 Lorenzi et al 2009 Lustria et al 2011 Morton amp Wiedenbeck 2010 Palacio et al 2009 Randeree 2007 Simon et al 2007 Tyler 2001 Yoon-Flannery et al 2008 Zheng et al 2012)
bull ProductivityPhysiciansrsquo concerns about losses in productivity and time have been discussed throughout my literature reviews and in this part Some users reported decrease in productivity right after the implementation of an EHR system (Cresswell amp Sheikh 2012) There are numerous research papers especially qualitative stud-ies that recorded interviews with physicians and other users of the system describing issues of productivity with selection and implementation of an EHR system (Andreacute et al 2008 Boonstra amp Broekhuis 2010 Bowens et al 2010 Chaudhry et al 2006 Davidson amp Heineke 2007 Ford et al 2006 Hatton et al 2012 Maumlenpaumlauml et al 2009 McGinn et al 2011 Morton amp Wiedenbeck 2009 Piliouras et al 2011 Police et al 2011 Storey amp Buchanan 2008 Yi et al 2006 Yoon-Flannery et al 2008) According to a survey of Medical Group Management Association Report more than four out of five users of paper records (783 ) believed that there would be a ldquosignificantrdquo to ldquovery signifi-cantrdquo loss of provider productivity during implementation and two-thirds (674 ) had concerns about the loss of physician productivity after the transi-tion period with EHR (MGMA 2011)
L Hogaboam and TU Daim
171
7217 Interpersonal Criterion
bull Sharing among doctors (doctor-doctor relationship)bull Interconnectivity between doctor and nurses (doctor-nurse relationship)bull Sharing with patients (doctor-patient relationship)
The importance of various relationships in peoplersquos lives and workplaces can impact decision-making processes Perceived impact of dynamics of the relation-ship whether itrsquos doctor-doctor doctor-nurse and doctor-patient should not be overlooked Interpersonal criterion has some elements of social organizational and personal dynamics (Cresswell amp Sheikh 2012) The importance of sharing and communication among various levels in the organization and outside (doctor- patient) and the ability of EHR software to provide that capability and perhaps improve the communication and important flow of information should be consid-ered during an EHR selection process Interpersonal issues have been discussed in the research literature (Beckett et al 2011 Chen amp Hsiao 2012 Cheng 2012 Chiasson et al 2007 Duumlnnebeil et al 2012 Frambach amp Schillewaert 2002 Liu amp Ma 2005 Wu et al 2007 Yang 2004 Yarbrough amp Smith 2007 Yu et al 2009 Yusof et al 2008) Kumar and Aldrich performed an SWOT analysis of a nationwide EMR system implementation in USA and in the section of ldquothreatsrdquo included statements that greater standardization could remove the ldquohuman touchrdquo between healthcare practitioners and patients and the doctor-patient relationship might turn into a new triad where EMR could be acting as a proxy for all who provide patient with care
The following hypotheses will be examined
HA1-A2 Perceived usefulness and ease of use have a high influence in the process of decision making for EHR selection
HB Interpersonal implementation factors greatly impact the EHR selection process
HC Financial factors significantly impact physicianrsquos decision-making process for EHR selection
HD Organizational factors significantly impact physicianrsquos decision-making pro-cess for EHR selection
HE1-E2 Productivity and privacy play an important role in EHR selection from physicianrsquos point of view
7218 Methodology
Multi-criteria decision tools like Saatyrsquos Analytic Hierarchy Process (AHP) (Saaty 1977) and HDM (Kocaoglu 1983) have some important steps in the application process
1 Structuring the decision problem into levels consisting of objectives and their associated criteria
7 Decision Models Regarding Electronic Health Records
172
2 Eliciting decision makerrsquos preferences through pairwise comparison among all variables at every hierarchical level of the decision model
3 Processing the input from the decision maker and calculating the priorities of the objectives
4 Checking consistency of the decision makerrsquos responses to ensure logical and not random comparison of the criteria
The last level of the hierarchy will be the software choices By the time the research is conducted the software selection might need to be evaluated again but currently according to the literature search performed for this exam the software choices are listed in Table 69
In HDM a variance-based approach is used for the inconsistency calculations and 10 limit is recommended on it in the constant sum method (CSM) While the HDM approach is similar to Saatyrsquos AHP the computational phase uses the CSM instead of the eigenvectors (Kocaoglu 1983) As explained by Dr Kocaoglu in the hierarchical decision process the problem is considered as a network of relation-ships among major levels (impact target and operational) of hierarchy with multi- criteria objectives at the top leading to multiple benefits and at the bottommdashmultiple outputs resulting from multiple actions (Kocaoglu 1983)
The CSM (Kocaoglu 1983) consists of the following
1 n(n minus 1)2 are randomized for the n elements under consideration 2 The decision makers distribute a total of 100 points between elements with
respect to each other (If they are of equal importance both elements get 50 points if one is four times highermore important with respect to another the allocation will be 80ndash20 points etc)
3 The data is written into Matrix A through comparing column elements with row elements
4 Matrix B is obtained by taking the ration of comparisons for each pair from Matrix A
5 Matrix C is constructed through division of each element in a column of Matrix B by the element in the next column
6 Element d is assigned a value of 1 and the calculation of other elements is per-formed by ratios as the mean of each column in Matrix C
73 The Use of EHR with Focus on Impacts
In the study about impacts of EHR system use itrsquos important to consider impact factors found in the literature For example such effect factors were described by DesRoches et al in the New England Journal of Medicine (DesRoches et al 2008) with percentages of positive survey responses upon adoption of EHR Those were
bull Quality of clinical decisionsbull Quality of communication with other providers
L Hogaboam and TU Daim
173
bull Quality of communication with patientsbull Prescription refillsbull Timely access to medical recordsbull Avoiding medication errorsbull Delivery of preventive care that meets guidelinesbull Delivery of chronic illness care that meets guidelines
While the positive effect was shown in many cases the significance of p lt 0001 was reported only for the quality of clinical decisions delivery of preventive care that meets guidelines and delivery of chronic illness care that meets guidelines
Lanham at al who focused on social underpinning of EHR use or the ldquohuman elementrdquo of EHR acceptance implementation and use also noted about research in the area of EHR impacts particularly EHR influence of fundamental outcomes like cost and quality of healthcare delivery as well as reshaping organizational culture and clinical workflow (Lanham et al 2012)
Goroll et al (2008) also talked about the impact on safety and impact on quality Those types of EHR impacts may be hard to assess but are extremely important in growing the healthcare information management field and constantly improving it Chaudhry et al (2006) performed systematic review of the impact of HIT on qual-ity efficiency and cost The researchers outlined the components of an HIT imple-mentation (Chaudhry et al 2006)
bull Technological (for example system applications)bull Organizational process change (workflow redesign)bull Human factors (user friendliness)bull Project management (archiving project milestones)
Chaudhry et al (2006) also discussed what elements are behind the major effects of quality efficiency and cost
1 Effect on quality was predominantly in the role of increasing adherence (with decision support) to guideline- or protocol-based care In addition to the men-tioned variable clinical monitoring based on large-scale screening and aggrega-tion of data could show how health IT can support new ways of care delivery Reduction of medication errors was also reported measure of the effect on quality
2 Effects on efficiency
(a) Utilization of care (could be measured through the monetized estimates through the average cost of the examined service at the researched institu-tion could be analyzed through provided decision support (display of labo-ratory test costs computerized reminders display of previous test results automated calculation of pretest probability for diagnostic tests) at the point of care)
(b) Provider time (physician time could be examined in relation to computer use)
7 Decision Models Regarding Electronic Health Records
174
3 Effects on costs (changes in utilization of services cost data on aspects of system implementation or maintenance)
A summary table indicating key points of the systematic review on impacts of HIT from (Chaudhry et al 2006) is displayed in Table 74 above
While a lot of studies on barriers to adoption and impacts of EHR have been mentioned in this exam one particular study by Yusof et al (2008) examined previ-ous models of IS evaluation particularly the IS success model and the IT-organization fit model as well as introduced another HOT-fit model based on the system of human organization and technology-fit factors Before our EHR impacts model will be introduced letrsquos look at the theoretical history behind it
Updated DeLone and McLean IS success model was developed in 2003 based on the original DeLone and McLean IS success model introduced 20 years ago as a framework and model for measuring the complex-dependent variable in IS research (DeLone amp McLean 2003) The model is shown in Fig 78
As can be seen from the framework (Fig 78) the measures are included in the six system dimensions (Yusof et al 2008 DeLone amp McLean 2003)
bull System quality (the measures of the information processing system itself)bull Information quality (the measures of IS output)bull Service quality (the measures of technical support or service)bull Information use (recipient consumption of the output of IS)bull User satisfaction (recipient response to the use of the output of IS)bull Net benefits (IS impact overall)
While the model illustrates clear grounded well-observed and specific dimen-sions or impacts of IS successeffectiveness and their relationships it does not include organizational factors which have been included in HOT-fit model (Yusof et al 2008) Before depicting HOT-fit model there is another model that requires our attention in order to improve understanding of our research model
Table 74 Summary points of impact studies Chaudhry et al (2006)
Main summary points of impact studies
Health information technology has been shown to improve quality throughbull Increasing adherence to guidelinesbull Enhancing disease surveillancebull Decreasing medication errors
Primary and secondary preventive care holds much evidence on quality improvement
Decreased utilization of care is reported as the major efficiency benefit
Effect on time utilization is mixed
Empirically measured data on the aspects of costs is limited and inconclusive
Four benchmark research institutions supply most of the high-quality literature on multifunctional HIT systems
Effect of multifunctional commercially developed systems is not well documented
Interoperability and consumer HIT impacts have little evidence
Generalizability is a major limitation in the literature
L Hogaboam and TU Daim
175
IT-organizational fit model was presented in 1991 by Scott Morton and includes both internal and external elements of fit Modelrsquos internal fit is attained through combination and dynamic equilibrium of organizational components of business strategy organizational structure management processes and roles and skills while modelrsquos external fit is achieved due to formulation of organizational strategy grounded in environmental trends and market industry and technology changes (Yusof et al 2008) The enablermdashITmdashis shown to affect the management process also impacting organizational performance and strategy IT-organizational fit model (Yusof et al 2008) is shown in Fig 79
In 2008 Yusof et al combined elements of both models to create humanndashorga-nizationndashtechnology fit (HOT-fit) framework and proposed it for applications in healthcare while testing it with subjectivist case study strategy approach employ-ing qualitative methods (Yusof et al 2008) The researchers also presented exam-ples (Table 75) of the evaluation measures of the proposed network The HOT-fit proposed framework is shown in Fig 710
In our research model we are going to use hierarchical decision modeling in order to study impacts of EHR system as perceived by physicians of small ambula-tory clinics The criteria in the levels have been explained through the theoretical background and literature sources The methodology has been explained in detail
NETBENEFITS
USERSATISFACTION
INTENTIONTO USE
INFORMATIONQUALITY
SYSTEM QUALITY
SERVICEQUALITY
USE
Fig 78 Updated DeLone and McLean IS success model (DeLone amp McLean 2003)
Structure
Strategy
External EnvironmentRoles amp Skills
ManagementProcess
InformationTechnology
Fig 79 IT-organizational fit model by Scott Morton
7 Decision Models Regarding Electronic Health Records
176
during the use of HDM for the second study explained in this exam Just like in the previous model the components of the model are arranged in an ascending hierar-chical order At each level those criteria and subcriteria are compared with each other using a pairwise comparison scheme (also explained in the previous study) The questionnaire will be administered online through Qualtrics and the results will be put into PCM software for pairwise comparisons as well as Excel and pos-sibly SPSS to analyze some additional demographic and other information (age gender job position years of experience years of experience with EHR type and brand of EHR system implemented year of implementation number of implemen-tation (first system or replacement))
Table 75 Explanation of impact criteria through evaluation measures
Impact criteria Subcriteria Evaluation measures
Technology System quality Data accuracy data currency database contents ease of use ease of learning availability usefulness of system features and functions flexibility reliability technical support security efficiency resource utilization response time turnaround time
Information quality
Importance relevance usefulness legibility format accuracy conciseness completeness reliability timeliness data entry methods
Service quality Quick responsiveness assurance empathy follow-up service technical support
Human System use Amountduration (number of inquiries amount of connect time number of functions used number of records accessed frequency of access frequency of report requests number of reports generated) use by whom (direct vs chauffeured use) actual vs reported use nature of use (use for intended purpose appropriate use type of information used) purpose of use level of use (general vs specific) recurring use report acceptance percentage used voluntaries of use motivation to use attitude expectationsbelief knowledgeexpertise acceptance resistancereluctance training
User satisfaction
Satisfaction with specific functions overall satisfaction perceived usefulness enjoyment software satisfaction decision-making satisfaction
Organization Structure Nature (type size) culture planning strategy management clinical process autonomy communication leadership top management support medical sponsorship champion mediator teamwork
Environment Financial source government politics localization competition interorganizational relationship population served external communication
Net benefits Clinical practice (job effects task performance productivity work volume morale) efficiency effectiveness (goal achievement service) decision- making quality (analysis accuracy time confidence participation) error reduction communication clinical outcomes (patient care morbidity mortality) cost
L Hogaboam and TU Daim
177
TECHNOLOGY
HUMAN
ORGANIZATION
SystemQuality
InformationQuality
ServiceQuality
System Use
Net Benefits
Fit
Influence
User Satisfaction
Structure
Environment
Fig 710 The HOT-fit proposed framework (Yusof et al 2008)
Some open-ended questions will be asked in this questionnaire since they may provide important qualitative information and depending on the response rate will be used for further descriptive or other statistical analysis for example
bull How many clinical measures are reported by your systembull What clinical measures are reported by your system Please at least name the
main five you use or perceive useful if there are too many to reportbull What are the three major benefits to your practice from EHRbull What are the three main frustrations with your EHRbull Are you happy with your EHR system (5-point Likert scale) Why
(Fig 711)
Impacts of EHR system
Technological
Sys
tem
Qua
lity
Info
rmat
ion
Qua
lity
Ser
vice
Qua
lity
Human
Sys
tem
Use
Use
r S
atis
fact
ion
Organizational
Str
uctu
re
Env
ironm
ent
Net BenefitsC
linic
al
Fin
anci
ial
Fig 711 HDM of EHR impacts (Study 3)
7 Decision Models Regarding Electronic Health Records
178
The following hypotheses will be analyzed
HA1-A3 Quality measures (system quality information quality and service quality) have higher importance as EHR impact from physicianrsquos point of view
HB1-B2 EHR use greatly impacts organizational criteria of structure and environment
HC EHR use improves clinical outcomesHD EHR use saves costs
References
Aggelidis VP Chatzoglou PD (2009) Using a modified technology acceptance model in hospitals International Journal of Medical Informatics 78(2)115ndash126 Retrieved October 29 2012 from httpwwwncbinlmnihgovpubmed18675583
Ajzen I Madden TJ (1986) Prediction of goal-directed behavior Attitudes intentions and per-ceived behavioral control Journal of Experimental Social Psychology 22(5)453ndash474 Retrieved from httplinkinghubelseviercomretrievepii0022103186900454
Alkhateeb FM Khanfar NM Loudon D (2009) Physiciansrsquo adoption of pharmaceutical E-detailing application of Rogers innovation-diffusion model Services Marketing Quarterly 31(1) 116ndash132 Retrieved November 12 2012 from httpwwwtandfonlinecomdoiabs101080 15332960903408575
Alper J amp Olson S (2010) Report to the President realizing the full potential of health informa-tion technology to improve healthcare for Americans The path forward
Andreacute B et al (2008) Experiences with the implementation of computerized tools in health care units A review article International Journal of Human-Computer Interaction 24(8)753ndash775 Retrieved November 12 2012 from httpwwwtandfonlinecomdoiabs101080 10447310802205768
