Faculty of Economics and Business MSc BA – Change Management
January 2017
A Study on the Implementation Factors that Influence the Adoption of E-health Applications by Employees in Intramural Mental and Elderly
Healthcare Organisations
A Systematic Literature Review & Qualitative Case Study
Jamie Assen (S2083949)
Supervisor: Dr. M.A.G. van Offenbeek Second Assessor: M.L. Hage
ABSTRACT
Since the worldwide elderly population is growing, it is urgent to make elderly healthcare more cost-effective. The use of e-health can help in making elderly healthcare more efficient. However, limited research has been conducted on the mechanisms underlying successful implementation of e-health applications in healthcare settings. As a reaction to this, this study explores the implementation factors that can influence the adoption of intramural e-health applications by employees in mental and elderly healthcare organisations. By means of a systematic literature review and a case study, a conceptual framework for an action-based e-health implementation strategy is developed. Twenty-six action-steps that can influence adoption of e-health applications in intramural mental and elderly healthcare organisations are identified.
Keywords: e-health, implementation, adoption, systematic literature review
Word Count: 13.365 (excluding figures, tables, and appendices)
Table of Contents
1. Introduction 3
1.1 Research Aim 4
1.2 Research Process 4
1.3 Practical & Theoretical Relevance 5
2. Theoretical Background 6
2.1 E-health Implementation as Techno-Change 6
2.2 Technology Adoption in Healthcare Organisations 7
2.2.1 Managerial Influences on Individual Adoption 7
2.3 Domotics in Healthcare 8
2.4 Categories of Implementation Factors 8
2.5 The Implementation Stages Model of Rosenheck 9
3. Method 10
3.1 Research Cycle I: Systematic Literature Review 10
3.1.1 Step I: Framing Review Questions 10
3.1.2 Step II: Identifying Relevant Work 10
3.1.2.1 Stage I 10
3.1.2.2 Stage II 11
3.1.3 Step III: Assessing the Quality of Studies 12
3.1.4 Step IV: Summarising the Evidence 13
3.1.5 Step V: Interpreting the Findings 13
3.2 Research Cycle II: Qualitative Case Study 13
3.2.1 Research Context 14
3.2.2 Interviews 14
3.2.3 Data Analysis 15
3.2.3.1 Evaluation and Recommendations Reports 15
3.2.3.2 The Conceptual Framework for an Action-Based E-health 15
Implementation Strategy
3.3 Reliability and Validity of the Study 16
4. Results 17
4.1 Descriptive Results: Systematic Literature Review 17
4.2 Descriptive Results: Case Study x Systematic Literature Review 19
4.2.1 Need for User Involvement 20
4.2.1.1 Literature Review Cycle 20
4.2.1.2 Case Study Cycle 20
4.2.2 Need for Theoretical and Practical Training 21
4.2.2.1 Literature Review Cycle 21
4.2.2.2 Case Study Cycle 21
4.2.3 Need for Super Users 22
4.2.3.1 Literature Review Cycle 22
4.2.3.2 Case Study Cycle 22
4.2.4 Need for Evaluation 23
4.2.4.1 Literature Review Cycle 23
4.2.4.2 Case Study Cycle 23
4.2.5 Need for Post-Implementation User Support 24
4.2.5.1 Literature Review Cycle 24
4.2.5.2 Case Study Cycle 24
4.3 Interpretive Results 25
4.3.1 The conceptual Framework for an Action-Based E-health 25
Implementation Strategy
4.3.2 The Importance of the Third Implementation Stage 25 4.3.3 The Interpretation of the Fourth implementation Stage 27
5. Discussion 27
5.1 Answers to Research Questions 28
5.2 Identified Gaps in Literature and Future Research 29
5.2.1 Ignorance of the Post-Implementation and Post-Adoption Stage 29 5.2.2 Shallow Descriptions of the Adoption and Implementation Stages 29 5.2.3 A Lack of Implementation Process Research in the Field of Domotics 30
5.3 Practical and Theoretical Contributions 31
5.4 Limitations 32
6. Conclusion 33
Acknowledgements 33
References 34
Appendix A Five Steps for a Systematic Literature review 39
Appendix B Codebook Selection Criteria 40
Appendix C Characteristics of Included Studies 42
Appendix D Codebook Factor Categorisation 44
Appendix E Interview Template 45
Appendix F Evaluation & Recommendations Report 55
Appendix G List of Inductive Codes 69
Appendix H Codebook Implementation Stages 70
Appendix I Elaboration of the 26 Action-Steps 73
1. INTRODUCTION