Angst CM et al (2010) Social contagion and information technology diffusion The adoption of electronic medical records in US hospitals Management Science 56(8)1219ndash1241 Retrieved November 12 2012 from httpmanscijournalinformsorgcgidoi101287mnsc11001183
Ash J Bates D (2005) Factors and forces affecting EHR system adoption report of a 2004 ACMI discussion Journal of the American Medical Informatics 128ndash13 Retrieved May 15 2012 from httpwwwsciencedirectcomsciencearticlepiiS1067502704001495
Ash J S et al (2001) A diffusion of innovations model of physician order entry Proceedings of the AMIA hellip Annual symposium AMIA Symposium (pp 22ndash6) httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2243456amptool=pmcentrezamprendertype=abstract
Ayatollahi H Bath PA Goodacre S (2009) Paper-based versus computer-based records in the emergency department staff preferences expectations and concerns Health Informatics Journal 15(3)199ndash211 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed19713395
Bates DW et al (2003) A proposal for electronic medical records in US primary care Journal of American Informatics Association 10(1)1ndash10
Becker A et al (2011) A new computer-based counselling system for the promotion of physical activity in patients with chronic diseasesndashresults from a pilot study Patient Education and Counseling 83(2)195ndash202 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed20573467
Beckett M et al (2011) Bridging the gap between basic science and clinical practice The role of organizations in addressing clinician barriers Implementation Science 6(1)35 Retrieved May 14 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=3086857amptool=pmcentrezamprendertype=abstract
L Hogaboam and TU Daim
179
Blumenthal D (2009) Stimulating the adoption of health information technology New England Journal of Medicine 360(15)1477ndash1479 Retrieved May 14 2012 from httpwwwnejmorgdoifull101056NEJMp0901592
Blumenthal D (2010) Launching HITECH The New England Journal of Medicine 362(5)382ndash385 httpwwwncbinlmnihgovpubmed20042745
Bodenheimer T Grumbach K (2003) Electronic technology a spark to revitalize primary care JAMA 290(2)259ndash264
Boonstra A Broekhuis M (2010) Barriers to the acceptance of electronic medical records by physi-cians from systematic review to taxonomy and interventions BMC Health Services Research 10231 httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2924334amptool=pmcentrezamprendertype=abstract
Bowens F M Frye P A amp Jones W A (2010) Health information technology integration of clinical workflow into meaningful use of electronic health records Perspectives in health infor-mation managementAHIMA American Health Information Management Association 7 p 1d httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2966355amptool=pmcentrezamprendertype=abstract
Box TL et al (2010) Strategies from a nationwide health information technology implementation the VA CART story Journal of General Internal Medicine 25(Suppl 1)72ndash76 Retrieved March 6 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2806964amptool=pmcentrezamprendertype=abstract
Brand C et al (2005) Clinical practice guidelines barriers to durability after effective early implementation Internal Medicine Journal 35(3)162ndash169 httpwwwncbinlmnihgovpubmed15737136
Bufalino V J et al 2011 The American Heart Associationrsquos recommendations for expanding the applications of existing and future clinical registries a policy statement from the American Heart Association Retrieved May 14 2012 from httpwwwncbinlmnihgovpubmed 21482960
Burton-Jones A Hubona GS (2006) The mediation of external variables in the technology accep-tance model Information and Management 43(6)706ndash717 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0378720606000504
Carayon P et al (2011) ICU nursesrsquo acceptance of electronic health records Journal of the American Medical Informatics Association 18(6)812ndash819 Retrieved November 8 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=3197984amptool=pmcentrezamprendertype=abstract
Chau PYK Hu PJ-H (2002) Investigating healthcare professionalsrsquo decisions to accept telemedi-cine technology An empirical test of competing theories Information and Management 39(4)297ndash311 httplinkinghubelseviercomretrievepiiS0378720601000982
Chaudhry B et al (2006) Systematic review Impact of health information technology on qual-ity efficiency and costs of medical care Annals of Internal Medicine 144(10) 742ndash752 Wndash168 ndashWndash185
Chen R-F Hsiao J-L (2012) An investigation on physiciansrsquo acceptance of hospital information systems A case study International Journal of Medical Informatics 601ndash11 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed22652011
Chen Y-P et al (2010) An agile enterprise regulation architecture for health information security management Telemedicine Journal and E-Health 16(7)807ndash817 Retrieved April 24 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2956519amptool=pmcentrezamprendertype=abstract
Cheng Y-M 2012 Exploring the roles of interaction and flow in explaining nursesrsquo e-learning acceptance Nurse Education Today Retrieved November 10 2012 from httpwwwncbinlmnihgovpubmed22405340
Chiasson M et al (2007) Expanding multi-disciplinary approaches to healthcare information tech-nologies what does information systems offer medical informatics International Journal of Medical Informatics 76(Suppl 1)S89ndashS97 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed16769245
7 Decision Models Regarding Electronic Health Records
180
Choi YK Totten JW (2012) Self-construalrsquos role in mobile TV acceptance Extension of TAM across cultures Journal of Business Research 65(11)1525ndash1533 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0148296311000695
Chow M Chan L et al 2012 Exploring the intention to use a clinical imaging portal for enhancing healthcare education Nurse Education Today 1ndash8 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed22336478
Chow M Herold DK et al (2012b) Extending the technology acceptance model to explore the intention to use Second Life for enhancing healthcare education Computers and Education 59(4)1136ndash1144 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0360131512001327
Chumbler NR Haggstrom D Saleem JJ (2011) Implementation of health information technology in Veterans Health Administration to support transformational change telehealth and personal health records Medical Care 49(Suppl 12)S36ndashS42 httpwwwncbinlmnihgovpubmed 20421829
Cresswell K amp Sheikh A (2012) Organizational issues in the implementation and adoption of health information technology innovations An interpretative review International Journal of Medical Informatics Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS1386505612001992
Davidson S Heineke J (2007) Toward an effective strategy for the diffusion and use of clinical information systems Journal of the American Medical Association 14(3)361ndash367 Retrieved November 12 2012 from http17167114118content143361abstract
Davis FD (1985) A technology acceptance model for empirically testing new end-user information systems Theory and results Massachusetts Institute of Technology Sloan School of Management ∎ httpenscientificcommonsorg7894517
Davis F (1989) User acceptance of computer technology a comparison of two theoretical models Management Science 35(8)982ndash1003 Retrieved November 12 2012 from httpmansci journalinformsorgcontent358982short
Davis F (1993) User acceptance of information technology system characteristics user percep-tions and behavioral impacts International Journal of Man-Machine Studies 38475ndash487 Retrieved November 12 2012 from httpdeepbluelibumicheduhandle20274230954
Davis FD Venkatesh V (1996) A critical assessment of potential measurement biases in the tech-nology acceptance model three experiments International Journal of Human-Computer Studies 45(1)19ndash45 httplinkinghubelseviercomretrievepiiS1071581996900403
Degoulet P Jean FC Safran C (1995) The health care professional multimedia workstation development and integration issues International Journal of Bio-Medical Computing 39(1)119ndash125 httpwwwncbinlmnihgovpubmed7601524
DeLia D et al (2004) What matters to low-income patients in ambulatory care facilities Medical Care Research and Review 61(3)352ndash375 Retrieved May 14 2012 from httpwwwncbinlmnihgovpubmed15358971
DePhillips H (2007) Initiatives and barriers to adopting health information technology A US per-spective Disease Management Health Outcomes 15(1)1ndash6 Retrieved May 10 2012 from httpwwwingentaconnectcomcontentadisdmho20070000001500000001art00001
DesRoches CM et al (2008) Electronic health records in ambulatory care mdash A national survey of physicians The New England Journal of Medicine 35950ndash60
Dillon A Morris MG (1996) User acceptance of new information technology - Theories and mod-els Annual Review of Information Science and Technology 313ndash32 Williams M (ed)
Dixon DR (1999) The behavioral side of information technology International Journal of Medical Informatics 56(1-3)117ndash123 httpwwwncbinlmnihgovpubmed10659940
Dixon BE Zafar A Overhage JM (2010) A Framework for evaluating the costs effort and value of nationwide health information exchange Journal of the American Medical Informatics Association 17(3)295ndash301 Retrieved March 14 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2995720amptool=pmcentrezamprendertype=abstract
L Hogaboam and TU Daim
181
Dulcic Z Pavlic D Silic I (2012) Evaluating the intended use of Decision Support System (DSS) by applying Technology Acceptance Model (TAM) in business organizations in Croatia Procedia ndash Social and Behavioral Sciences 581565ndash1575 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS1877042812046058
Duumlnnebeil S et al (2012) Determinants of physiciansrsquo technology acceptance for e-health in ambu-latory care International Journal of Medical Informatics 81(11)746ndash760 Retrieved November 6 2012 from httpwwwncbinlmnihgovpubmed22397989
Fonkych K Taylor R (2005) The state and pattern of health information technology adoption Retrieved May 10 2012 from httpbooksgooglecombookshl=enamplr=ampid=qiALR-nsUrcCampoi=fndamppg=PP1ampdq=The+State+and+Pattern+of+Health+Information+Technology+Adoptionampots=Esaxti6UfVampsig=5XaJzkf0bVuTuwVPnZs5ybWZ8n4
Ford E Menachemi N Phillips T (2006) Predicting the adoption of electronic health records by physicians When will health care be paperless Journal of the American Medical Inform Assoc 13106ndash113 Retrieved May 14 2012 from httpjamiabmjjournalscomcon-tent131106short
Frambach RT Schillewaert N (2002) Organizational innovation adoption a multi-level framework of determinants and opportunities for future research Journal of Business Research 55(2) 163ndash176 httplinkinghubelseviercomretrievepiiS0148296300001521
Furukawa MF (2011) Electronic medical records and the efficiency of hospital emergency depart-ments Medical Care Research and Review 68(1)75ndash95 Retrieved May 14 2012 from httpwwwncbinlmnihgovpubmed20555014
Glaser J et al (2008) Advancing personalized health care through health information technology An update from the American Health Information Communityrsquos Personalized Health Care Workgroup Journal of the American Medical Informatics Association 15(4)391ndash396
Goldberg DG (2012) Primary care in the United States the practice-based innovations and factors that influence adoption Journal of Health Organization and Management 26(1)81ndash97
Goldzweig C L et al(2009) Costs and benefits of health information technology new trends from the literature Health Affairs (Project Hope) 28(2) w282ndash93 Retrieved March 29 2012 from httpwwwncbinlmnihgovpubmed19174390
Goroll AH et al (2008) Community-wide implementation of health information technology the Massachusetts eHealth Collaborative experience Journal of the American Medical Informatics Association 16(1)132ndash139 Retrieved March 29 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2605598amptool=pmcentrezamprendertype=abstract
Greenhalgh T et al (2009) Tensions and paradoxes in electronic patient record research A system-atic literature review using the meta-narrative method The Milbank Quarterly 87(4)729ndash788 Retrieved May 14 2012 from httponlinelibrarywileycomdoi101111j1468-00092009 00578xfull
Handy J Hunter I Whiddett R (2001) User acceptance of inter-organizational electronic medical records Health Informatics Journal 7(2)103ndash107 Retrieved November 12 2012 from httpjhisagepubcomcgidoi101177146045820100700208
Hatton JD Schmidt TM Jelen J (2012) Adoption of electronic health care records physician heu-ristics and hesitancy Procedia Technology 5706ndash715 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS2212017312005099
Helfrich C D et al (2007) Adoption and implementation of mandated diabetes registries by community health centers American Journal of Preventive Medicine 33(1 Suppl) S50ndashS58 quiz S59ndash65 Retrieved May 14 2012 from httpwwwncbinlmnihgovpubmed17584591
Holden RJ Karsh B-T (2010) The technology acceptance model its past and its future in health care Journal of Biomedical Informatics 43(1)159ndash172 Retrieved October 26 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2814963amptool=pmcentrezamprendertype=abstract
Hung S-Y Ku Y-C Chien J-C (2012) Understanding physiciansrsquo acceptance of the Medline system for practicing evidence-based medicine a decomposed TPB model International Journal of Medical Informatics 81(2)130ndash142 Retrieved November 5 2012 from httpwwwncbinlmnihgovpubmed22047627
7 Decision Models Regarding Electronic Health Records
182
Im I Kim Y Han H-J (2008) The effects of perceived risk and technology type on usersrsquo accep-tance of technologies Information amp Management 45(1)1ndash9 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0378720607000468
Janczewski L Shi FX (2002) Development of information security baselines for healthcare infor-mation systems in New Zealand Computers amp Security 21(2)172ndash192 Retrieved November 12 2012 from httpwwwsciencedirectcomsciencearticlepiiS0167404802002122
Jeng DJ-F Tzeng G-H (2012) Social influence on the use of clinical decision support systems Revisiting the unified theory of acceptance and use of technology by the fuzzy DEMATEL technique Computers amp Industrial Engineering 62(3)819ndash828 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0360835211003895
Jian W-S et al (2012) Factors influencing consumer adoption of USB-based personal health records in Taiwan BMC Health Services Research 12(1)277 Retrieved November 12 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=3465237amptool=pmcentrezamprendertype=abstract
Jung S (2006) The perceived benefits of healthcare information technology adoption Construct and survey development Retrieved March 22 2013 from httpetdlsuedudocsavailableetd-11162006-125102
Karahanna E Straub DW (1999) The psychological origins of perceived usefulness and ease-of- use Information amp Management 35(4)237ndash250 httplinkinghubelseviercomretrievepiiS0378720698000962
Kazley AS Ozcan YA (2007) Organizational and environmental determinants of hospital EMR adoption A national study Journal of Medical Systems 31(5)375ndash384 Retrieved May 14 2012 from httpwwwspringerlinkcomindex101007s10916-007-9079-7
Kazley AS Ozcan YA (2008) Do hospitals with electronic medical records (EMRs) provide higher quality care An examination of three clinical conditions Medical Care Research and Review 65(4)496ndash513 Retrieved May 14 2012 from httpwwwncbinlmnihgovpubmed18276963
Kim D Chang H (2006) Key functional characteristics in designing and operating health informa-tion websites for user satisfaction an application of the extended technology acceptance model International Journal of Medical Informatics 76(11-12)790ndash800 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed17049917
King WR He J (2006) A meta-analysis of the technology acceptance model Information amp Management 43(6)740ndash755 Retrieved November 2 2012 from httplinkinghubelseviercomretrievepiiS0378720606000528
Kukafka R et al (2003) Grounding a new information technology implementation framework in behavioral science a systematic analysis of the literature on IT use Journal of Biomedical Informatics 36(3)218ndash227 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS1532046403000844
Kumar S Aldrich K (2010) Overcoming barriers to electronic medical record (EMR) implementa-tion in the US healthcare system A comparative study Health Informatics Journal 16(4)306ndash318 Retrieved March 12 2012 from httpwwwncbinlmnihgovpubmed21216809
Lanham HJ Leykum LK McDaniel RR (2012) Same organization same electronic health records (EHRs) system different use exploring the linkage between practice member communication patterns and EHR use patterns in an ambulatory care setting Journal of the American Medical Informatics Association 19382ndash391 Retrieved April 9 2012 from httpwwwncbinlmnihgovpubmed21846780
Lapinsky SE et al (2008) Survey of information technology in intensive care units in Ontario Canada BMC Medical Informatics and Decision Making 85 Retrieved March 16 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2233621amptool=pmcentrezamprendertype=abstract
Lee G Xia W (2011) A longitudinal experimental study on the interaction effects of persuasion quality user training and first-hand use on user perceptions of new information technology Information amp Management 48(7)288ndash295 Retrieved November 12 2012 from httplinkin-ghubelseviercomretrievepiiS0378720611000772
L Hogaboam and TU Daim
183
Legris P Ingham J Collerette P (2003) Why do people use information technology A critical review of the technology acceptance model Information amp Management 40(3)191ndash204 httplinkinghubelseviercomretrievepiiS0378720601001434
Leu MG et al (2008) Centers speak up the clinical context for health information technology in the ambulatory care setting Journal of General Internal Medicine 23(4)372ndash378 Retrieved March 1 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2359517amptool=pmcentrezamprendertype=abstract
Liang H Xue Y Chase SK (2011) Online health information seeking by people with physical dis-abilities due to neurological conditions International Journal of Medical Informatics 80(11)745ndash753 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed21917511
Linder JA et al (2007) Electronic health record use and the quality of ambulatory care in the United States Archives of Internal Medicine 167(13)1400ndash1405 httpwwwncbinlmnihgovpubmed17620534
Lorence DP Churchill R (2005) Incremental adoption of information security in health-care orga-nizations Implications for document management IEEE Transactions on Information Technology in Biomedicine 9(2)169ndash173 httpwwwncbinlmnihgovpubmed16138533
Lorenzi NM et al (2009) How to successfully select and implement electronic health records (EHR) in small ambulatory practice settings BMC Medical Informatics and Decision Making 9(15)1ndash13 Retrieved May 14 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2662829amptool=pmcentrezamprendertype=abstract
Ludwick DA Doucette J (2009) Adopting electronic medical records in primary care lessons learned from health information systems implementation experience in seven countries International Journal of Medical Informatics 78(1)22ndash31 Retrieved February 29 2012 from httpwwwncbinlmnihgovpubmed18644745
Maumlenpaumlauml T et al (2009) The outcomes of regional healthcare information systems in health care a review of the research literature International Journal of Medical Informatics 78(11)757ndash771 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed19656719
Martich G amp Cervenak J (2007) Eyes wide shut The ldquohiddenrdquo costs of deploying health infor-mation technology Journal of Critical Care 7ndash8 Retrieved November 12 2012 from httpwwwjournalselsevierhealthcomperiodicalsyjcrcarticleS0883-9441(06)00217-6abstract
McFarland DJ Hamilton D (2006) Adding contextual specificity to the technology acceptance model Computers in Human Behavior 22(3)427ndash447 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS074756320400130X
McGinn CA et al (2011) Comparison of user groupsrsquo perspectives of barriers and facilitators to implementing electronic health records A systematic review BMC Medicine 9(46)1ndash10 httpwwwpubmedcentralnihgovarticlerenderfcgiartid=3103434amptool=pmcentrezamprendertype=abstract
Melas CD et al (2011) Modeling the acceptance of clinical information systems among hospital medical staff an extended TAM model Journal of Biomedical Informatics 44(4)553ndash564 Retrieved November 7 2012 from httpwwwncbinlmnihgovpubmed21292029
Menachemi N Brooks RG (2006) Reviewing the benefits and costs of electronic health records and associated patient safety technologies Journal of Medical Systems 30(3)159ndash168 Retrieved March 27 2012 from httpwwwspringerlinkcomindex101007s10916-005- 7988-x
Menachemi N et al (2008) The relationship between local hospital IT capabilities and physician EMR adoption Journal of Medical Systems 33(5)329ndash335 Retrieved May 14 2012 from httpwwwspringerlinkcomindex101007s10916-008-9194-0
Miller RH Sim I (2004) Physiciansrsquo use of electronic medical records barriers and solutions Health Affairs (Project Hope) 23(2)116ndash126 httpwwwncbinlmnihgovpubmed22533131
Moores TT (2012) Towards an integrated model of IT acceptance in healthcare Decision Support Systems 53(3)507ndash516 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0167923612001108
7 Decision Models Regarding Electronic Health Records
184
Morton M E amp Wiedenbeck S (2009) A framework for predicting EHR adoption attitudes a physician survey Perspectives in health information management AHIMA American Health Information Management Association 6 p1a httpwwwpubmedcentralnihgovarticleren-derfcgiartid=2804456amptool=pmcentrezamprendertype=abstract
Morton M E amp Wiedenbeck S (2010) EHR acceptance factors in ambulatory care a survey of physician perceptions Perspectives in health information management AHIMA American Health Information Management Association 7 p1c httpwwwpubmedcentralnihgov articlerenderfcgiartid=2805555amptool=pmcentrezamprendertype=abstract
Ortega Egea JM Romaacuten Gonzaacutelez MV (2011) Explaining physiciansrsquo acceptance of EHCR sys-tems An extension of TAM with trust and risk factors Computers in Human Behavior 27(1)319ndash332 Retrieved November 7 2012 from httplinkinghubelseviercomretrievepiiS0747563210002530
Pai F-Y Huang K-I (2011) Applying the technology acceptance model to the introduction of healthcare information systems Technological Forecasting and Social Change 78(4) 650ndash660 Retrieved November 12 2012 from httplinkinghubelseviercomretrievepiiS0040162510002714
Palacio C Harrison JP Garets D (2009) Benchmarking electronic medical records initiatives in the US a conceptual model Journal of Medical Systems 34(3)273ndash279 Retrieved May 12 2012 from httpwwwspringerlinkcomindex101007s10916-008-9238-5
Pareacute G Sicotte C (2001) Information technology sophistication in health care an instrument vali-dation study among Canadian hospitals International Journal of Medical Informatics 63(3)205ndash223 httpwwwncbinlmnihgovpubmed11502433
Police RL Foster T Wong KS (2011) Adoption and use of health information technology in physi-cian practice organisations Systematic review Informatics in Primary Care 18245ndash259
Rahimpour M et al (2008) Patientsrsquo perceptions of a home telecare system International Journal of Medical Informatics 77(7)486ndash498 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed18023610
Randeree E (2007) Exploring physician adoption of EMRs A multi-case analysis Journal of Medical Systems 31(6)489ndash496 Retrieved April 23 2012 from httpwwwspringerlinkcomindex101007s10916-007-9089-5
Rind D M amp Safran C (1993) Real and imagined barriers to an electronic medical record Computer Application in Medical Care 74ndash78 Retrieved May 15 2012 from httpwwwncbinlmnihgovpmcarticlesPMC2248479
Rosemann T et al (2010) Utilisation of information technologies in ambulatory care in Switzerland Swiss Medical Weekly 140(September) pw 13088 Retrieved April 20 2012 from httpwwwncbinlmnihgovpubmed20853193
Roth CP et al (2009) The challenge of measuring quality of care from the electronic health record American Journal of Medical Quality 24(5)385ndash394 Retrieved May 14 2012 from httpwwwncbinlmnihgovpubmed19482968
Schoen C et al (2006) On the front lines of care primary care doctorsrsquo office systems experi-ences and views in seven countries Health Affairs (Project Hope) 25(6) w555ndashw571 Retrieved March 15 2012 from httpwwwncbinlmnihgovpubmed17102164
Shen JJ Ginn GO (2012) Financial position and adoption of electronic health records a retrospec-tive longitudinal study Journal of Health Care Finance 38(3)61ndash77 Retrieved May 15 2012 from httpwwwncbinlmnihgovpubmed22515045
Shields AE et al (2007) Adoption of health information technology in community health centers results of a national survey Health Affairs (Project Hope) 26(5)1373ndash1383 Retrieved March 26 2012 from httpwwwncbinlmnihgovpubmed17848448
Simon S et al (2007) Correlates of electronic health record adoption in office practices A statewide survey Journal of the American Medical Informatics Association 14(1)110ndash117 Retrieved May 15 2012 from httpwwwsciencedirectcomsciencearticlepiiS1067502706002143
Simon S et al (2008) Electronic health records Which practices have them and how are clinicians using them Journal of Evaluation in Clinical Practice 1443ndash47 Retrieved May 15 2012 from httponlinelibrarywileycomdoi101111j1365-2753200700787xfull
L Hogaboam and TU Daim
185
Storey J Buchanan D (2008) Healthcare governance and organizational barriers to learning from mistakes Journal of Health Organisation and Management 22(6)642ndash651 Retrieved November 12 2012 from httpwwwemeraldinsightcom10110814777260810916605
Tsiknakis M Katehakis DG Orphanoudakis SC (2002) An open component-based information infrastructure for integrated health information networks International Journal of Medical Informatics 68(1-3)3ndash26 httpwwwncbinlmnihgovpubmed12467787
Valdes I et al (2004) Barriers to proliferation of electronic medical records Informatics in Primary Care 123ndash9 Retrieved May 15 2012 from httpwwwingentaconnectcomcontentrmpipc20040000001200000001art00002
Vedvik E Tjora AH Faxvaag A (2009) Beyond the EPR Complementary roles of the hospital- wide electronic health record and clinical departmental systems BMC Medical Informatics and Decision Making 929 Retrieved May 10 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2700794amptool=pmcentrezamprendertype=abstract
Vishwanath A Brodsky L Shaha S (2009) Physician adoption of personal digital assistants (PDA) Testing its determinants within a structural equation model Journal of Health Communication 14(1)77ndash95 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed19180373
Wagner H amp Weibel S (2005) The Dublin Core Metadata Registry Requirements implementa-tion and experience Journal of Digital Information 1ndash20 Retrieved May 15 2012 from httpdialnetuniriojaesservletarticulocodigo=1416626
Weiner BJ et al (2011) Use of qualitative methods in published health services and management research a 10-year review Medical Care Research and Review 68(1)3ndash33 Retrieved March 4 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=3102584amptool=pmcentrezamprendertype=abstract
Wu J-H Chen Y-C Greenes RA (2009) Healthcare technology management competency and its impacts on IT-healthcare partnerships development International Journal of Medical Informatics 78(2)71ndash82 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed18603470
Wu J-H Wang S-C Lin L-M (2007) Mobile computing acceptance factors in the healthcare indus-try a structural equation model International Journal of Medical Informatics 76(1)66ndash77 Retrieved November 12 2012 from httpwwwncbinlmnihgovpubmed16901749
Yang H (2004) Itrsquos all about attitude revisiting the technology acceptance model Decision Support Systems 38(1)19ndash31 Retrieved November 9 2012 from httpportlandstateworldcatorgtitleits-all-about-attitude-revisiting-the-technology-acceptance-modeloclc198488645amp referer=brief_results
Yarbrough AK Smith TB (2007) Technology acceptance among physicians A new take on TAM Medical Care Research and Review 64(6)650ndash672 Retrieved May 14 2012 from httpwwwncbinlmnihgovpubmed17717378
Yi MY et al (2006) Understanding information technology acceptance by individual professionals Toward an integrative view Information amp Management 43(3)350ndash363 Retrieved November 4 2012 from httplinkinghubelseviercomretrievepiiS0378720605000716
Yoon-Flannery K et al (2008) A qualitative analysis of an electronic health record (EHR) imple-mentation in an academic ambulatory setting Informatics in Primary Care 16277ndash285
Yu P Li H Gagnon M-P (2009) Health IT acceptance factors in long-term care facilities a cross- sectional survey International Journal of Medical Informatics 78(4)219ndash229 Retrieved November 7 2012 from httpwwwncbinlmnihgovpubmed18768345
Yusof MM et al (2008) An evaluation framework for Health Information Systems human organi-zation and technology-fit factors (HOT-fit) International Journal of Medical Informatics 77(6)386ndash398 Retrieved October 29 2012 from httpwwwncbinlmnihgovpubmed 17964851
Zandieh SO et al (2008) Challenges to EHR implementation in electronic- versus paper-based office practices Journal of General Internal Medicine 23(6)755ndash761 Retrieved April 15 2012 from httpwwwpubmedcentralnihgovarticlerenderfcgiartid=2517887amptool=pmcentrezamprendertype=abstract
Zaroukian MH (2006) Benefiting from ambulatory EHR implementation Solidarity six sigma and willingness to strive JHIM 20(1)53ndash60
7 Decision Models Regarding Electronic Health Records
Part III Adoption Factors of Electronic Health
Record Systems
Orhun M Koumlk Nuri Basoglu and Tugrul U Daim
Todayrsquos rapidly changing regulations increasing healthcare costs and most impor-tantly globalization have made health record keeping an important issue Electronic health record systems are rising as a crucial and unavoidable way of record keeping for healthcare However as other information technology implementations elec-tronic health records also have their own adoption processes and diffusion factors The main goal of this study is to defi ne a model to analyze adoption process of electronic health record systems and to understand the diffusion factors
Results of the study indicate that there are different factors affecting the adop-tion process via a literature research and quantitative fi eld survey Model has been tested and constructs have been grouped under intermediary dependent and exter-nal factors
189copy Springer International Publishing Switzerland 2016 TU Daim et al Healthcare Technology Innovation Adoption Innovation Technology and Knowledge Management DOI 101007978-3-319-17975-9_8
Chapter 8 Adoption Factors of Electronic Health Record Systems
Orhun Mustafa Koumlk Nuri Basoglu and Tugrul U Daim
Todayrsquos rapidly changing regulations increasing healthcare costs and most importantly globalization have made health record keeping an important issue Electronic health record systems are rising as a crucial and unavoidable way of record keeping for healthcare However as other information technology imple-mentations electronic health records also have their own adoption processes and diffusion factors The main goal of this study is to defi ne a model to analyze the adoption process of electronic health record systems and to understand the diffusion factors
Results of the study indicate that there are different factors affecting the adoption process via a literature research and quantitative fi eld survey Models have been tested and constructs have been grouped under intermediary dependent and exter-nal factors
81 Introduction
In Turkey 368 of the people over the age of 15 have health problems affecting their daily activities (Turkstat Health Statistics 2012a 2012b ) Seventy-six percent of the healthcare expenditure in Turkey is conducted via government in 2011
O M Koumlk PwC Strategyamp Ernst and Young Advisory Istanbul Turkey
N Basoglu İzmir Institute of Technology Urla Turkey
T U Daim () Portland State University Portland OR USA e-mail tugruludaimpdxedu
190
(Euromonitor 2012 ) In 2020 it is expected that 20 of the Turkish population will be older than 50 years (Euromonitor 2012) The Ministry of Health has started a transformation program in 2003 and offering e-health services is an important part of the program The ministry has created database and data collection standards for all types of healthcare organizations (Ministry of Health Statistics 2012 ) In 2010 there are 16651 patient care institutions and ~123000 physicians in Turkey (Turkstat 2010 ) This proves that effi cient integration and information sharing are required between these institutions and physicians In order to establish this pur-pose the government is planning to integrate all healthcare organizations within a network and in the later steps telehealth and telemedicine applications will go live in the future (Ministry of Health Statistics 2012 )
Healthcare systems are facing with increasing demand rising costs inconsis-tency and lowering interoperability (Lluch 2011 ) As the increasing demand gets combined with the lowering funds of the governments healthcare providers started to look for less costly alternatives (Al-Qirim 2007 )
In our era with the innovations in the telecommunications and information tech-nologies the use of electronic services has increased in many areas Health is one of these areas affected by technologies In the last decades health information systems (HIS) have developed many new technologies Telemedicine telehealth and elec-tronic health records can be counted as the main areas in this industry (Haux 2010 ) Behkami and Daim stated that electronic health records and their adoption are an important research area for technology adoption and medical information research-ers ( 2012 ) Technology is used in many areas in health services Medical informat-ics is a discipline which focuses on data storing processing and information and knowledge management related to healthcare (Haux 2010 )
Health information systems are used by many different types of users such as patients doctors administration employees and application developers So they all have diffi culties in both using and developing these systems This research will focus on the factors that affect users using the electronic health record (EHR) from the technological and organizational perspective
As the healthcare processes are getting more complicated the public expects to move from hard copy of records to electronic-based record keeping (Tavakoli Jahanbakhsh Mokhtari amp Tadayon 2011 ) On the other hand many healthcare IT projects are failing or being abandoned due to lack of understanding of the health-care adoption factors (Kijsanayotin Pannaruthonai amp Speedie 2009 )
Healthcare providers and payers need more collaboration and communication than they ever did (Al-Qirim 2007 ) Electronic health records are an important layer to establish this communication Healthcare providers who try to implement health information systems face with challenging problems in technical social and organizational areas (Ovretveit Scott Rundall Shortell amp Brommels 2007 )
This study has been conducted to bring an understanding to the adoption factors of EHR systems To reach this goal diffusion of information systems diffusion of
OM Koumlk et al
191
health information systems and diffusion of electronic health records have been analyzed This study has researched and sought answers for the following topics
bull The technology diffusion process and factors affecting the technology adoption bull Health information system implementation and main barriers affecting the
implementation process bull Electronic health record evolution and main benefi ts of electronic health record
usage bull Electronic health record diffusion models and factors affecting the electronic
health record adoption process
82 Literature Review
821 Electronic Health Records
The International Organization for Standardization defi nes the electronic health record as a digital information format which contains the health progress of a patient (ISO 2005 ) The electronic health record is also implied as a computerized patient record (CPR) computer-based patient record computerized medical record elec-tronic medical record (EMR) electronic patient record (EPR) electronic healthcare record (EHCR) virtual EHR and digital medical record (DMR) which all have been determined during the last 30 years (Wen Ho Wen-Shan Li amp Hsu 2007 )
Developments in technology and health information systems would result to increase in the quality of healthcare (Tange Hasman Robbe amp Schouten 1997 ) However the developments in technology and telecommunications have not really improved the EHR systems (Brender Nohr amp McNair 2000 )
EHR systems are used by different types of users such as healthcare professionals and upper management Moreover healthcare professionals including physicians nurses radiologists pharmacists laboratory technicians and radiographers use differ-ent modules of EHR systems (Hayrinen et al 2008 ) Early adopters of EHR systems have already started to develop and expand their systems (Collins amp Wagner 2005 )
Transition from old paper-based records to new electronic record systems is a hard and long process which needs to satisfy several stakeholders (Estebaranz amp Castellano 2009 )
As demand of health system stakeholders increases too much healthcare providers cannot serve them until new developments have been taken in (Ludwick amp Doucette 2009 ) EHR 2003 systems are preferred over the paper-based records in the meaning of being portable more accurate and easier to report and also because in some cases they can be used as input for decision support systems (Holbrook et al 2003 )