The worldwide elderly population is growing (United Nations, 2013). Between 2000 and 2050 the proportion of people over 60 years of age will double from about 11 to 22% (Bujnowska- Fedak & Pirogowicz, 2014). Higher life expectancies and declining fertility rates lead to an extreme drop in the ratio of 15-year-olds to 60-year-olds (United Nations, 2013). Consequently, there will be a dramatic decrease in potential young individuals working in the elderly healthcare sector, while the amount of people that moves to residential care or nursing homes will further increase (Chan et al., 2009). When global demand for healthcare increases, total costs of professional healthcare will increase as well (Meulendijk et al., 2011). Consequently, urgency is there to make elderly healthcare more cost-effective. More efficiency in the elderly healthcare sector can be reached by the use of innovative solutions, like e-health (Bujnowska-Fedak &
Pirogowicz, 2014). Therefore, it is increasingly important to investigate the options and concerns regarding the use of e-health in elderly healthcare.
According to Eysenbach (2001), the following often-used description is a thorough definition of the term e-health and is used in this research as well:
E-health is an emerging field in the intersection of medical informatics, public health and business, referring to health services and information delivered or enhanced through the Internet and related technologies. In a broader sense, the term characterizes not only a technical development, but also a state-of-mind, a way of thinking, an attitude, and a commitment for networked, global thinking, to improve healthcare locally, regionally, and worldwide by using information and communication technology. (Eysenbach, 2001, p. 1)
Many studies about e-health implementations focus mainly on the usability and usefulness of
technology and the initial costs of implementation (Karsh, 2004). Though, according to Karsh
(2004), focusing solely on these technological factors or costs is not sufficient to achieve
successful implementation and adoption of e-health. The European Union has argued that
implementation of e-health applications “has almost everywhere proven to be much more
complex and time-consuming than initially anticipated” (Watson, 2010). Productivity loss due to
the disruption caused by implementation of IT technologies in healthcare settings has led to
many examples of failure (Jie Mein, Guodong & Agarwal, 2011). Implementing e-health
applications in an appropriate way could prevent those failures. For this, knowledge about the
factors concerning failure-free IT implementation in healthcare settings is required. Despite the
extensive studies about the value of IT solutions in healthcare literature, limited research has
been conducted on the mechanisms underlying successful implementation of e-health
applications in healthcare settings (Jie Mein et al., 2011). It is essential to identify
implementation factors in order to develop an appropriate and effective implementation strategy
that is adapted to these implementation factors (Fleuren, Wiefferink & Paulussen, 2004). A better
understanding of the factors that are influential in the implementation of e-health applications in the elderly healthcare sector will lead to better profiting from investments in e-health.
Subsequently, there is an urgency to investigate the implementation factors concerning e-health applications in elderly healthcare settings.
1.1 Research Aim
This study was a reaction to the limited research about the mechanisms underlying successful implementation of e-health applications in healthcare settings, as suggested by Jie Mein et al.
(2011). In order to fill this literature gap, this study intended to answer the following research question: ‘Which implementation factors influence the adoption of intramural e-health applications by employees in mental and elderly healthcare organisations?’.
To further clarify the aim of this research question, we now specify choices we made in conducting the research. First, whereas an e-health application can only be useful when it is substantially adopted by its users (Aubert & Hamel, 2001), this study focused on adoption of e- health implementations in mental and elderly healthcare settings. Furthermore, the scope of this study was limited to research focusing on work processes. The implementation factors that effect adoption of e-health implementations in current work processes were assessed. Additionally, this study focused on the use of e-health applications in care delivery processes in intramural mental and elderly healthcare institutions. Other healthcare settings, like homecare, were not included in this study. Finally, we chose to focus on one user group. Only healthcare employees as users of the e-health application were included in this study. Despite the importance of the effect of e-health implementations on intramural clients, we choose not to focus on this user group.