An electronic healthcare record should include information about patientrsquos con-ditions and situation for doctors administrative data for administrative services and data required for the management of the healthcare organization (Estebaranz amp Castellano 2009 ) Moreover electronic health record systems can be used as a great
8 Adoption Factors of Electronic Health Record Systems
192
input for decision support systems with their long-term storage functionality reliable data structure and exceptional sharing capabilities (Hannan 1999 ) Usage of EHR may lead to reducing costs enhancing higher quality of care increased reli-ability and access to more accurate results (Kierkegaard 2011 ) Changing policies healthcare payers and governments require more accurate standardized and detailed data in order to clearly understand the situation to develop statistics and to segment their customers (Gonzalez-Heydrich et al 2000 ) Electronic health records can play an important role to fulfi ll these requirements (Gonzalez-Heydrich et al 2000 ) Although there are many policies regulating the electronic health record and healthcare information systems they are not totally practiced (Ovretveit et al 2007 ) All countries are changing their system from paper-based records to elec-tronic health records however only some of them could succeed in this operation (Jahanbakhsh Tavakoli amp Mokhtari 2011 ) Health information technologies and electronic health records are rising as a method to increase quality of care produc-tivity and security (Jha Doolan Grandt Scott amp Bates 2008 ) Also EHR offers an easy process for disease management processes with its functionalities and easy sharing (Wright et al 2009 )
822 Technology Adoption Models
Some models have been defi ned to understand the behaviors of people in the adop-tion process The theory of reasoned actions (Fishbein amp Ajzen 1975 ) Technology Acceptance Model (Davis 1989 ) Technology Acceptance Model 2 (Venkatesh amp Davis 2000 ) and unifi ed theory of acceptance and use of technology (Venkatesh Morris Davis amp Davis 2003 ) can be taken as the most signifi cant ones Also most of the researchers are taking these models as base asset and then specify their researches on these
The theory of reasoned action which can be seen in Fig 81 takes subjective norm and attitude toward act as its main constructs Subjective norm refers to ldquothe personrsquos beliefs that specifi c individuals or groups think heshe should or should not perform the behavior and hisher motivation to comply with the specifi c referentsrdquo (Fishbein amp Ajzen 1975 ) on the other hand attitude refers to ldquothe personrsquos beliefs that the behavior leads to certain outcomes and hisher evaluations of these out-comesrdquo (Fishbein amp Ajzen 1975 )
Attitude Toward Act
Subjective Norm
Behaviroal Intention Behavior
Fig 81 Theory of reasoned actions (Fishbein amp Ajzen 1975 )
OM Koumlk et al
193
Davis came up with the idea of the Technology Acceptance Model ( 1989 ) Perceived usefulness and perceived ease of use are taken as the two main drivers In fi nal behavioral intention brings the actual use result (Davis 1989 ) This modelrsquos main purpose is to predict user adoption behavior toward the technological develop-ments Figure 82 explains how the Technology Acceptance Model (TAM) is struc-tured (Davis 1989 ) TAM can be considered a future step for the theory of reasoned actions (Fishbein amp Ajzen 1975 ) and theory of planned behavior (Ajzen 1991 )
Venkatesh and Davis have made some additions to the Technology Acceptance Model and developed a further model with new factors in 2000 Factors such as experience and voluntariness affect the perceived usefulness Also the perceived ease of use has determinants such as subjective norm image job relevance output quality and demonstrability (Venkatesh amp Davis 2000 ) In Fig 83 TAM2 is explained (Venkatesh amp Davis 2000 )
Perceived Ease of Use
Attitude BehavioralIntention
Perceived Usefulness
Fig 82 Technology Acceptance Model (Davis 1989 )
Image
Job Relevance
Output Quality
Subjective Norm
Result Demonstability
Experience Voluntariness
Perceived Usefulness
Perceived Ease of Use
Attitude Behavioral Intention
Fig 83 Technology Acceptance Model 2 (Venkatesh amp Davis 2000 )
8 Adoption Factors of Electronic Health Record Systems
194
The unifi ed theory of acceptance and use of technology (UTAUT) has been defi ned by Venkatesh et al as a combination of different adoption theories such as the Technology Acceptance Model theory of reasoned actions and theory of planned behavior ( 2003 )
UTAUT (Fig 84 ) has three direct determinants on behavioral intention to use such as expectations from performance expectations from effort and the infl uence of the social environment (Venkatesh et al 2003 ) Intention to use and facilitating conditions affect the use behavior (Venkatesh et al 2003 )
DeLone and McLean have proposed a model for information systems success which correlates system quality and information quality with the actual system use and user satisfaction (1992) Furthermore it is stated that these categories are mul-tidimensional and also affect both individual and organizational impact (DeLone amp McLean 1992 ) (Fig 85 )
In 2003 the information systems success model has been updated and new vari-ables have been added intention to use net benefi ts and service quality (DeLone amp McLean) (Fig 86 )
Performance Expectancy
Effort Expectancy
Social Influence
Facilitating Conditions
Behavioral Intention Use Behavior
Gender Age Experience Voluntariness
Fig 84 UTAUT (Venkatesh et al 2003 )
System Quality
Information Quality
Use
User Satisfaction
Individual Impact
Organizational Impact
Fig 85 Information systems success model (DeLone amp McLean 1992 )
OM Koumlk et al
195
823 Health Information System Adoption
Researchers have developed adoption models specifi cally for health information systems
Yu and Gagnon have extended TAM2 and proposed taxonomy for health IT acceptance factors They have added subjective norm image and computer level as antecedent factors of ease of use Job role and subjective norm are defi ned as sub- factors of usefulness It is expressed that image has a negative effect on behavioral intention (Kargin et al 2009 ) (Fig 87 )
A further step has been taken on UTAUT and it is updated for hospital technol-ogy acceptance It is stated that anxiety has a negative effect on self-effi cacy (Aggelidis amp Chatzoglou 2009 ) Also self-effi cacy has positive effects on perceived ease of use and behavioral intention (Aggelidis amp Chatzoglou 2009 )
Electronic health records have different adoption factors than the other technolo-gies because their focus is mostly on the physicians and hospital administrations unlike the other technologies which mostly focus on citizen workers or students (Gagnon et al 2003 )
In order to increase the adoption effectiveness EHR systems have to be designed to be applicable with the workfl ows of the healthcare employees otherwise practi-cal application of the EHR system would take longer than expected (Hyun Johnson Stetson amp Bakken 2009 )
Another model combines the technology adoption model with new variables for health information adoption factors including computer self-effi cacy and perceived fi nancial cost variables (Tung amp Chang 2008 )
Health information-seeking behavior is related with EHR system usage Availability creditability and comprehensiveness are important factors in health information-seeking behavior (Basoglu et al 2010 ) Improved quality of care is an important adoption factor for EHR systems however privacy concern cost and implementation diffi culties are the main barriers (Greenshup 2012 )
International HL7 standards are defi ned in order to establish communication between healthcare organizations in terms of effi ciency with improved quality of care (Dosswell et al 2010 )
Information Quality
System Quality
Service Quality
Intention to Use
User Satisfaction
Net Benefits
Use
Fig 86 Updated information systems success model (DeLone amp McLean 2003)
8 Adoption Factors of Electronic Health Record Systems
196
The dynamically changing healthcare industry requires software which can adapt to new changes and a platform that works effi ciently at a low cost (Daim Basoglu amp Tan 2010 )
Unlike the old times present-day healthcare organizations need to combine tech-nology with information in order to meet the organizationrsquos IT requirements (Blue amp Tan 2010 )
Topacan stated that compatibility quality of support and information quality have a positive impact on usefulness (2011) On the other hand self-effi cacy has a positive effect on the ease of use (Topacan 2009 ) Figure 88 implies Topacanrsquos detailed model
Accesibility
Service Quality
Quality of Sup
Information Qua
Usage Time
Compatibility
Social Influence
Understandibility
Image
Cost
Ease of Use
Usefulness
Attitude
Intention
Self- Efficacy
Fig 88 Topacanrsquos e-health services framework (2009)
Image
Subjective Norm
Job Role
Computer Level
Usefulness
Ease of Use
Behavioral Intention
Fig 87 Health IT acceptance factors (Yu Li amp Gagnon 2009 )
OM Koumlk et al
197
Challenges during the implementation of EHR systems would be divided into two categories structural and infrastructural (Jahanbakhsh et al 2011 ) Infrastructural challenges can be summarized as IT-based problems communi-cation problems between stakeholders cultural problems and lack of require-ment analysis (Jahanbakhsh et al 2011 )
Usage of electronic records brings functionalities such as directly getting the required information through filtering and search capabilities (Wang Chase Markatou Hripcsak amp Friedman 2010 ) Selected information posi-tively affects quality of care and increases the performance of diagnosis (Wang et al 2010 )
Usability of the EHR software depends on many variables Rose et al defi ned the relationship with the usability of EHR systems with the user interface fl exibility and workfl ow of the implemented system ( 2005 ) Also Edwards et al said that fl ex-ibility and workfl ow are the main elements of the usability ( 2008 ) However it is implied that there is a trade-off between the fl exibility and consistency (Edwards Moloney Jacko amp Franccedilois 2008 )
According to Ross et al increasing quality of care effi ciency workfl ow man-agement and different functionalities are the main adoption factors of the health information systems ( 2010 ) It is stated that for each system users need different functionalities which are mainly described as the search ability through patient records report creation and electronic prescribing (Ross Schilling Fernald Davidson amp West 2010 )
A study which has been conducted in Korea has shown that adoption of the EHR systems has been generally blocked by lack of workfl ow-related EHR lack of IT knowledge and concern of privacy and security (Yoon Chang Kang Bae amp Park 2012 )
Vest has categorized EHR adoption factors under three groups technological organizational and environmental context ( 2010 ) Figure 89 implies the grouping of the factors
After the adoption of EHR systems organizations are looking for further benefi -ciary actions and auditing such as warningblocking a healthcare responsible of prescribing penicillin to someone who is already stated as allergic to penicillin (Brown amp Warmington 2002 )
One of the main adoption factors of EHR is standardized guidelines which can direct the user during the healthcare process and turn the processes in a standardized way starting with data entry and at each step of procedures (Vesely Zvarova Peleska Buchtela amp Zdenek 2006 )
Likourezos et al expressed that satisfaction of nurses and physicians mainly depends on computer experience perception regarding the use of EHR and EHRrsquos effects on quality of care ( 2004 )
Lenz and Kuhn implied that organizational structure vendor capabilities and changes in the processes with new software are the main barriers for EHR system adoption ( 2004 )
8 Adoption Factors of Electronic Health Record Systems
198
Iakovidis described that standardization effort for certain organizations cultural attitude and technological challenges are the main barriers for EHR implementation ( 1998 )
Sagiroglu stated that integration with other systems and devices is an important success factor of EHR systems ( 2006 ) It is identifi ed that functionalities of elec-tronic health records and its alignment with organizational structure can be taken as a leading adoption factor (Sagiroglu 2006 ) Meyer et al stated that adoption of electronic health record systems through the means of information saving heavily depends on the regulations regarding the privacy of personal records ( 1998 )
To ensure easier adoption health information systems are required to have fl ex-ible architecture which can easily fi t in to the new requirements of the users or technological developments (Toussiant amp Lodder 1998 )
For a successful adoption health information systems need to integrate with other systems or equipment with certain standards (Blazona amp Koncar 2007 ) Moreover electronic health records provide inter-organizational communication which offers a great chance for elderly people that need home care (Helleso amp Lorensen 2005 )
Technological Readiness
Certified EHR
Point-to-point connection technologies
Vertical Integration
Information Needs
Competition
Uncompensated care burden
Horizontal Integration
Control
Environmental Context
Organizational Context
Technological Context
Health information exchange adoption amp implementation
SizeOrganizational complexityNo of potential partnersDays cash on handUrban Rural
Control Variables
Fig 89 Categorization of adoption factors (Vest 2010)
OM Koumlk et al
199
83 Framework
In order to develop a model and taxonomy detailed literature review and semi- structured interviews have been conducted Constructs have been analyzed and then grouped under four categories external intermediary dependent and demographic categories
Table 81 implies the constructs that have been gathered via literature review and semi-structured interviews (L) refers to a construct that has been gathered from literature review (I) refers to a construct that has been gathered from the semi- constructed interviews
Literature has been deeply researched and factors affecting the technology adop-tion health information system adoption and electronic health record adoption have been analyzed Table 82 refers to the subjects and articles of the literature research
Thanks to the expert focus group and semi-structured interviews some of the constructs have been selected for a deeper analysis These constructs have struc-tured the base of our study The list of constructs and their explanations are implied in Table 83
Table 84 lists the major constructs and the literatures that they have been implied before
There are dependent items which are affected by the external factors via the intermediary factors
Table 81 Construct list from interviews and literature
Access validation (L) Disaster recovery (L) Reliability (L)
Accuracy (L) (I) Easy access (I) Reporting (I) Age (L) Ease of learning (I) Response time (L) Attitude (L) Ease of use (L) (I) Search ability (L) (I) Auditing L) Effi ciency (L) Self-confi dence (L) Authorization (L) (I) Flexibility (L) (I) Security (L) Comparison (L) (I) Image (L) Sharing (L) (I) Complexity of treatment (I) Integration (L) (I) Staff anxiety (L) Computer skills (L) Input effort (L) (I) Standardization (L) (I) Completeness (L) (I) Input time (L) Statistics (L) (I) Compatibility (L) Job experience (L) Subjective norm (L) Consistency (L) Job level (L) Support quality (L) (I) Copy (L) Medical assistant (I) Taskndashtechnology fi t (L) (I) Cost (L) Medical history (L) (I) Time saving (L) (I) Customization (L) (I) Normative beliefs (L) Training time (L) Data migration (L) Organization type (L) Usage goal (L) (I) Data preservation (L) (I) Online consultation (I) User interface (L) (I) Decision effectiveness (L) Privacy (L) (I) Usefulness (L) (I) Decision support system (L) Providepatient relations (L) Voluntariness (L) Developer support (I) Quality of care (L) (I)
8 Adoption Factors of Electronic Health Record Systems
200
H1 Usefulness of the systems positively affects the quality of care H2 Attitude toward the system use positively affects the quality of care
Quality of care provided by the physicians can be defi ned as rate of successful treatments and rate of successful diagnosis Higher quality of care can be reached with a more useful system and a more positive approach to the EHR usage (Brown amp Warmington 2002 Cho Kim Kim Kim amp Kim 2010 Collins amp Wagner 2005 Ludwick amp Doucette 2009 )
H3 Diffusion is positively affected by usefulness H4 Attitude signifi cantly and positively affects diffusion H5 Infusion is signifi cantly and positively affected by attitude H6 Infusion is signifi cantly and positively affected by ease of use
Usefulness and ease of use are important factors of an individualrsquos acceptance and wide usage of an information system (Davis 1989 Venkatesh amp Davis 2000 )
H7 Usefulness of the system positively affects the attitude toward system use H8 Ease of use of the system signifi cantly and positively affects the attitude toward
the system use
Table 82 Researched literature
Subject Article
Technology adoption models
Holden and Karsh ( 2010 ) Fishbein and Ajzen ( 1975 ) Ajzen and Fishbein ( 1980 ) Kerimoglu ( 2006 ) Davis ( 1989 ) Davis Jr ( 1985 ) Venkatesh and Davis ( 2000 ) Venkatesh et al ( 2003 ) Dishaw and Strong ( 1999 ) Kerimoglu Basoglu and Daim ( 2008 )
Health adoption models
Al-Qirim ( 2007 ) Aggelidis and Chatzoglou ( 2009 ) Basoglu Daim Atesok and Pamuk ( 2010 ) Behkami and Daim ( 2012 ) Blue and Tan ( 2010 ) Brender et al ( 2000 ) Daim et al ( 2010 ) Dossler et al (2010) Gagnon et al ( 2003 ) Greenshup ( 2012 ) Hyun et al ( 2009 ) Jha et al ( 2008 ) Kijsanayotin (2009) Lenz and Kuhn ( 2004 ) Lluch ( 2011 ) Sagiroglu et al ( 2006 ) Stowe and Harding ( 2010 ) Topacan ( 2009 ) Toussiant and Lodder ( 1998 ) Tung and Chang ( 2008 ) Vest ( 2010 )
Electronic health records
Bergman ( 2007 ) Blazona and Koncar ( 2007 ) De-Meyer Lundgren De Moor and Fiers ( 1998 ) Edwards et al ( 2008 ) Haas et al ( 2010 ) Hannan ( 1999 ) Helleso and Lorensen ( 2005 ) Holbrook Keshavjee Troyan Pray and Ford ( 2003 ) International Organization for Standardization ( 2005 ) Kierkegaard ( 2011 ) Scott et al ( 2007 ) Tange et al ( 1997 ) Ueckert et al ( 2003 ) Wen et al ( 2007 ) Wang et al ( 2010 ) Wright et al ( 2009 ) Yoshihara ( 1998 )
Electronic health record adoption
Bernstein Bruun-Rasmussen Vingtoft Andersen and Nohr ( 2005 ) Brown and Warmington ( 2002 ) Cho et al ( 2010 ) Collins and Wagner ( 2005 ) Dobbing ( 2001 ) Estebaranz and Castellano ( 2009 ) Gonzalez-Heydrich et al ( 2000 ) Iakovidis ( 1998 ) Jahanbakhsh et al ( 2011 ) Likourezos et al ( 2004 ) Ludwick and Doucette ( 2009 ) Natarajan et al ( 2010 ) Ovretveit et al ( 2007 ) Rose et al ( 2005 ) Ross et al ( 2010 ) Saitwal et al ( 2010 ) Tavakoli et al ( 2011 ) Vesely et al ( 2006 ) Yoon et al ( 2012 ) Yu Li and Gagnon ( 2009 )