This decision was made because the study was commissioned by a business administration master thesis.
1.2 Research Process
This study followed Ramondt’s (1996) cyclical research process of local and scientific knowledge integration. In Figure 1, the two cyclical research processes are shown. The two different time axes show a difference in interest of knowledge between researchers and organisations. By means of two research cycles, the two knowledge fields were integrated. We now elaborate on the two research cycles that we followed. First, in order to answer the main research question, a systematic literature review was conducted. Such a review is a good method to synthesise the findings of a group of studies (Tranfield et al., 2003). The scientific world has reached consensus about the desirable methodological characteristics of the systematic literature review (Davies &
Crombie, 1998). Moreover, scholars argue that a systematic literature review provides the most
efficient and high-quality method for identifying and evaluating extensive literatures (Mulrow,
1994). The systematic literature review was the first research cycle of this study and attempted
to answer the ‘what’-oriented main research question. In order to obtain more ‘how’-oriented
answers, we conducted a second research cycle with a corresponding ‘how’-oriented research
question.
Figure 1. The two cyclical research processes of this study.
The second research cycle consisted of a case study at a large Dutch mental and elderly healthcare provider that recently implemented a new e-health application. Interviews were conducted to refine and contextualise implementation factors found in the systematic literature review. The case study attempted to answer the refined research question: ‘Which action-steps need to be taken by the implementing unit in which implementation stage(s) to influence the adoption of intramural e-health applications by employees in mental and elderly healthcare organisations?’. These action-steps can help the implementing unit to better effectuate the adoption of the use of e-health applications in current work processes. The output of this case study is a conceptual framework for an action-based e-health implementation strategy.
1.3 Practical & Theoretical Relevance
A systematic literature review can be seen as pragmatic management research, which intends to serve both practitioners and academics (Tranfield et al., 2003). On the one hand, this study is of practical relevance. By gathering knowledge from a variety of studies, our systematic literature review helps implementing units by developing a reliable knowledge base. In addition to this, the case study provides a set of “field-tested and grounded technological rules” (Van Aken, 2004, p.
Scienti'ic knowledge:
Literature about e-health implementations
Scienti'ic knowledge:
Outcomes of the systematic literature
review Analysis: A systematic
literature review
Diagnosis: Categorisation of implementation factors Action: Conducting a case study to re<ine the outcomes
Local knowledge:
Employees’ experiences with and ideas about e- health implementations Local knowledge:
Contextual knowledge about how to implement
e-health applications
Organisation problem: Need
for scienti<ic knowledge about e-health implementations
Organisation problem: Need
for knowledge about the action-
steps to take when implementing e-
health
Analysis: A case study Diagnosis: Phasing and categorisation of action-steps Action: Testing and verifying the outcomes (qualitative and quantitative) Organisation
T1
Organisation T3 Organisation
T2
T1 Researcher
T2 Researcher
T3
Researcher
221), which is of practical relevance. When an implementing unit knows these rules for the implementation of e-health in intramural mental and elderly healthcare, it will be better equipped to meet the potential benefits of e-health applications. When an application is implemented unsuccessfully, i.e., without being accurately adopted, the application will not meet its full potential. On the other hand, our systematic literature review is of theoretical relevance as well. The literature review identifies and synthesises research-based evidence and presents it in an accessible format (Green & Higgins, 2011). Simultaneously, by doing a case study in the specific field of intramural mental and elderly healthcare, we assisted in developing context sensitive science (Nowotny, Scott & Gibbons, 2001). This type of science blurs the boundaries between science, policy and practice. To conclude, our systematic literature review is a type of management research that helps both theorists and practitioners in making more sensitive judgements, which is, according to Tranfield et al. (2003), the ultimate aim of a literature review in management science.
This paper is structured as follows. The next chapter gives an overview of relevant background literature. Chapter three discusses the methodology used during the two research cycles. The fourth chapter presents the results of this study. There, an overview of the descriptive results of the two research cycles is given. Subsequently, the results are interpreted and the resulting conceptual framework for an action-based implementation strategy is presented. In chapter five, the findings are discussed in light of other relevant literature. The paper ends with a compact conclusion.