OM Koumlk et al
201
Tabl
e 8
3 E
xpla
natio
n of
the
cons
truc
ts
Con
stru
ct
Exp
lana
tion
Age
A
ge o
f th
e us
er
Ent
ity ty
pe
The
org
aniz
atio
n th
at th
e pa
rtic
ipan
t is
empl
oyed
at (
eg
hos
pita
l cl
inic
fam
ily h
ealth
cen
ter
Goa
l O
rgan
izat
ionrsquo
s go
al f
or u
sing
the
elec
tron
ic h
ealth
rec
ord
syst
em s
uch
as fi
nanc
ial
med
ical
and
adm
inis
trat
ive
Flex
ibili
ty
Syst
emrsquos
abi
lity
of a
dapt
ing
the
inte
rfac
e an
d w
orkfl
ow
acc
ordi
ng to
use
r re
quir
emen
ts (
Pola
t 20
10)
Bog
azic
i U
nive
rsity
MIS
Dep
artm
ent M
aste
r T
hesi
s U
ser
inte
rfac
e A
ll us
er-f
acin
g gr
aphi
cal i
nter
face
incl
udin
g bu
ttons
men
us o
ptio
ns v
isua
lizat
ion
and
use
r-fr
iend
lines
s Se
curi
ty
The
arc
hite
ctur
e th
at k
eeps
the
reco
rds
from
una
utho
rize
d ac
cess
dat
a lo
ss a
nd d
ata
man
ipul
atio
n (B
lobe
l 20
06 )
Task
ndashtec
hnol
ogy
fi t
Info
rmat
ion
syst
em w
hich
hav
e a
fl exi
ble
wor
kfl o
w a
nd a
cle
ar g
raph
ical
inte
rfac
e ca
n ea
sily
ada
pt to
the
task
s of
an
indi
vidu
al (
Dis
haw
amp S
tron
g 1
999 )
In
tegr
atio
n ha
rdw
are
Syst
emrsquos
inte
grat
ion
capa
bilit
y w
ith m
edic
al d
evic
es s
uch
as u
ltras
ound
lab
equ
ipm
ent
etc
In
tegr
atio
n so
ftw
are
Syst
emrsquos
org
aniz
atio
n ca
pabi
lity
with
oth
er s
oftw
are
syst
ems
such
as
acco
untin
g n
atio
nal i
dent
ity d
atab
ase
an
d in
sura
nce
com
pani
es (
Med
ula
Mer
nis)
Thi
s fu
nctio
nalit
y pr
ovid
es d
ata
cons
iste
ncy
amon
g sy
stem
s an
d al
so s
ave
criti
cal t
ime
for
the
user
s D
ose
func
tiona
lity
(Fun
cDos
e)
Syst
emrsquos
fun
ctio
nalit
y of
kee
ping
dos
e in
form
atio
n re
gard
ing
the
patie
ntrsquos
med
icat
ion
Ran
ge f
unct
iona
lity
(Fun
cRan
ge)
Syst
emrsquos
fun
ctio
nalit
y of
kee
ping
min
imum
max
imum
val
ues
rega
rdin
g th
e te
st r
esul
ts b
lood
val
ues
etc
M
edic
al in
form
atio
n fu
nctio
nalit
y (F
uncX
Med
) Sy
stem
rsquos f
unct
iona
lity
of p
rovi
ding
req
uire
d ad
ditio
nal m
edic
al in
form
atio
n to
the
user
s in
the
case
of
nece
ssity
Acc
essA
LL
U
serrsquo
s ac
cess
to a
ll re
quir
ed in
form
atio
n in
pat
ient
rec
ords
A
ccur
acy
Syst
emrsquos
cap
abili
ty to
hav
e ac
cura
te a
nd s
ensi
tive
info
rmat
ion
(Hay
rine
n et
al
200
8)
Com
plet
enes
s Sy
stem
rsquos c
apab
ility
to h
ave
com
plet
e in
form
atio
n (O
vret
veit
et a
l 2
007 )
U
p-to
-dat
enes
s Sy
stem
rsquos c
apab
ility
to u
pdat
e in
form
atio
n re
gula
rly
(con
tinue
d)
8 Adoption Factors of Electronic Health Record Systems
202
Tabl
e 8
3 (c
ontin
ued)
Con
stru
ct
Exp
lana
tion
Stan
dard
izat
ion
Syst
emrsquo f
unct
iona
lity
to k
eep
info
rmat
ion
alig
ned
with
nat
iona
l and
inte
rnat
iona
l sta
ndar
ds (
Yos
hiha
ra 1
998 )
M
obili
ty
Syst
emrsquos
fun
ctio
nalit
y to
off
er u
ser
acce
ssib
ility
fro
m a
nyw
here
at a
ny ti
me
Sys
tem
rsquos d
egre
e to
the
user
rsquos e
ase
of a
cces
s to
the
info
rmat
ion
(Top
acan
200
9 )
Priv
acy
unau
thor
ized
acc
ess
(Pri
vacy
UA
) Sy
stem
rsquos f
unct
iona
lity
to p
reve
nt u
naut
hori
zed
acce
ss b
ut le
tting
aut
hori
zed
user
s to
acc
ess
requ
ired
info
rmat
ion
(Dob
bing
200
1 )
Med
ical
info
rmat
ion
shar
ing
(Pri
vacy
MD
) U
serrsquo
s at
titud
e to
pat
ient
info
rmat
ion
bein
g se
en b
y ot
her
care
take
rs
Kno
wle
dge
shar
ing
Use
rrsquos
attit
ude
to s
hare
med
ical
info
rmat
ion
with
co-
wor
kers
for
con
sulta
tion
(Uec
kert
et a
l 2
003 )
Su
ppor
t qua
lity
The
qua
lity
of th
e su
ppor
t pro
vide
d by
gui
delin
es s
yste
m h
elp
func
tiona
lity
ven
dor
team
and
co-
wor
kers
Se
lf-c
onfi d
ence
In
divi
dual
rsquos o
wn
skill
s ow
n co
mpu
ter
usag
e (T
anog
lu 2
006 )
E
ase
of le
arni
ng
Syst
emrsquos
rat
e on
how
eas
ily it
can
be
lear
ned
(Hol
broo
k et
al
200
3 )
Eas
e of
use
Sy
stem
rsquos r
ate
on h
ow it
can
be
used
with
leas
t eff
ort (
Dav
is 1
989 )
U
sefu
lnes
s Sy
stem
rsquos p
ositi
ve e
ffec
ts o
n th
e en
hanc
ing
indi
vidu
alrsquos
wor
k (D
avis
198
9 )
Atti
tude
In
divi
dual
rsquos p
ositi
ve o
r ne
gativ
e pe
rcep
tion
abou
t the
sys
tem
(Fi
shbe
in amp
Ajz
en 1
975 )
Q
ualit
y of
car
e R
ate
of th
e pr
oduc
tivity
in th
e he
alth
care
ser
vice
s in
clud
ing
num
ber
of s
ucce
ssfu
l tre
atm
ents
num
ber
of
succ
essf
ul d
iagn
osis
etc
(L
udw
ick
amp D
ouce
tte 2
009 )
E
ffi c
ient
use
R
ate
on h
ow th
e in
divi
dual
effi
cie
ntly
use
s th
e sy
stem
D
iffu
sion
R
ate
on h
ow th
e sy
stem
is s
prea
d w
ithin
the
orga
niza
tion
Infu
sion
R
ate
on h
ow th
e in
divi
dual
use
s th
e of
feri
ngs
of th
e sy
stem
U
se d
ensi
ty
Rat
e on
how
foc
used
the
indi
vidu
al u
sed
the
syst
em
Satis
fact
ion
Rat
e on
how
hap
py th
e in
divi
dual
is o
n us
ing
the
syst
em
OM Koumlk et al
203
Tabl
e 8
4 M
ajor
con
stru
cts
and
thei
r lit
erat
ure
Con
stru
ct
Ana
lyze
d lit
erat
ure
Age
Sh
abbi
r et
al
( 201
0 ) V
enka
tesh
et a
l ( 2
003 )
E
ntity
type
Ja
hanb
akhs
h et
al
( 201
1 ) H
elle
so a
nd L
oren
sen
( 200
5 ) S
agir
oglu
( 20
06 )
Iak
ovid
is (
1998
) Se
curi
ty
Uec
kert
et a
l ( 2
003 )
Dob
bing
( 20
01 )
Ovr
etve
it et
al
( 200
7 ) H
olbr
ook
et a
l ( 2
003 )
Haa
s et
al
( 201
0 )
Jaha
nbak
hsh
et a
l ( 2
011 )
Ta
skndasht
echn
olog
y fi t
N
atar
ajan
et a
l ( 2
010 )
Hol
broo
k et
al
( 200
3 ) C
ayir
( 20
10 )
Dis
haw
and
Str
ong
( 199
9 ) H
yun
et a
l ( 2
009 )
Sa
giro
glu
( 200
6 )
Satis
fact
ion
Hay
rine
n et
al
(200
8) D
eLon
e an
d M
cLea
n (1
992
200
3) L
ikou
rezo
s et
al
( 200
4 )
Eas
e of
use
D
avis
( 19
89 )
Ven
kate
sh e
t al
( 200
3 ) Y
u et
al
( 200
9 ) H
olbr
ook
et a
l ( 2
003 )
Sai
twal
et a
l ( 2
010 )
Top
acan
( 20
09 )
Use
fuln
ess
Yu
et a
l ( 2
009 )
Hol
broo
k et
al
( 200
3 ) S
habb
ir e
t al
( 201
0 ) D
avis
( 19
89 )
Ven
kate
sh e
t al
( 200
3 ) V
enka
tesh
and
D
avis
( 20
00 )
Top
acan
( 20
09 )
Atti
tude
Fi
shbe
in a
nd A
jzen
( 19
75 )
Dav
is (
1989
) V
enka
tesh
and
Dav
is (
2000
) T
opac
an (
2009
) E
ase
of le
arni
ng
Hol
broo
k et
al
( 200
3 ) H
ayri
nen
et a
l (2
008)
DeL
one
and
McL
ean
(200
3)
Info
H
ayri
nen
et a
l (2
008)
Yos
hiha
ra (
1998
) O
vret
veit
et a
l ( 2
007 )
Cay
ir (
2010
) B
asog
lu e
t al
(200
9)
Jaha
nbak
hsh
et a
l ( 2
011 )
Wan
g et
al
( 201
0 )
Qua
lity
of c
are
Lud
wic
k an
d D
ouce
tte (
2009
) H
ayri
nen
et a
l (2
008)
Col
lins
and
Wag
ner
( 200
5 ) B
row
n an
d W
arm
ingt
on (
2002
)
Cho
et a
l ( 2
010 )
Tan
ge e
t al
( 199
7 ) D
ossl
er e
t al
(201
0)
Self
-con
fi den
ce
Tano
glu
( 200
6 ) D
avis
( 19
89 )
Yu
et a
l ( 2
009 )
Agg
elid
is a
nd C
hatz
oglo
u ( 2
009 )
Tun
g an
d C
hang
( 20
08 )
Priv
acy
Dob
bing
( 20
01 )
Lud
wic
k an
d D
ouce
tte (
2009
) H
aas
et a
l ( 2
010 )
Saf
ran
and
Gol
derb
erg
(200
0) B
lobe
l ( 20
06 )
Use
r in
terf
ace
Saitw
al e
t al
( 201
0 ) W
ang
et a
l ( 2
010 )
Dob
bing
( 20
01 )
Pol
at (
2010
) B
row
n an
d W
arm
ingt
on (
2002
)
8 Adoption Factors of Electronic Health Record Systems
204
Relationship among usefulness ease of use and attitude is explained in the TAM (Davis 1989 ) and TAM2 (Venkatesh amp Davis 2000 )
H9 Privacy function of the system which avoids unauthorized access to confi den-tial patient data positively affects the attitude
H10 Caretakerrsquos attitude toward information sharing with hisher co-workers has in impact on attitude toward system use
H11 The systemrsquos ease of learning has an impact on attitude toward system use
Holbrook et al stated that provided support on the system and ease of learning of the system have an impact on the implementation of EHR systems ( 2003 )
H12 Ease of use positively affects the satisfaction H13 Usefulness positively impacts the satisfaction H14 Electronic health record systemrsquos integration with medical equipment posi-
tively affects the satisfaction H15 Usefulness signifi cantly and positively impacts use density of the system H16 Attitude toward use signifi cantly impacts the use density of the system
(Table 85 )
In the second aspect the relationship between external factors and intermediary constructs will be analyzed
H1 Ease of use positively affects usefulness H2 Information quality positively and signifi cantly impacts usefulness H3 Flexibility of the system positively affects usefulness H4 Mobility of the system positively affects usefulness H5 Self-confi dence of the user positively affects usefulness
Table 85 Hypothesis list for dependent items
Hypotheses Dependent Independent Relationship
H1 Quality of care Usefulness Positive H2 Quality of care Attitude Positive H3 Diffusion Usefulness Positive H4 Diffusion Attitude Positive H5 Infusion Usefulness Positive H6 Infusion EoU Positive H7 Attitude Usefulness Positive H8 Attitude EoU Positive H9 Attitude PrivacyUA Positive H10 Attitude PrivacyMD Positive H11 Attitude EoL Positive H12 Satisfaction EoU Positive H13 Satisfaction Usefulness Positive H14 Satisfaction IntegrationHW Positive H15 Use density Usefulness Positive H16 Use density Attitude Positive
OM Koumlk et al
205
H6 Ease of learning of the system signifi cantly and positively affects usefulness H7 User interface signifi cantly and positively affects usefulness H8 The systemrsquos functionality related to keeping dose information of the medica-
tion positively affects usefulness H9 The systemrsquos ease of learning positively impacts the systemrsquos ease of use H10 User interface of the system positively and signifi cantly impacts the ease of
use of the system H11 Mobility of the system positively and signifi cantly affects the systemrsquos ease of
use H12 Information quality signifi cantly affects the ease of use H13 Privacy measure for avoiding unauthorized access negatively affects the ease
of use (Table 86 )
In the third model factors affecting userrsquos effi cient use of the system will be analyzed
H1 Taskndashtechnology fi t of the system signifi cantly and positively affects the effi -cient use
H2 User interface signifi cantly and positively impacts the effi cient use of the systems H3 Userrsquos ability to access all required information positively affects the effi cient
use of the system H4 The systemrsquos functionality of offering basic medical information signifi cantly
and positively impacts the effi cient use of the system H5 Information quality in the system positively impacts the effi cient use of the
systems H6 The systemrsquos integration with other software signifi cantly and positively
affects the effi cient use of the system H7 The systemrsquos functionality related to keeping dose information of the medica-
tion positively affects the effi cient use of the system (Table 87 )
Table 86 Hypothesis list for intermediary constructs
Hypotheses Dependent Independent Relationship
H1 Usefulness EoU Positive H2 Usefulness Info Positive H3 Usefulness Flexibility Positive H4 Usefulness Mobility Positive H5 Usefulness Self confi dence Positive H6 Usefulness Ease of learning Positive H7 Usefulness User interface Positive H8 Usefulness FuncDose Positive H9 EoU EoL Positive H10 EoU User interface Positive H11 EoU Mobility Positive H12 EoU Info Positive H13 EoU Privacy Negative
8 Adoption Factors of Electronic Health Record Systems
206
84 Methodology
This research study has started in September 2010 From that time many inter-views surveys literature research and observations have been conducted to deeply understand the topic and to develop hypotheses
Firstly literature research has been done between September 2010 and July 2011 Literature related to electronic health records health information systems technology adoption models and health technology adoption has been analyzed and main constructs and variables have been extracted
Furthermore to combine the literature information between September 2010 and December 2010 semi-structured interviews have been conducted with healthcare employees who use electronic health record systems Results of the literature research and semi-structured interviews have been consolidated and published in the PICMET 2011 Conference (Kok Basoglu amp Daim 2011 ) Also these studies have helped us to develop hypotheses
In the second phase of the study we have conducted a focus group study with information systems and medical experts A construct list has been provided to them to select their top preferences
In the third phase a pilot survey has been conducted with 15 participants to check the reliability of the items in the survey
In the fourth phase in order to test our hypotheses quantitative fi eld survey study has been completed with 301 participants (Table 88 )
841 Qualitative Study
Semi-structured face-to-face interviews were conducted to widen electronic health record adoption taxonomy Literature review fi ndings were aimed to be corrected and new fi ndings were expected
Interviewees were doctors who were selected from different hospitals and dif-ferent specialties Questions were prepared in a Word document which have included both factors gained from literature review and questions to discover factors which were not faced yet
Table 87 Hypothesis list for effi cient use
Hypotheses Dependent Independent Relationship
H1 Effi cient use TTF Positive H2 Effi cient use User interface Positive H3 Effi cient use AccessALL Positive H4 Effi cient use FuncXMed Positive H5 Effi cient use Info Positive H6 Effi cient use Integration SW Positive H7 Effi cient use FuncDose Positive
OM Koumlk et al
207
We targeted the doctors as our interview group as they are the main users of EHR systems However there are other users of the systems such as administrations nurses medical assistants etc These groups were not included in the face-to-face interviews
Eight interviews were conducted and the factors have been analyzed with their existence ratio rate of the factorrsquos occurrence in total of the interviews
Questions list can be found in Appendix 1
842 Expert Focus Group Study
After the defi nition of constructs an expert focus group has been conducted in order to prioritize the constructs Figure 810 implies the expert focus group study example
A focus group has been performed with eight experts Participants were experi-enced medical doctors and software development engineers The expert focus group questionnaire was based on Excel which has been sent to the experts and can be found in Appendix 2 Studied constructs are listed in Table 89
843 Pilot Study
Before the quantitative fi eld survey study two pilot studies were conducted to improve the fi eld survey studyrsquos quality and accuracy
The fi rst pilot study was conducted with three people with a survey of 65 ques-tions Participants have completed the survey with us and shared their comments regarding the quality or wording of the questions that we have prepared Also one of the participants requested a question to be added
Table 88 Steps of the study
Step Date Explanation
Semi-structured interviews
September 2010 Interviews were conducted with eight participants from our main target group doctors Results of the study have been published in PICMET-2011 conference
Expert focus group study
August 2011 A focus group study has been conducted with eight participants including doctors and software developers Participants were asked to choose 20 most important constructs from the construct list that we have provided
Pilot study January 2012 In order to test the research instrument a pilot study has been conducted with 15 participants Sixty-fi ve questions survey has been conducted with participants Then reliability analysis and factor analysis have been conducted
Quantitative fi eld survey study
February 2012 Quantitative fi eld survey study has been conducted with 301 participants Reliability analysis factor analysis regression modeling ANOVA analysis and clustering have been done with the results
8 Adoption Factors of Electronic Health Record Systems
208
The second pilot study was shared via a web survey system Fifteen people have participated in the second pilot study Results of the pilot study have been used as an input for the reliability and factor analysis test in the Statistical Package for Social Sciences (SPSS)
844 Quantitative Field Survey
After the pilot study the survey has been prepared in a web-based tool and shared via e-mail through different channels Initially three hospitals were targeted Then with efforts of the Manisa City Health Department the survey is shared with the
Fig 810 Expert focus group construct list
Table 89 Constructs studied in focus group
Accessibility Guidelines Quality of support
Accuracy Habit Successful treatment Adequate resources Hospital size Successful decision Age Image Successful diagnosis Behavioral control Income Response time Clinical specialty Information quality Risk Compatibility Job experience Satisfaction Computer experience Job relevance Security Computer literacy Managerial support Social infl uence Ease of learning Marital status Standardization Ease of use Medical Taskndashtechnology fi t Educational level Occupation Tool experience Facilitating conditions Other clinical variables Trust Flexibility Peer support Usefulness Functionality characteristics Place of residence User interface Gender Population serviced Vendor support Geographic area Professional support Voluntariness
OM Koumlk et al
209
family practitioners of the city of Manisa They have shown great participation and the quantitative fi eld survey study has been applied to 301 people in total Mostly the participants were family health practitioners in the city of Manisa
85 Findings
851 Qualitative Study Findings
Semi-structured face-to-face interviews have been conducted with eight participants
bull 375 + of the participants were females bull 50 of the employees had more than 15 years of work experience bull Only one participant had his own clinic the remaining ones were working at a
hospital bull Average age of the interviewees was 41
General characteristics of the interviewees can be found in Table 810 Constructs which two or more interviewees have implied are listed in Table 811
with their frequency and frequency rate during the interviews (in total eight interviews)
Several important factors have been defi ned via combination of literature review and qualitative research
8511 Sharing and Privacy
Easy sharing is the one of the other important factors It is implied that unlike the paper records medical records can be shared easier and faster without making phys-ical transaction such as photocopying (Safran amp Golderberg 2000 )
Also interviewers told that sometimes they are exchanging information about patients with their colleagues Moreover interviewers working in government
Table 810 Profi le of the interviewees
Specialty Age Organization Gender Experience
Brain surgeon 49 Hospital A Male 20+ Internist 50 Hospital B Male 20+ Pediatrician 46 Own clinic Male 20+ Earndashnosendashthroat 32 Hospital A Male 6 Earndashnosendashthroat 36 Hospital C Male 10 Pediatrician 38 Hospital C Female 12 Dermatologist 35 Hospital C Female 11 Pediatrician 40 Hospital C Female 15
8 Adoption Factors of Electronic Health Record Systems
210