2. THEORETICAL BACKGROUNDS
In this chapter, we elaborate on the importance of e-health implementations as a techno change, on the phenomenon of technology adoption in healthcare organisations and on domotics as an e- health application. Further, we explain why the implementation stages model of Rosenheck (2001) and the categorisation of determinants of Fleuren, Wiefferink & Paulussen (2004) both serve as appropriate theoretical frameworks for this study.
2.1 E-health Implementation as Techno-Change
Technology system implementation models already highlight that technological factors alone are
not enough to accomplish successful adoption of the system. The organisational aspects must
also be addressed (Or et al., 2014). As well as the technological aspects of technology
implementation, change management is an important factor and needs to be taken into
consideration. According to e.g. Markus (1983; 2004), a change in IT needs to be seen as a
techno-change, instead of just a technological project. IT projects focus on improving technical
performance, but techno-change involves great potential impact on users. It is important to be
aware of both technological and organisational change aspects of an IT change. Like other IT
implementations, it is important to align implementation of e-health with organisational change.
This study focused on change management related factors of e-health implementations.
The results are analysed from the perspective of the implementing unit. In this paper, the term implementing unit intends to describe the people who manage the implementation process of an e-health application. These can be external or internal change agents, people from the management of the organisation or employees from the organisation who get the job of organising the implementation. When constructing the conceptual framework for an action- based e-health implementation strategy, we kept the implementing unit in mind. We included factors that can be made actionable for people who are managing the e-health implementation.
Hence, we focused on the action-based results of the two research cycles.
2.2 Technology Adoption in Healthcare Organisations
The implementation of a new system in an organisation is worth nothing for the improvement of on-going care processes without the system being adopted by its users (Crossan & Apaydin, 2010). Without adoption of the new system, the potential benefits of the system will not be met.
The word institutionalisation refers to the adoption process in organisations, and means the routinisation of practices of different kinds in everyday social life (May & Finch, 2009).
Institutionalisation refers to the point where some practice has become largely habitualised (May
& Finch, 2009). Furthermore, implementation is a political concept where the organisational setting and its actors are the antecedents that need to create a fit between the technology and the organisation (Langstrup, 2008). In this way, implementation is a highly purposive and directed set of activities. The process of how these activities become routinely integrated into their social contexts, or not, needs to be explored. This is also the case for mental and elderly healthcare settings; a context in which such business related research is less common (Jie Mein, Guodong &
Agarwal, 2011).
Effective technology adoption requires adaptation of work practices, reorientation, and organisational change, especially in the knowledge-intensive healthcare sector (Mayo-Smith &
Agrawal, 2007). Healthcare organisations have a substantial degree of autonomy, which is difficult to change (Tsiknakis & Kouroubali, 2009). This compared to organisations that typically have strong central control and where IT is usually uniform and centralised, such as banking. In healthcare organisations, strong incidental initiative is allowed or even encouraged. In this way, there is much less central control and a substantial diversity of practice. Subsequently, it is a big challenge to initiate the changes in practices that are necessary for adoption of e-health implementations in the healthcare sector. According to Tsiknakis & Kouroubali (2009), such an implementation requires an in depth analysis of organisational change factors that influence implementation.
2.2.1 Managerial Influences on Individual Adoption
Adoption of innovations by individuals in organisations is always subject to strong
managerial influences (Leonard-Barton & Deschamps, 1988). Regarding the use of innovations in
the work place, individuals rarely have complete autonomy. It is the management who can
encourage adoption of innovations explicitly through orders and communicated preferences
(Moore and Benbasat, 1991) and implicitly through incentives and reward systems (Leonard- Barton, 1987). The management can control the physical access to hardware and software needed to use the innovation. Likewise, the infrastructure that supports adoption, like training and support, can be controlled by them (Leonard-Barton & Deschamps, 1988). Accordingly, the management strongly influences adoption of innovations by employees. In the form of the resulting conceptual framework for an action-based e-health implementation strategy, this study offers a possible recipe for implementing units to influence adoption of e-health by employees in intramural mental and elderly healthcare organisations.