hospitals explained that some of the government hospitals have been using a com-mon system and they can easily share fi les through them This also brings out that systems can be used for consultation and some EHR system can be developed with this functionality This can also be related with the doctorrsquos title and work experience One of the interviewers stated that
For some specifi c cases I request consultation over the system from more experienced doc-tors Even for some cases I share the fi le over the system with other departments to consult their opinion (Brain Surgeon 49)
Moreover it stated that many organizations started to look for exchanging healthcare data and patient data faster through networks as a result of the development in commu-nications technologies (Ueckert Maximilian Goerz Tessmann amp Prokosch 2003 )
So easier and accurate sharing is an important adoption factor of EHR systems It brings more fl exibility than paper-based records
8512 User Interface
User interface highly affects the usage of EHR systems It defi nes the mental opera-tions needed to be done and also the physical steps to take for completing a task (Saitwal Xuan Walji Patel amp Zhang 2010 )
In the in-depth interview we made we gained the feedback that most of the users have complaints about the UIs of the EHR systems Some of the doctors stated that they have diffi culties to compare the results of the tests that they requested with their pre-diagnoses and the patient complaints Because all of these are kept in different places in the system and from one UI they canrsquot view them all
Also one of the interviewers has stated that for some tasks she needs to deal with many steps
For some simple tasks even I need to go to 2ndash3 different UIs and have to click a few buttons (Female 35)
User interface affects the ease of use positively
Table 811 Frequency of the constructs
Construct Frequency Frequency rate ()
User interface 8 100 Archiving 7 88 Quality of care 6 75 Sharing 4 50 Data preservation 4 50 Search criteria 4 50 Accuracy 3 38 Time saving 2 25 Medical assistant 2 25 Standardization 2 25 Search ability 2 25
OM Koumlk et al
211
8513 Perceived Ease of Use
Davis defi ned the perceived ease of use as ldquothe degree to which a person believes that using a particular system would be free of effortrdquo ( 1989 )
8514 Perceived Usefulness
Perceived usefulness is defi ned as ldquoextent to which a person believes that using the system will enhance his or her job performancerdquo (Davis 1989 )
It is modeled that if users believe that a system has high usefulness users will gain high performance when the system is used (Davis 1989 )
8515 Information Quality
Use of EHR brings standardization of the medical terms in the use of medical records Even though standardization of the terms may cause problems in the begin-ning of the adoption process such as requiring assistance to enter standardized names in the long term users will start to use it more effi ciently Also for effective statistics standardized records are the main base asset (Yoshihara 1998 )
One of the interviewers stated that
Electronic health records provide us to the chance to compare them with other patients and to be able to get statistics The data that I get is more qualifi ed (Male Internist 50)
Also standardization of the procedures might have a positive impact on the qual-ity of the processes (Nowinski et al 2007 ) Usage of EMR has distinctive changes on the way that physicians keep their records (Bergman 2007 ) From this stand-point we can say that getting easier statistics with standardized information is one of the important adoption factors of electronic health records We can assume that it has positive interaction with the perceived usefulness
8516 Quality of Care
Most of our interviewees have stated that EHR usage has many effects on the qual-ity of care provided EHR lets the user see the medical history of the patient consis-tently Physicians have access to see the past injuries of the patient and the treatments that have been applied to himher
If physicians do not have the enough information about the medical history of the patient they would not be able to give the right decisions The patient care process also includes the process of getting data turning it to information and then using it in the decision-making (Collins amp Wagner 2005 ) Keeping accurate and correct information is important otherwise with wrong data wrong clinician actions can be taken on the patients (Brown amp Warmington 2002 ) It has been proven in many studies that EHR has a positive effect on the quality of care
8 Adoption Factors of Electronic Health Record Systems
212
To be able to offer better healthcare diagnostics and treatments healthcare pro-viders should have good information about the patientrsquos situation Nowadays EHR is upcoming as the most preferred way to keep up with patient data (Haas Wohlgemuth Echizen Sonehara amp Muumlller 2010 ) Also some studies have shown that with EHR input to decision support systems for some specifi c cases like chronic illnesses quality of care has signifi cantly increased (Cho et al 2010 )
So we can assume that quality of care is an important factor on the usage of the EHR system Quality of care affects the usefulness of the systems positively
8517 Job Relevance TaskndashTechnology Fit (TTF)
As gathered from both interviews and literature EHR usage reduces the time spent in the healthcare Input time does not really decrease with the EHR usage but time spent for gathering the information and viewing the patientrsquos medical history occurs much faster (Dobbing 2001 ) Also it is stated that sometimes data entry takes a little more time than the data entry on paper-based records (Shabbir et al 2010 ) The more customized the workfl ows of the system can be the faster the user can adapt to the system (Dishaw amp Strong 1999 )
Our interviewees did not really give specifi c responses about the time that they saved during the data entry However they specifi ed that EHR usage really reduces the time spent during the search of the records and also they spend less time when they want to look for some specifi c information
8518 Functionality
Interviewees had a general opinion about EHR having many advantages with search abilities than paper-based records Users can easily and quickly search health records over the system In the old-fashioned way doctors needed to search the fi les manually between folders However our interviewees have stated that the EHR sys-tem is not fully functional about search now
If my patients have two names itrsquos hard to fi nd and identify them I need another criteria to be able to search (Earndashnosendashthroat 32)
Also another interviewee stated that
I can search with the name or identity number of the patient It could be more useful if I have some other criteria (Pediatrician 38)
With the increasing data in the EHR systems search abilities will play a very critical role to fi nd the accurate and required information (Natarajan Stein Jain amp Elhadad 2010 )
We can say that search abilities are an important factor in adoption of EHR As the search abilities are developed more it would have more effect on the use of
OM Koumlk et al
213
EHR EHR systems can offer different functionalities such as integration with other required software (IntegrationSW) integration with medical devices eg ultra-sound (IntegrationHW) keeping limit dosage values for medicines (FuncDose) containing basic health and diagnosis information to assist healthcare responsible (FuncXMed) and critical ranges for lab results (FuncRange)
8519 Archiving and Data Preservation
Medical records are essential for healthcare Thus archiving plays a critical role
With EHR system we gained a better archiving We are the master of the data now 10ndash15 years ago I was giving my patients the reports lab results and etc about them They needed to archive them in their house by themselves However mostly they were not able to keep the records They generally lost them and for next appointments they came to me without any records So this was limiting my knowledge about the patientsrsquo background and the treatments have been applied Now I keep all the records in my computer and the data is preserved (Neurobiologist 49)
One of the interviewees stated that
Papers can always get lost even if they are stored by me or the patient itself Archiving the records in computers are more reliable (Pediatrician 40)
Paper-based records bring high costs to save keep and then use again Sometimes they are transferred to different departments and sometimes they are not returned thus the data get lost (Safran amp Golderberg 2001)
Keeping the medical data is very important also for healthcare At least the health information which can be used as input for clinical decision-making should be kept and archived in systems (Estebaranz amp Castellano 2009 ) EHR history should be recorded with its updates and also should be aimed to be kept long term as required (Toyoda 1998)
85110 Medical Assistant
We found another specifi c item which is the medical assistant Medical assistants are the clerks in the hospital who are occupied for up to 2ndash3 doctors They handle the offi ce work of the doctors Some doctors stated that they let their medical assistants keep their medical records
852 Expert Focus Group Findings
Constructs gained from literature review and qualitative study have been com-piled in Excel Then the Excel file has been sent to the expert via e-mail Experts were asked to determine the 20 most favorable constructs out of 51
8 Adoption Factors of Electronic Health Record Systems
214
The list had the Turkish meaning English meaning and explanation of the construct
bull 125 of the participants were female bull 50 of the participants had work experience over 20 years bull Half of the participants were software experts and the other half were medical
experts (Table 812 )
Participants had consistent responses Age and ease of use constructs were selected by all participants Satisfaction compatibility usefulness and accuracy were the other signifi cant constructs
These results have been analyzed by us and the responses are used as an input to the pilot and quantitative fi eld survey studies
Detailed results can be viewed in Table 813 The selection of constructs has been done and items for the pilot study have been
chosen
853 Pilot Study Findings
8531 Participant Characteristics
Fifteen participants were involved in pilot study
bull 733 of the participants were aged between 18 and 25 bull 50 of the participants had at least a university degree bull 733 of the participants were from government hospitals
Characteristics of the pilot study participants can be viewed in Table 814
8532 Reliability and Factor Analysis
After conducting reliability analysis and factor analysis redundant items were elim-inated Table 815 shows the constructs and their related items for the quantitative fi eld survey study
Table 812 Characteristics of participants
Specialty Age Organization Gender Experience
Brain surgeon 40+ Hospital A Male 20+ Brain surgeon 40+ Hospital B Male 20+ Brain surgeon 40+ Hospital B Female 20+ Doctor 40+ Hospital C Male 20+ Software project manager 30+ Organization A Male 15+ Software architect 30 Organization B Male 10+ Software designer 20+ Organization C Male 5+ Software expert 20+ Organization D Male 5+
OM Koumlk et al
215
Tabl
e 8
13
Exp
ert f
ocus
stu
dy r
esul
ts
Con
cept
Fr
eque
ncy
Con
cept
Fr
eque
ncy
Con
cept
Fr
eque
ncy
Age
8
Occ
upat
ion
4 G
uide
lines
3
Eas
e of
use
8
Eas
e of
lear
ning
4
Com
pute
r lit
erac
y 3
Com
patib
ility
7
Geo
grap
hic
area
4
Gen
der
2 Sa
tisfa
ctio
n 7
Popu
latio
n se
rvic
ed
4 C
linic
al s
peci
alty
2
Use
fuln
ess
6 H
ospi
tal s
ize
4 Pr
ofes
sion
al s
uppo
rt
2 A
ccur
acy
6 V
endo
r su
ppor
t 4
Tool
exp
erie
nce
2 Se
curi
ty
6 So
cial
infl u
ence
4
Rat
e of
suc
cess
ful t
reat
men
ts
2 Q
ualit
y of
sup
port
5
Func
tiona
l cha
ract
eris
tics
4 H
abit
2 St
anda
rdiz
atio
n 5
Acc
essi
bilit
y 4
Tru
st
2 In
form
atio
n qu
ality
5
Rat
e of
suc
cess
ful d
iagn
osis
4
Mar
ital s
tatu
s 1
Secu
rity
5
Edu
catio
nal l
evel
3
Job
expe
rien
ce
1 Fa
cilit
atin
g co
nditi
ons
5 Ta
skndasht
echn
olog
y fi t
3
Ade
quat
e re
sour
ces
1 Jo
b re
leva
nce
5 R
isk
3 Fl
exib
ility
1
Rat
e of
dec
isio
n ef
fi cie
ncy
5 Pl
ace
of r
esid
ence
3
Beh
avio
ral c
ontr
ol
1 R
espo
nse
time
5 M
anag
eria
l sup
port
3
Com
pute
r ex
peri
ence
1
Use
r in
terf
ace
5 V
olun
tari
ness
3
8 Adoption Factors of Electronic Health Record Systems
216
Table 815 Reliability analysis of pilot study
Construct c Alpha Items before deletion Items after deletion
User interface 0736 8 8 Usefulness 0773 7 7 Info 0613 5 5 EoU 0429 4 4 Satisfaction 0851 4 2 Flexibility 0694 3 3 Sharing 0328 3 0 TTF 0596 3 3 Mobility 0474 3 3 Quality of care 0714 3 3 Security 0254 2 2 Support quality 0691 2 2 Attitude toward use 0851 2 2
Table 814 Participant characteristics of pilot study
Item Range Frequency Percentage
Age 18ndash25 11 733 26ndash35 1 67 35ndash45 0 00 45ndash55 3 200 55+ 0 00
Education High school 7 500 University 5 357 Masters 0 00 PhD 2 143
Goal Medical 13 867 Management 2 133 Financial 0 00
Entity type Family treatment 0 00 Government 4 277 Private hospital 11 733
Reliability analysis has been conducted between the constructs Generally reli-ability results were over 0600 and items were considerably reliable However con-structs such as mobility security and sharing had lower reliabilities The main reason for this situation is related to the low number of observations and low num-ber of items in the test These results have been ignored and constructs have been kept same Detailed results of the reliability analysis can be seen in Table 815
OM Koumlk et al
217
Table 816 Profi le of the respondents
Item Range Frequency Percentage
Age 18ndash25 23 76 26ndash35 24 80 35ndash45 130 432 45ndash55 110 365 55+ 14 42
Education High school 23 77 University 189 632 Masters 45 151 PhD 42 140
Goal Medical 257 854 Management 39 132 Financial 5 17
Entity type Family treatment center 251 839 Government 4 13 Private hospital 44 147
Seven components have been extracted with the factor analysis for all items Detailed results for factor analysis of the pilot study can be found in Appendix 3 Factor analysis results have also supported our hypotheses
854 Quantitative Field Survey Study Findings
A study aimed to explore and understand factors affecting the adoption of electronic health record systems A web-based data collection tool has been used to gather data via questionnaire from healthcare employees from different organizations with different purposes
8541 Profi le of the Respondents
Most of the respondents were university graduates (432 ) and majority of the respondents were in the age between 36 and 45 (632 ) Systems in the respon-dentrsquos work locations were mainly used for medical purposes Doctors employed in the family treatment centers constituted the majority of the respondents with 854 (Tables 816 ndash 818 )
8 Adoption Factors of Electronic Health Record Systems
218
Table 818 Respondent profi le by entity goal and centrality
Entity type Central
Goal
Medical Admin Finance Total
Family HC No 1 1 Government Yes 215 32 3 250 Private Yes 2 2 4 Blank No 3 3 Family HC Yes 34 5 2 41 Government Yes 2 2 Total 257 39 5 301
Table 817 Respondent profi le by entity and education
Entity type Education High S Uni Masters PhD Blank Total
Family HC 6 176 40 28 1 251 Government 1 2 1 4 Private 15 11 4 13 1 44 Blank 1 1 2 Total 23 189 45 42 2 301
8542 Reliability and Factor Analysis
Responses from the survey have been evaluated with reliability analysis and factor analysis Validity of the constructs and reliability of the items have been investi-gated with these studies For multi-item constructs lowest c alpha value was calcu-lated as 0676 In general c alpha values were over 0800 which show that the consistencies of the items were relatively signifi cant However constructs such as support quality and fl exibility have lower consistencies compared to the others (Table 819 )
Factor analysis has been conducted on all constructs Ten main components have been extracted For intermediary construct group one component was extracted with 70 variance For dependent construct group one component was iterated with a variance of 67 Finally for external constructs four components have been devel-oped with a 57 variance Detailed factor analysis results can be seen in Appendix 3
8543 Descriptives
Descriptive statistics show us that participants do not have a certain decision about information sharing with our colleagues In average they all fi nd the electronic health records software easy to learn easy to use and useful They generally have a positive attitude to the electronic health record software usage They are mostly satisfi ed with the software and they believe that they are effi ciently using the soft-ware Descriptive results of the summated constructs can be found in Table 820
OM Koumlk et al
219
Table 819 Reliability analysis results
Construct of items c Alpha
Satisfaction 3 0943 Info 5 0915 Usefulness 7 0914 Attitude 2 0905 TTF 3 0863 EoU 4 0854 Security 2 0826 QualityofCare 3 0819 Mobility 3 0804 User interface 8 0770 Flexibility 3 0696 SupportQuality 2 0676
Table 820 Descriptive statistics for all constructs
Construct Mean Median Mode Min Max SD
IntegrationHW 174 1 1 1 5 155 IntegrationSW 055 1 1 0 1 050 FuncDose 057 1 1 0 1 050 FuncRange 050 1 1 0 1 050 FuncXMed 051 1 1 0 1 050 AccessALL 082 1 1 0 1 039 PrivacyUA 345 4 4 1 5 117 PrivacyMD 354 4 4 1 5 118 KnowledgeShare 323 3 4 1 5 122 SelfConfi dence 401 4 4 1 5 096 EoL 379 4 4 1 5 106 Effi cientUse 761 8 8 1 10 181 Diffusion 389 4 4 1 5 090 Infusion 371 4 4 1 5 103 UseDensity 405 4 4 1 5 089 Attitude 407 4 4 1 5 074 Security 379 4 4 1 5 092 SupportQuality 340 350 4 1 5 098 EoU 403 4 4 1 5 073 Flexibility 367 360 4 1 5 086 Mobility 368 4 4 1 5 095 QualityofCare 361 360 4 1 5 084 Satisfaction 395 4 4 1 5 090 TTF 389 4 4 1 5 088 Info 385 4 4 1 5 078 Usefulness 390 4 4 1 5 073 UserInterface 369 370 370 110 5 062
8 Adoption Factors of Electronic Health Record Systems
220
8544 Regression Model Results
Obtained data has been analyzed using the IBM SPSS v20 software Linear regres-sion modeling has been chosen as the applied methodology Results of the executed regression model for dependent items are listed in Tables 821 and 822
Based on the regression results two models have been developed One shows the relationship between the external factors intermediary factors and dependent fac-tors The second model shows the relationship between the external factors and effi cient use First model is implied in Fig 811 and second model is implied in Fig 812 (Table 823 )
Regression results show that usefulness and attitude are direct determinants of quality of care with coeffi cients 055 ( p lt 0001) and 024 ( p lt 0001) Usefulness ( p lt 0001) and attitude ( p lt 001) explains 0568 of the diffusion respectively On the other hand infusion is dependent on usefulness ( p lt 0001) and EoU ( p lt 0010) Our hypothesis that attitude is dependent on PrivacyUA PrivacyMD and EoL was not supported in the regression analysis However results showed that 0710 of attitude is dependent on usefulness with a coeffi cient of 068 ( p lt 0001) and on EoU with a coeffi cient of 020 ( p lt 0001) The relationship between attitude EoU and usefulness was also supported in Davisrsquos TAM model (Davis 1989 ) Although EoU ( p lt 0001) and usefulness ( p lt 0001) explain the 0710 of satisfaction analysis did not imply that hardware integration (IntegrationHW) affects satisfaction Usefulness ( p lt 0001) and attitude ( p lt 0100) explain the 0417 of use density (Table 824 )
Information quality ( b 030 p lt 0001) ease of use ( b 020 p lt 0010) fl exibility of the software ( b 014 p lt 0010) mobility of the software ( b 014 p lt 0010) self- confi dence of the individual( b 011 p lt 0010) user interface of the software ( b 015 p lt 0100) and dose functionality of the software ( b 007 p lt 0100) explain the 0752 of usefulness factor Results also show similarities with other models An unsupported hypothesis was that privacy negatively affects ease of use and ease of learning affects usefulness (Table 825 )
Effi cient use of the system is explained mainly with taskndashtechnology fi t ( b 027 p lt 0001) and user interface ( b 028 p lt 0001) is then affected with AccessALL ( b 014 p lt 0002) medical information functionality of the software ( b 009 p lt 0100) information quality ( b 017 p lt 0010) integration of the system with other software ( b 011 p lt 0100) and dose functionality of the system ( b 009 p lt 0100)
8545 ANOVA Results
ANOVA analysis has been conducted on demographic values including age entity goal and education
Signifi cant results for ANOVA analysis based on age construct can be found in Table 826 Participants are grouped under fi ve different age categories 18ndash25 26ndash35 36ndash45 46ndash55 and 55+ It can be seen that participants in the age of 55+ are more satisfi ed with their EHR system and use the system more densely People in
OM Koumlk et al
221
Tabl
e 8
21
Reg
ress
ion
resu
lts f
or d
epen
dent
fac
tors
Dep
ende
nt
Inde
pend
ent
Coe
ffi c
ient
bet
a St
anda
rdiz
ed c
oeffi
cie
nt
Sign
ifi ca
nce
R 2
Adj
uste
d R
2
Qua
lity
of c
are
(Con
stan
t)
005
0
786
057
8 0
575
Use
fuln
ess
063
0
55
000
0 A
ttitu
de
027
0
24
000
0 E
ffi c
ient
use
(C
onst
ant)
minus
020
0
697
054
2 0
529
TT
F 0
57
027
0
000
Use
rInt
erfa
ce
079
0
28
000
0 A
cces
sAL
L
068
0
14
000
2 Fu
ncX
Med
0
33
009
0
049
Info
0
39
017
0
009
Inte
grat
ionS
W
039
0
11
001
8 Fu
ncD
ose
034
0
09
004
4 D
iffu
sion
(C
onst
ant)
0
10
061
1 0
572
056
9 U
sefu
lnes
s 0
67
054
0
000
Atti
tude
0
29
024
0
001
Infu
sion
(C
onst
ant)
minus
024
0
346
046
4 0
460
Use
fuln
ess
069
0
49
000
0 E
oU
031
0
22
000
1 U
se d
ensi
ty
(Con
stan
t)
085
0
000
042
1 0
417
Use
fuln
ess
062
0
51
000
0 A
ttitu
de
019
0
16
004
4 Sa
tisfa
ctio
n (C
onst
ant)
minus
043
0
009
071
2 0
710
EoU
0
56
045
0
000
Use
fuln
ess
054
0
44
000
0
8 Adoption Factors of Electronic Health Record Systems
222
Table 822 Regression results for intermediary factors
Dependent Independent Coeffi cient beta
Standardized coeffi cient Signifi cance R 2 Adjusted R 2
Attitude (Constant) 056 0000 0712 0710 Usefulness 069 068 0000 EoU 020 020 0000
Usefulness (Constant) 011 0464 0759 0752 Info 028 030 0000 EoU 019 020 0002 Flexibility 012 014 0002 Mobility 011 014 0003 SelfConfi dence 009 011 0006 UserInterface 017 015 0010 FuncDose 011 007 0027
EoU (Constant) 017 0238 0775 0771 UserInterface 046 038 0000 Info 025 027 0000 EoL 019 024 0000 Mobility 013 017 0000
020
044
045
051
054
055
Diffusion
Infusion
Attitude
Use Density
Satisfaction
Quality of Care
EoU
Usefulness
Use Density
EoL
Self Confidence
Func Dose
User Int
Mobility
Info
Flexibility
p lt 0100 p lt 0010 p lt 0001
Fig 811 Factors affecting the EHR adoption
OM Koumlk et al
223
027 Efficient Use
FuncXMed
Func Dose
User Interface
TTF
Access All
Info
IntegrationSW
p lt 0100 p lt 0010 p lt 0001
Fig 812 Factors affecting the effi cient use of EHR
Table 823 Results for dependent items
Hypotheses Dependent Independent Supported Signifi cance
H1 Quality of care Usefulness Yes 0000 H2 Quality of care Attitude Yes 0000 H3 Diffusion Usefulness Yes 0000 H4 Diffusion Attitude Yes 0001 H6 Infusion Usefulness Yes 0000 H7 Infusion EoU Yes 0001 H8 Attitude Usefulness Yes 0000 H9 Attitude EoU Yes 0000 H10 Attitude PrivacyUA No ndash H11 Attitude PrivacyMD No ndash H12 Attitude EoL No ndash H13 Satisfaction EoU Yes 0000 H14 Satisfaction Usefulness Yes 0000 H15 Satisfaction IntegrationHW No ndash H16 Use density Usefulness Yes 0000 H17 Use density Attitude Yes 0044
8 Adoption Factors of Electronic Health Record Systems
Table 824 Results of intermediary items
Hypotheses Dependent Independent Supported Signifi cance
H1 Usefulness EoU Yes 0000 H2 Usefulness Info Yes 0002 H3 Usefulness Flexibility Yes 0002 H4 Usefulness Mobility Yes 0003 H5 Usefulness Self-confi dence Yes 0006 H6 Usefulness Ease of learning No ndash H7 Usefulness User interface Yes 0010 H8 Usefulness FuncDose Yes 0027 H9 EoU EoL Yes 0000 H10 EoU User interface Yes 0000 H11 EoU Mobility Yes 0000 H12 EoU Info Yes 0000 H13 EoU PrivacyUA No ndash
Table 825 Results of effi cient use
Hypotheses Dependent Independent Supported Signifi cance
H1 Effi cient use TTF Yes 0000 H2 Effi cient use User interface Yes 0000 H3 Effi cient use AccessALL Yes 0002 H4 Effi cient use FuncXMed Yes 0049 H5 Effi cient use Info Yes 0009 H6 Effi cient use IntegrationSW Yes 0018 H7 Effi cient use FuncDose Yes 0044
Table 826 ANOVA results for age
Construct F Sig 18ndash25 26ndash35 36ndash45 46ndash55 55+
23 24 130 110 14 IntegrationHW 1561 0000 386 252 153 151 100 Satisfaction 720 0000 307 381 410 397 417 SelfConfi dence 676 0000 313 413 411 410 357 UserInterface 590 0000 313 360 378 374 366 EoL 567 0000 291 358 396 385 350 UseDensity 541 0000 330 383 415 410 429 SupportQuality 520 0000 265 313 353 342 379 TTF 484 0001 317 382 401 388 410 Info 438 0002 326 375 394 387 410 Flexibility 438 0002 306 338 373 377 376 Mobility 423 0002 312 336 382 364 407 Attitude 408 0003 352 410 417 404 421 Usefulness 381 0005 336 389 398 389 403 EoU 363 0007 350 397 410 407 414 IntegrationSW 333 0011 057 077 060 042 062 PrivacyMD 324 0013 296 346 377 346 314 Diffusion 293 0021 335 392 402 385 386 FuncRange 283 0025 065 058 040 059 043
225
the age between 26 and 36 have more self-confi dence than other participants Participants in the age of 36ndash45 fi nd their system easier to learn
Signifi cant ANOVA results for education (Table 827 ) show that participants with a PhD have higher self-confi dence than other participants and also they care less about privacy issues
ANOVA results for entity types show that (Table 828 ) participants from family treatment centers are more satisfi ed with their system and they believe that their system is aligned with their workfl ow On the other hand government and private hospital participants stated that their systems are effectively integrated with diag-nostic healthcare devices
ANOVA results for software usage goal show that participants who use the sys-tem for medical purposes fi nd the system more useful and show a more positive attitude to the usage of the system On the other hand participants who use the system for management and fi nance purposes are more self-confi dent and keen on information sharing Whole results are implied in Table 829
8546 Cluster Analysis
Sample clustering has been applied to the participants with two different construct sets Two- three- and four-group cluster analysis have been applied and the four- group analysis has given the most signifi cant results in both sets Case numbers have been shown for each group in Table 830 for the fi rst analysis
The fi rst cluster is the moderately satisfi ed cluster They have an average attitude and average satisfaction with most of the constructs The second cluster is the least satisfi ed cluster with low satisfaction rates The third cluster is the totally satisfi ed one with high satisfaction rates and positive attitude They are also pleasant about the general functionalities and specifi cations The last cluster is the partially adopted group They are not pleasant about all the functionalities or specifi cations of the system Thus they are partially satisfi ed
Table 827 ANOVA results for education
Construct F Sig High S Uni Masters PhD
23 189 45 42 IntegrationHW 1521 0000 389 149 173 186 EoL 565 0001 296 385 398 381 SelfConfi dence 561 0001 330 408 387 419 FuncDose 433 0005 070 062 045 037 IntegrationSW 366 0013 085 054 041 060 UserInterface 323 0023 340 374 379 355 Satisfaction 320 0024 346 402 406 381 EoU 288 0036 375 409 417 385 Mobility 279 0041 317 375 374 358 PrivacyMD 273 0044 330 357 387 319
8 Adoption Factors of Electronic Health Record Systems
226
Table 828 ANOVA results for entity
Construct F Sig FHC Gov- Pri
251 48 IntegrationHW 11306 0000 139 367 Satisfaction 8033 0000 414 300 UserInterface 7707 0000 382 306 TTF 5837 0000 404 308 Infusion 5676 0000 389 277 EoU 4272 0000 415 345 Diffusion 3947 0000 403 319 Flexibility 3924 0000 380 301 SupportQuality 3584 0000 355 268 Mobility 3580 0000 382 298 Usefulness 3516 0000 401 337 UseDensity 3118 0000 416 342 EoL 2629 0000 393 310 Info 2383 0000 395 338 QualityofCare 2014 0000 371 314 Attitude 1675 0000 415 369 Effi cientUse 1561 0000 779 669 SelfConfi dence 1336 0000 410 356 Security 1183 0001 387 339 PrivacyUA 887 0003 354 300 PrivacyMD 593 0015 362 317 FuncRange 535 0021 047 066
Results of the fi rst clustering can be seen in Fig 813 and Table 831 Second clustering has been done related to characteristics of the systems and
user behavior (Table 832 ) The fi rst group was the average systems Their characteristics were fulfi lling the
user expectations somehow The second cluster was the least functional systems The third cluster was the moderate systems They had similar performance to the average system cluster however their performance was shown on different charac-teristics The fourth cluster was the capable systems They had high-performance characteristics in each area Detailed results of the clustering can be seen in Table 833 and Fig 814
8547 Participant Comments
At the end of the questionnaire two open-ended questions were asked to the participants regarding their requests for modifi cations and extra functionalities related to the systems The following quotes include selected responses from the participants
OM Koumlk et al
227
Table 829 ANOVA results for goal
F Sig Medical MngmtmdashFin
257 44 SupportQuality 834 0004 347 301 Satisfaction 809 0005 401 360 Usefulness 692 0009 394 363 Flexibility 620 0013 372 337 Security 560 0019 384 349 EoU 556 0019 408 380 FuncDose 519 0024 059 040 QualityofCare 511 0024 366 335 Mobility 483 0029 373 339 Infusion 462 0032 377 341 Attitude 395 0048 410 386 AccessALL 355 0060 084 071 Diffusion 354 0061 393 366 PrivacyUA 325 0072 350 316 Info 227 0133 388 369 UserInterface 190 0169 371 357 Effi cientUse 181 0180 767 727 UseDensity 167 0197 407 389 TTF 122 0270 391 375 SelfConfi dence 095 0331 398 414 IntegrationSW 079 0374 054 062 FuncRange 078 0377 051 044 EoL 034 0562 381 370 PrivacyMD 028 0599 356 345 IntegrationHW 021 0648 172 184 KnowledgeShare 015 0696 322 330 FuncXMed 000 0973 051 051
Table 830 Cluster distribution
Cluster of cases in each cluster Percentage
Moderate 103 342 Least satisfi ed 17 56 Totally satisfi ed 161 535 Partially adopted
20 66
Currently we only have access to the patient records related to the family health centers In order to make a full assessment we need to see the whole medical history of the individual (Healthcare Practitioner)
We should be able to request laboratory tests x-ray diagnosis and etc for patient via online channel from other institutions Also the results should be delivered via same mod-ule quickly and effectively (Healthcare Practitioner)
The system should be integrated with the MEDULA (Social Insurance Medicine System) Otherwise we canrsquot be able to see which medicines the patient has been prescribed
8 Adoption Factors of Electronic Health Record Systems
228
0123456789
EoU
EoL
Usefulness
Attitude
Satisfaction
QualityofCare
EfficientUse
UseDensity
Diffusion
Infusion
1
2
3
Fig 813 Cluster analysis 1
Table 831 Cluster analysis 1 results
1 2 3 4
EoU 383 263 442 320 EoL 346 276 414 355 Usefulness 366 278 432 271 Attitude 391 315 442 280 Satisfaction 367 202 448 280 QualityofCare 344 245 401 230 Effi cientUse 641 335 880 790 UseDensity 386 229 448 295 Diffusion 376 229 435 225 Infusion 346 171 426 235
Table 832 Cluster analysis 2 distribution
Cluster of cases in each cluster Percentage
Average systems 65 223 Least functional systems 29 100 Moderate systems 125 430 High-performance systems 72 247
OM Koumlk et al
229
to and their dosages This creates problems when we need to prescribe to the patient (Healthcare Practitioner)
These three quotes defi nitely show that caretakers require integration with other healthcare institutions Integration with other institutions will provide access to the full medical history of the patients and also the whole medical examination and testing process will be kept in a common environment
System has low response times This creates delays in our caretaking process (Healthcare Practitioner)
In the user interface warnings should come up about the patientrsquos allergies vaccine deadline and etc (Healthcare Practitioner)
Table 833 Cluster analysis 2 results
1 2 3 4
Flexibility 372 247 352 442 Info 390 250 370 465 AccessALL 083 079 075 093 KnowledgeShare 263 259 334 382 Mobility 370 226 346 462 PrivacyMD 198 328 398 424 PrivacyUA 314 266 320 449 Security 388 233 359 467 SelfConfi dence 398 303 383 476 SupportQuality 356 198 320 417 TTF 369 263 382 469 UserInterface 373 266 360 428
000
100
200
300
400
500
600Flexibility
Info
AccessALL
KnowledgeShare
Mobility
PrivacyMD
PrivacyUA
Security
SelfConfidence
SupportQuality
TTF
UserInterface
1
2
3
4
Fig 814 Cluster analysis 2
8 Adoption Factors of Electronic Health Record Systems
230
I canrsquot make changes in the past information sometimes mistakes or mistypes exist in the recorded data (Healthcare Practitioner)
These three comments raise the caretakersrsquo main problems regarding the sys-temrsquos performance or user interface The last one discusses the data update mecha-nism However that request needs a detailed and secure process map in order to be successful since there are certain privacy data quality and security issues
Sometimes properly working modulesfunctions of the systems are being altered due to testing new functions This creates problems as they also break the properly working mod-ules (Healthcare Practitioner)
This request is related with the updates in the system and their effects Developers should consider the ongoing work of the caretakers and system updates should not go live without a proper testing period that does not affect the live system
A mobile version of this system should be developed since we often conduct on-site visits to patient homes or villages out of the city center (Healthcare Practitioner)
This quote is mainly aligned with the requirements of our era Many software offer mobile applications and mobile versions After the main developments are complete in the system developers should consider the mobile version of the appli-cations as the next step
86 Conclusion
As the usage of electronic health record systems increases developers systems architects and project managers will focus on them more Adoption process and diffusion factors will be the main input for the implementation and development of electronic health record systems This study has focused on the adoption factors and developed a model implying the interaction of intermediary dependent and exter-nal factors and their effects on the use and attitude
Main determinants for EHR adoption process have been defi ned as attitude ease of use and usefulness These results also align with TAM TAM2 and UTAUT It is also found that attitude ease of use usefulness and ease of learning have effects on satisfaction infusion diffusion and use density processes
Effi cient use of the electronic health record systems is mainly affected by the functionalities of the systems user interface integration taskndashtechnology fi t infor-mation quality and accessibility Taskndashtechnology fi t was also investigated by Hyun et al in 2009 and it was stated that the system should fi t with workfl ows of the healthcare employees
In conclusion this study provided a model in light of a quantitative fi eld survey study and is supported by the prior literature The relationship among dependent factors intermediary factors and external factors has been analyzed
OM Koumlk et al
231
861 Limitations
This study had some limitations First of all it has been applied among three hospi-tals and Manisa family health practitioners Results may differ when the quantitative fi eld survey study has been applied in different geographic regions and among differ-ent professionals Secondly all participants of the survey were using centralized record systems Ones that have their own individual systems for record keeping might have different adoption factors It would be sounder if we could recruit strati-fi ed representative health professional samples from different health units of the country such as state hospitals university hospitals private hospitals primary health-care facilities and those who use specialized record systems such as a cancer regis-try As another restriction the majority of our data come from the primary healthcare facilities of Manisa in which the data were collected via an announcement from the province health directorate of Manisa This might positively bias the results
862 Implications
During this study main adoption factors of EHR system usage have been analyzed
Effi cient use of the EHR system is found to be mainly related with the alignment between the systemrsquos workfl ow and the individualrsquos daily tasks It can be stated that the more the developers adapt their systemsrsquo workfl ows to the individualsrsquo tasks the more effi ciently their system will be used or this can be considered vice versa Also effi cient use of the system is found to be mainly dependent on the functionalities of the system and its integration with other required software Developers should focus on offering more functionality with their system such as dose functionality and medical critical value range Other factors that developers or software architects should take into account are information quality user interface and accessibility
The information quality factor is considered a multi-construct factor in our study We defi ned information quality from completeness accuracy and up-to-dateness aspects Future studies may also include other aspects and take into account differ-ent factors
Quality of care was found to be an important factor during the whole research since caretakers aim to offer the best care The relationship between quality of care and EHR systems is found to be usefulness of the system and the individualrsquos attitude
Infusion rate is found to be dependent on usefulness and ease of use of the sys-tem So developers should try to focus on creating systems which are found to be more useful and easy to use
Usefulness of the system is defi ned with information quality fl exibility mobility user interface and ease of use factors in the developed model Moreover the individualrsquos
8 Adoption Factors of Electronic Health Record Systems
232
self-confi dence is taken into account as an important factor This shows that individuals who have more computer experience will fi nd the system more useful
Ease of use of the system is found to be correlated with information quality ease of learning mobility and user interface of the system We can say that software developers should focus on the user interface of their product and make it easier to learn with guidelines Also this study proves that mobility is an important adoption factor and should be considered with priority
Outputs of this study and the developed model can be a really useful input for further researches More comprehensive or more detailed frameworks can be devel-oped from this research
87 Appendices
871 1 Interview Questions
1 Adınız 2 Yaşınız 3 Medikal Kayıt Sistemlerini daha oumlnce kullandınız mı 4 Medikal Kayıt Sistemlerini kullanmanın gerekli olduğunu duumlşuumlnuumlyor musunuz
Nedenleri nelerdir 5 Medikal Kayıt Sistemlerinin kullanım kolaylığı hakkında ne duumlşuumlnuumlyorsunuz 6 Medikal Kayıt Sistemlerinin sizce sağladığı faydalar neledir 7 Medikal Kayıt Sistemleri kullanmanız gerektiği durumlarda kayıtları kendiniz
mi tutuyorsunuz yoksa bu konuda daha yetkin kişilerden yardım mı alıyorsunuz 8 Medikal Kayıt Sistemleri geliştirilirken hangi konulara dikkat edilmesi
gerektiğini duumlşuumlnuumlyorsunuz 9 Medikal Kayıt Sistemleri kullanırken aradığınız bilgiye ulaşmakta ne gibi zor-
luklar ccedilekmektesiniz 10 Hastalarınız medikal kayıtlarının dijital ortamda tutulduğundan haberdarlar mı 11 Meslektaşlarınızla medikal kayıtları paylaşarak bilgi aktarımında bulunmakta
mısınız 12 Medikal Kayıt sistemleri kullanırken teknolojik zorluklarla karşılaştınız mı 13 Medikal Kayıt Sistemlerinde size goumlre bulunması zorunlu fonksiyonaliteler
nelerdir 14 Medikal kayıtlarınızı kendiniz mi tutmaktasınız yoksa bu konuda medikal
sekreterlerasistanlarınızdan yardım aldığınız olmakta mıdır 15 Medikal kayıtlarınızı başkalarına tutturdugunuz durumlarda kayıtların oumlnem
derecesi (ilgili hasta operasyon hastalık) bu kararı vermenizde etken oluyor mu
OM Koumlk et al
233
(con
tinue
d)
Oumlze
llikl
er
Anl
am
Accedilı
klam
a D
emog
raph
ics
Dem
ogra
fi k
Kul
lanı
cını
n de
mog
rafi k
oumlze
llikl
eri
1 A
ge
Yaş
K
ulla
nıcı
nın
yaşı
2
Edu
catio
nal L
evel
E
ğitim
Duumlz
eyi
Kul
lanı
cını
n eğ
itim
duumlz
eyi
3 G
ende
r C
insi
yet
Kul
lanı
cını
n ci
nsiy
eti
4 In
com
e G
elir
K
ulla
nıcı
nın
aylık
gel
iri
5 M
arita
l sta
tus
Evl
ilik
Dur
umu
Kul
lanı
cını
n ev
lilik
dur
umu
6 Jo
b ex
peri
ence
İş
Den
eyim
i K
ulla
nıcı
nın
iş d
eney
imi
7 Pl
ace
of r
esid
ence
İk
amet
Yer
i K
ulla
nıcı
nın
ikam
et y
erin
in ouml
zelli
liği (
koumly
ilccedile
şeh
ir m
erke
zi)
8 O
ccup
atio
n M
esle
k K
ulla
nıcı
nın
mes
leği
In
term
edia
ry
Ara
cı Ouml
zelli
kler
K
ulla
nıcı
yaz
ılım
la e
tkile
şim
e ge
ccediltiğ
i sır
ada
orta
ya ccedil
ıkan
oumlz
ellik
ler
kiş
inin
yaz
ılım
ı kul
land
ığın
da k
azan
dığı
fay
da
kulla
nım
ın k
olay
olm
ası g
ibi
9 E
ase
of u
se
Kol
ay K
ulla
nım
Y
azılı
mın
kol
ay k
ulla
nım
ı 10
U
sefu
lnes
s Fa
yda
Yaz
ılım
ın k
ulla
nım
dan
doğa
n fa
yda
11
Eas
e of
lear
ning
K
olay
Oumlğr
enm
e Y
azılı
mı k
ulla
nmay
ı oumlğr
enm
enin
kol
aylığ
ı C
linic
al v
aria
bles
K
linik
Oumlze
llikl
eri
Has
tane
ile
ilgili
değ
işke
nler
12
G
eogr
aphi
c ar
ea
Coğ
rafi
Kon
um
Has
tane
nin
coğr
afi k
onum
u (ş
ehir
mer
kezi
ilccedil
e k
oumly g
ibi)
13
Po
pula
tion
serv
iced
H
izm
et E
ttiği
Nuumlf
us
Has
tane
nin
hizm
et v
erdi
ği k
işi s
ayıs
ı 14
H
ospi
tal s
ize
Has
taha
ne B
uumlyuumlk
luumlğuuml
H
asta
neni
n fi z
ikse
l buumly
uumlkluuml
ğuuml
15
Oth
er c
linic
al v
aria
bles
D
iğer
Değ
işke
nler
H
asta
ne il
e ilg
ili d
iğer
değ
işke
nler
16
A
dequ
ate
reso
urce
s K
ayna
klar
H
asta
neni
n se
rvis
icin
ayi
rabi
lece
gi k
ayna
klar
17
C
linic
al s
peci
alty
U
zman
lık A
lanı
H
asta
neni
n ge
nel u
zman
lık a
lanı
Su
ppor
t D
este
k Y
azılı
mı k
ulla
nanl
ara
veri
len
tekn
ik d
este
k
87
2 2
Exp
ert F
ocus
Gro
up Q
uest
ionn
aire
8 Adoption Factors of Electronic Health Record Systems
234
18
Man
ager
ial s
uppo
rt
Youmln
etim
Des
teği
Y
oumlnet
icile
rin
serv
isin
kul
lanı
lmas
ı iccedili
n ve
rdiğ
i des
tek
19
Peer
sup
port
A
rkad
aş D
este
ği
Yaz
ılım
kul
lanı
mı s
ıras
ında
yaş
ıtlar
ının
dan
veya
ark
adaş
ları
ndan
al
dığı
des
tek
20
Prof
essi
onal
sup
port
Pr
ofes
yone
l Des
tek
Yaz
ılım
kul
lanı
mı s
ıras
ında
pro
fesy
onel
lerd
en a
lınan
des
tek
21
Ven
dor
supp
ort
Satıc
ı Des
teği
Sa
tıcı fi
rm
anın
sağ
ladı
ğı y
ardı
m v
e de
stek
22
Q
ualit
y of
sup
port
D
este
ğin
Kal
itesi
V
erile
n ya
rdım
ve
dest
eğin
kal
itesi
23
So
cial
infl u
ence
So
syal
Etk
enle
r Y
azılı
mı k
ulla
nan
kişi
nin
ccedilevr
esin
deki
lerd
en
aldı
ğı e
tki
24
Com
patib
ility
U
yum
lulu
k Y
azılı
mı ouml
ncek
i suumlr
uumlmle
ri v
eya
ccedilalış
tırıld
ığı o
rtam
daki
diğ
er
sist
emle
re u
yum
u C
onte
nt
Serv
is İ
ccedileri
ği
Yaz
ılım
ın s
undu
ğu b
ilgin
in iccedil
eriğ
i 25
A
ccur
acy
Doğ
rulu
k Su
nula
n bi
lgin
in d
oğru
luğu
26
St
anda
rdiz
atio
n St
anda
rd
Bilg
inin
sta
ndar
t bir
şek
ilde
sunu
lmas
ı 27
In
form
atio
n qu
ality
B
ilgi K
alite
si
Sunu
lan
iccediler
iğin
kal
itesi
28
Se
curi
ty
Bilg
inin
Guumlv
enliğ
i İccedil
eriğ
in b
aşka
ları
nın
eriş
emey
eceğ
i bir
ort
amda
sak
lanm
ası
29
Tool
exp
erie
nce
Den
eyim
K
ulla
nıcı
nın
benz
er s
ervi
s ya
uumlruuml
n ile
ilgi
li ge
ccedilmiş
den
eyim
leri
30
Im
age
İmaj
K
ulla
nıcı
ları
n et
rafl a
rınd
aki i
nsan
lara
ken
dile
rini
far
klı
ayrı
calık
lı ve
oumlnc
uuml gouml
ster
me
iste
ği
31
Satis
fact
ion
Mem
nuni
yet
Kul
lanı
cını
n ya
zılım
dan
mem
nun
kalm
ası
32
Vol
unta
rine
ss
Goumln
uumllluuml
luumlk
Kul
lanı
cını
n yuuml
kuumlm
luumlluuml
ğuuml o
lmad
an is
teye
rek
yazı
lımı k
ulla
nmas
ı 33
Fa
cilit
atin
g co
nditi
ons
Kol
ayla
ştır
ıcı
Koş
ulla
r Y
azılı
mın
kul
lanı
mın
ı kol
ayla
ştır
acak
koş
ulla
r
34
Func
tiona
l cha
ract
eris
tics
Fonk
siyo
nel
Oumlze
llikl
er
Yaz
ılım
ın f
onks
iyon
el ouml
zelli
kler
i
35
Flex
ibili
ty
Kiş
isel
leşt
irile
bilir
lik
Yaz
ılım
ın f
onks
iyon
ları
nı is
teğe
goumlr
e de
ğişt
ireb
ilmek
Oumlrn
eğin
m
enuumln
uumln s
ıras
ı uumlze
rind
e de
ğişi
klik
yap
abilm
esi
Oumlze
llikl
erA
nlam
Accedilı
klam
aD
emog
raph
ics
Dem
ogra
fi kK
ulla
nıcı
nın
dem
ogra
fi k ouml
zelli
kler
i
(con
tinue
d)
OM Koumlk et al
235
36
Acc
essi
bilit
y U
laşa
bilir
lik
Yaz
ılım
ın k
ulla
nıcı
lar
tara
fınd
an k
olay
ula
şala
bilir
olm
ası
37
Beh
avio
ral c
ontr
ol
Kul
lanı
cını
n ya
zılım
ı kul
lanm
ak iccedil
in y
eter
li ye
tene
kler
inin
ka
ynağ
ının
ve
fırs
atın
ın o
lup
olm
adığ
ı alg
ısı
38
Job
rele
vanc
e Iş
e U
ygun
luk
Yaz
ılım
ın d
okto
run
işin
e uy
gunl
uğu
Med
ical
M
edik
al
Yaz
ılım
med
ikal
ala
ndak
i etk
ileri
39
R
ate
of s
ucce
ssfu
l tre
atm
ents
B
aşar
ılı T
edav
ileri
n O
ranı
Y
azılı
mın
kul
lanı
cını
n uy
gula
dığı
teda
vile
rin
oran
ını a
rtır
mas
ı 40
R
ate
of s
ucce
ssfu
l dia
gnos
is
Baş
arılı
Teş
hisl
erin
O
ranı
Y
azılı
mın
kul
lanı
cını
n ko
yduğ
u te
şhis
leri
n or
anın
ı art
ırm
ası
41
Rat
e of
dec
isio
n ef
fi cie
ncy
Kar
ar v
erm
e ve
rim
liliğ
inin
ar
tırılm
ası
Yaz
ılım
ın k
ulla
nıcı
nın
kara
r ve
rme
doğr
uluğ
unu
artır
mas
ı
42
Res
pons
e tim
e Si
stem
in Ccedil
alış
ma
Hız
ı Y
azılı
mın
kul
lanı
m z
aman
ı Y
azılı
mın
kul
lanı
mas
ı ccedilok
zam
an a
labi
lir v
e ku
llanı
cıla
rın
yete
rinc
e va
kti o
lmay
abili
r 43
G
uide
lines
D
oumlkuumlm
anta
syon
Y
azılı
mın
doumlk
uumlman
tasy
onu
44
Hab
it A
lışka
nlık
K
ulla
nıcı
nın
mev
cut a
lışka
nlar
ı 45
T
rust
G
uumlven
ilirl
ik
Kul
lanı
cını
n ya
zılım
a du
yduğ
u guuml
veni
C
ompu
ter
liter
acy
Bilg
isay
ar
Oku
ryaz
arlığ
ı K
ulla
nıcı
nın
bilg
isay
ar b
ilgis
i ve
okur
yaza
rlığ
ı
46
Com
pute
r ex
peri
ence
B
ilgis
ayar
Den
eyim
i K
ulla
nıcı
nın
kaccedil
yıld
ır b
ilgis
ayar
kul
land
ığı
47
Com
pute
r lit
erac
y B
ilgis
ayar
O
kury
azar
lığı
Kul
lanı
cını
n bi
lgis
ayar
kul
lanı
mın
ı ne
kad
ar iy
i bild
iği
48
Use
r in
terf
ace
Ekr
an G
oumlruumln
tuumlsuuml
Y
azılı
mın
kul
lanı
cı e
kran
ları
nın
oumlzel
likle
ri
49
Task
ndashtec
hnol
ogy
fi t
Tekn
oloj
iGoumlr
ev U
ygun
luğu
Y
azılı
mın
kul
lanı
cını
n ya
ptığ
ı goumlr
evle
re u
ygun
luğu
50
R
isk
Ris
k Y
azılı
mın
kul
lanı
lmas
ında
n do
ğabi
lece
k ol
an r
iskl
er
51
Secu
rity
G
uumlven
lik
Yaz
ılım
ın k
ulla
nılm
ası i
le o
luşa
n bi
lgik
ulla
nıcı
has
ta g
uumlven
liği
8 Adoption Factors of Electronic Health Record Systems
236
873 3 Factor Analysis Results for Pilot
1 2 3 4 5 6 7 Usef6 0967 0087 minus0151 0147 minus0059 minus0096 minus0017 UserInterface1 0943 0183 0036 0205 0062 0139 0107 EoU2 0931 minus0120 0001 0091 0075 0080 minus0314 Usef4 0918 0071 0152 0059 minus0292 0182 minus0082 EoU3 0902 minus0001 0033 0230 minus0293 minus0159 minus0146 FuncXMed 0868 0294 0041 0064 minus0073 minus0303 0238 EoU1 0868 0052 minus0148 minus0057 0027 0464 minus0058 UserInterface8 0855 minus0250 0273 0291 minus0185 minus0112 0019 UseDensity 0826 0209 minus0276 0346 minus0251 minus0115 minus0038 Effi cientUse 0824 minus0506 minus0116 0092 0081 0084 0173 SupportQ1 0789 minus0142 0335 0260 0204 0195 0312 Usef1 0774 0012 minus0399 0350 minus0035 0344 minus0011 Infusion 0765 minus0110 0120 0121 0539 0078 minus0276 Satisfaction2 0750 minus0018 0483 0408 minus0065 minus0053 minus0175 Diffusion 0710 minus0400 0054 0265 minus0249 0448 0016 TTF2 0710 minus0369 minus0237 0142 0426 0308 0091 Completeness 0670 0257 minus0399 0488 minus0255 0129 0078 UserInterface5 0668 0239 minus0116 0566 0327 0233 minus0042 UserInterface6 0595 minus0536 0078 0549 0002 0208 0092 UserInterface2 0543 minus0423 0413 0417 minus0211 0341 0144 Usef3 0027 0943 0157 0137 minus0247 minus0077 minus0016 QoCare1 0249 minus0915 0200 0033 0012 minus0038 minus0240 Attitude1 minus0065 0913 minus0005 0262 0195 minus0134 0194 TTF3 0237 0859 minus0166 0368 0201 minus0034 0038 UptoDate 0354 0769 minus0283 0294 minus0341 minus0032 0006 SupportQ2 0396 minus0684 0442 minus0041 minus0401 0098 minus0094 Attitude2 0271 0683 0163 minus0027 0111 minus0389 0519 Flexibility2 0374 minus0619 0040 0355 0388 0444 minus0018 SelfConfi dence 0128 0605 0145 0506 minus0489 minus0320 minus0004 PrivacyUA minus0346 minus0581 0385 0011 0383 0393 0304 IntegrationSW 0327 minus0561 minus0456 minus0242 minus0295 minus0147 0450 QoCare2 minus0120 minus0086 0965 0013 minus0118 0171 0058
(continued)
OM Koumlk et al
237
Usef7 0041 minus0035 0965 0004 0187 minus0149 0095 Consistency 0317 0247 0826 0265 minus0252 minus0065 minus0135 Mobility2 0033 minus0375 0822 minus0106 minus0276 0113 0289 Mobility3 0254 0487 0727 0041 minus0079 minus0395 0074 FuncDose minus0148 minus0259 0698 minus0266 0350 0174 minus0447 AccessALL minus0148 minus0259 0698 minus0266 0350 0174 minus0447 UserInterface3 minus0184 minus0398 0656 minus0257 minus0550 0092 minus0017 Usef5 minus0264 minus0139 0548 minus0358 0536 0389 0210 Security1 0244 0232 0086 0833 0094 minus0209 0364 Satisfaction3 0456 0250 0034 0812 0185 0068 minus0175 EoL 0258 0388 minus0256 0809 0075 0230 minus0067 Satisfaction1 0584 0102 0160 0771 0037 minus0031 minus0159 Accuracy 0634 0008 minus0174 0750 0061 0005 0021 Standardization 0127 minus0251 minus0433 0543 0363 0396 0388 FuncRange minus0251 0010 0238 0068 0934 0002 0050 PrivacyMD minus0044 0313 minus0258 0162 0830 minus0286 0193 TTF1 0467 minus0147 minus0336 0325 0693 0176 minus0176 Usef2 0360 0403 0471 minus0005 minus0612 minus0333 0033 Flexibility3 0210 minus0164 0310 0087 minus0147 0854 minus0273 Flexibility1 0500 minus0007 minus0008 minus0004 0180 0844 minus0073 IntegrationHW 0181 minus0623 0269 0130 minus0081 0645 minus0260 UserInterface4 0341 0408 0349 0218 minus0001 minus0584 0456 UserInterface7 0435 minus0430 minus0497 0064 0084 0568 0210 QoCare3 0506 0102 0431 minus0134 minus0299 minus0524 0407 Mobility1 0270 minus0041 0141 minus0096 minus0029 minus0003 minus0946 KnowledgeShare minus0011 0042 0181 minus0374 0069 minus0191 0886 EoU4 minus0195 0488 0319 0256 0021 minus0313 0677 Security2 0182 0388 minus0476 0401 0216 0017 0618
(continued)
8 Adoption Factors of Electronic Health Record Systems
238
Tabl
e 8
34
Fact
or a
naly
sis
for
all i
tem
s
1 2
3 4
5 6
7 8
9 10
U
sef3
0
822
012
4 0
147
028
1 0
135
001
1 0
104
minus0
096
minus0
116
003
4 Q
oCar
e3
079
4 0
177
014
2 0
240
003
3 0
105
021
1 0
031
minus0
006
005
6 U
sef2
0
793
018
8 0
146
025
3 0
128
minus0
001
011
6 minus
004
6 minus
011
8 0
038
Atti
tude
2 0
793
030
0 0
132
013
2 0
115
005
9 minus
014
6 0
011
001
8 0
098
Atti
tude
1 0
781
024
9 0
209
025
0 0
121
008
5 minus
012
5 minus
001
3 minus
001
1 0
055
Use
f1
077
4 0
233
019
8 0
249
008
3 minus
009
8 0
004
minus0
021
000
5 0
027
Dif
fusi
on
074
9 0
279
022
9 0
192
minus0
030
001
2 0
146
010
5 minus
002
3 0
036
QoC
are2
0
702
023
9 minus
001
2 0
037
010
9 0
087
020
8 0
132
020
4 0
048
Use
f6
070
1 0
389
028
3 0
186
007
8 0
031
minus0
137
006
5 minus
010
1 0
045
Use
f4
063
0 0
487
031
1 0
289
008
3 minus
004
5 0
038
009
4 0
038
002
9 Sa
tisfa
ctio
n3
055
9 0
482
043
1 0
255
004
2 minus
001
4 0
127
011
3 minus
006
5 0
018
Infu
sion
0
532
036
9 0
287
026
3 minus
001
5 0
056
021
0 0
253
minus0
103
011
1 U
seD
ensi
ty
052
2 0
359
032
3 0
397
minus0
041
minus0
093
002
1 0
082
minus0
047
008
6 Q
oCar
e1
052
0 0
179
008
2 0
075
004
9 0
104
050
4 0
068
020
6 0
140
Satis
fact
ion1
0
484
044
6 0
476
034
9 0
094
minus0
030
012
1 0
163
minus0
006
minus0
019
EoU
2 0
480
043
2 0
378
036
3 0
105
002
2 minus
016
4 0
193
004
6 minus
004
3 Sa
tisfa
ctio
n2
046
7 0
461
043
3 0
371
010
3 minus
003
2 0
098
017
0 minus
000
6 minus
002
2 U
sef7
0
448
031
9 0
145
033
0 0
212
022
9 0
162
029
9 0
114
004
1 Se
lfC
onfi d
ence
0
427
015
6 0
169
036
9 0
147
minus0
113
minus0
317
021
4 0
048
014
5 U
sef5
0
407
016
6 0
174
033
0 0
309
017
8 0
169
018
9 0
174
006
1 U
serI
nter
face
6 0
284
071
5 0
158
016
5 0
171
minus0
081
015
1 minus
005
0 0
085
001
8
87
4 4
Fac
tor
Ana
lysi
s R
esul
ts
OM Koumlk et al
239
Use
rInt
erfa
ce1
032
3 0
711
033
1 0
164
minus0
002
minus0
003
minus0
052
003
8 minus
005
0 0
068
Use
rInt
erfa
ce5
036
3 0
681
028
5 0
333
004
9 minus
001
5 minus
006
8 minus
001
9 0
038
010
8 E
oU4
043
2 0
616
017
0 0
245
017
2 0
132
minus0
086
001
5 0
048
014
2 U
serI
nter
face
2 0
208
061
5 0
357
024
8 0
115
minus0
010
021
4 0
079
010
3 minus
005
2 U
serI
nter
face
4 0
317
058
9 0
128
035
4 0
146
004
0 minus
014
5 minus
004
5 0
019
010
7 Fl
exib
ility
3 0
359
053
7 0
211
032
0 0
244
016
6 0
213
014
4 minus
003
9 minus
001
7 Fl
exib
ility
1 0
327
051
0 0
099
014
7 0
054
027
8 0
176
minus0
030
minus0
197
minus0
031
Use
rInt
erfa
ce8
035
5 0
509
018
9 0
212
015
4 0
007
011
9 0
158
020
4 minus
009
5 E
oU1
046
5 0
485
041
0 0
207
002
4 minus
006
7 minus
011
6 0
129
minus0
043
003
4 M
obili
ty1
033
3 0
458
031
2 0
097
012
6 minus
012
3 0
179
031
5 minus
016
3 0
126
TT
F2
014
7 0
200
074
5 0
262
012
8 0
052
027
8 0
042
008
8 0
029
TT
F3
031
9 0
233
067
5 0
211
021
8 0
079
006
0 0
019
minus0
085
002
9 E
oU3
033
7 0
289
065
5 0
171
009
5 minus
003
1 minus
013
9 0
141
003
9 minus
003
2 E
oL
012
1 0
287
065
0 0
178
005
1 0
083
minus0
228
014
4 minus
002
3 0
029
TT
F1
010
1 0
190
064
9 0
341
008
3 0
087
029
1 minus
001
4 minus
003
8 0
037
Use
rInt
erfa
ce7
028
7 0
221
063
2 minus
007
5 0
180
minus0
050
001
5 minus
003
6 0
077
006
8 E
ffi c
ient
Use
0
237
032
4 0
454
030
5 0
123
016
0 0
163
021
6 0
250
003
6 Pr
ivac
yUA
0
059
004
6 0
356
019
9 0
318
007
4 0
256
minus0
093
003
7 0
288
Acc
urac
y 0
396
025
8 0
186
066
6 0
124
minus0
049
001
8 0
014
007
3 0
004
Con
sist
ency
0
440
030
7 0
180
062
0 0
107
minus0
017
minus0
059
013
2 0
018
minus0
068
Stan
dard
izat
ion
040
7 0
319
024
9 0
614
016
4 minus
001
9 minus
014
5 0
153
minus0
008
007
4 Se
curi
ty1
026
3 0
262
015
6 0
610
005
2 0
098
014
9 0
021
minus0
020
035
2 U
ptoD
ate
047
0 0
222
018
2 0
596
019
3 0
061
minus0
033
005
2 0
021
minus0
016
(con
tinue
d)
8 Adoption Factors of Electronic Health Record Systems
240
Com
plet
enes
s 0
416
034
6 0
231
056
6 0
078
001
7 0
186
005
0 0
153
minus0
038
Secu
rity
2 0
300
030
9 0
180
055
1 0
139
007
5 minus
000
1 minus
009
1 minus
008
3 0
344
Supp
ortQ
1 0
351
040
6 0
202
050
1 0
208
007
5 0
117
007
4 minus
003
6 minus
005
0 M
obili
ty2
030
7 0
300
009
0 0
159
071
5 minus
001
9 0
043
013
0 0
058
010
2 U
serI
nter
face
3 0
066
000
7 minus
036
0 minus
013
7 minus
061
5 0
100
minus0
037
minus0
024
minus0
024
minus0
059
Mob
ility
3 0
366
033
8 0
158
025
8 0
593
013
5 minus
003
8 0
185
002
4 0
015
Func
Ran
ge
minus0
023
013
5 minus
002
9 0
023
008
2 0
751
minus0
036
007
6 0
054
minus0
097
Func
Dos
e 0
141
minus0
126
014
5 0
008
minus0
139
063
0 0
148
015
7 0
126
014
4 Fl
exib
ility
2 0
192
013
6 0
370
minus0
023
038
4 0
119
047
2 0
042
001
6 minus
002
9 A
cces
sAL
L
minus0
015
minus0
007
008
4 0
032
010
5 0
235
minus0
051
078
1 0
128
minus0
035
Supp
ortQ
2 0
298
025
0 0
054
035
5 0
124
minus0
084
019
9 0
420
minus0
034
minus0
002
Inte
grat
ionS
W
minus0
019
007
3 0
043
003
4 0
040
minus0
050
002
0 0
149
073
7 0
146
Inte
grat
ionH
W
minus0
172
minus0
141
minus0
106
minus0
132
minus0
022
036
9 minus
005
5 minus
001
2 0
547
minus0
143
Func
XM
ed
010
0 0
041
012
8 0
168
008
6 0
344
013
6 minus
011
7 0
503
minus0
222
Kno
wle
dgeS
hare
0
070
005
6 minus
005
6 0
162
004
8 minus
006
8 0
057
006
7 minus
001
4 0
792
Priv
acyM
D
012
6 minus
004
4 0
428
minus0
183
011
2 0
024
minus0
125
minus0
196
003
9 0
536
Ext
ract
ion
met
hod
pri
ncip
al c
ompo
nent
ana
lysi
s R
otat
ion
met
hod
var
imax
with
Kai
ser
norm
aliz
atio
n a R
otat
ion
conv
erge
d in
13
itera
tions
Tabl
e 8
34
(con
tinue
d)
12
34
56
78
910
OM Koumlk et al
241
Component 1
Satisfaction 0881 Diffusion 0879 Infusion 0860 UseDensity 0822 QualityofCare 0791 Effi cientUse 0697
Extraction method principal component analysis a One component extracted
Table 835 Factor analysis for dependent constructs
Table 837 Factor analysis for external constructs
1 2 3 4 Info 0875 0003 0033 0105 UserInterface 0839 0050 0041 minus0024 Mobility 0795 0036 0096 0103 SupportQuality 0789 0049 minus0041 0108 Flexibility 0765 0226 0106 minus0075 Security 0738 minus0058 0230 0102 TTF 0702 0203 0308 minus0088 SelfConfi dence 0607 minus0201 minus0013 0302 FuncXMed 0203 0630 minus0007 0060 FuncRange 0084 0622 minus0147 0117 IntegrationHW minus0345 0585 minus0038 0169 FuncDose 0045 0577 0218 0090 PrivacyMD 0045 minus0028 0792 minus0032 PrivacyUA 0355 0209 0555 minus0102 KnowledgeShare 0127 minus0341 0550 0456 IntegrationSW 0028 0259 0102 0656 AccessALL 0156 0268 minus0221 0582
Component 1
EoU 0925 Usefulness 0911 Attitude 0883 EoL 0582
Extraction method principal component analysis a One component extracted
Table 836 Factor analysis for intermediary constructs
8 Adoption Factors of Electronic Health Record Systems