2.3 Domotics in Healthcare
At the case organisation, a large Dutch mental and elderly healthcare provider, we analysed the implementation of one particular e-health application. The case study focused on three locations where intramural care delivery employees work with a so-called sensoring domotics application.
Domotics means ‘automation of houses’ (Lorente, 2004) and is in English generally called smart houses (Allen, 1996). In literature, the domotic application used at the case organisation is known as pervasive sensing technology (Prossegger & Bouchachia, 2014; Hsu et al., 2010). Pervasive sensors are used to collect data related to human physiology and human activity, as well as to the environment. Those data are processed in order to extract cues and patterns about various aspects such as the resident’s profile, health status, living environment, and the resident- environment interaction. The main reason for the rapid development of intelligent sensors like these is the cost-effective usefulness of these smart sensors in mental and elderly healthcare (Benmansour, Bouchachia, & Feham, 2015). However, cost-effectiveness is only the case when domotics are implemented in such a way that it effectuates adoption (Alemdar et al., 2013).
Since the case organisation implemented a domotics application, we intended to focus our systematic literature review on implementation of domotics in mental and elderly healthcare. However, during the first searches of the review process, we noticed a lack of sufficient literature about implementation factors and processes in the specific field of domotics in healthcare settings. This made us decide to broaden the subject of our research. Thus, this study focused on the broader subject of e-health, instead of domotics. Nevertheless, the lessons learned from the e-health literature can be translated to the field of domotics in healthcare, as the resulting conceptual framework is independent of the specific e-health application.
2.4 Categories of Implementation Factors
The introduction of innovations into healthcare organisations is generally recognised as a
complex process (Fleuren, Wiefferink & Paulussen, 2004). It is essential to gain insight into the
factors that may facilitate or hinder introduction of innovations in the healthcare sector. This
insight is necessary for designing an appropriate strategy for introducing innovations. Fleuren et
al. (2004) created a systematic overview of determinants of innovation processes in healthcare
organisations. By determinants, they mean factors that facilitate or impede actual change
(Bartholomew et al., 2001). They identified five categories of determinants: determinants related
to the socio-political context; determinants related to the organisation; determinants related to the adopting user/healthcare professional; determinants related to the innovation; and determinants related to facilities needed to implement the innovation.
We used these five categories to classify factors resulting from the systematic literature review and case study. This is a suitable categorisation for our study results, because an e-health implementation can be seen as the introduction of an innovation in a healthcare setting. The category of determinants related to the adopting user/healthcare professional was further differentiated into ‘factors that can be directly influenced by the implementing unit’ and ‘factors that are inherent to the individual’. This distinction was made to clearly group the factors that can be directly influenced by the implementing unit, which was the focus of this study.
2.5 The Implementation Stage Model of Rosenheck
The implementation stages of Rosenheck (2001) are used to structure the conceptual framework for an action-based implementation strategy. The study of Rosenheck is grounded in healthcare research and starts with the statement that too often clinical practices and service system innovations validated by research are not fully adopted in treatment settings (National Advisory Mental Health Council, 1998). Rosenheck’s model serves as a solution to this research-to-practice gap. He argues that effective dissemination of new treatment methods requires different organisational strategies at different phases of the implementation. He distinguishes four distinct phases on the road from research to practice. These stages are as follows: (1) the decision to implement a new procedure or program, (2) the initial implementation, (3) sustained maintenance over an extended period of time, and (4) termination or transformation of the implementation. Beside these stages, four strategies that can facilitate the introduction of research into real-world organisational settings were identified (Rosenheck, 2001). The four strategies are: constructing decision-making coalitions to support the innovation; linking new initiatives to legitimate organisational goals and values; quantitative monitoring of implementation and performance; and the development of self-sustaining communities of practice as well as learning organisations. These strategies are of central importance across the four stages. We chose to use this implementation model, because this was one of the few models with clear description of the concrete actions to be taken within each step. Furthermore, this model suits our study, because it emphasises the importance of the post-implementation stage as well. Many other models ignore this stage and end with the implementation stage itself. Hence, the combination of the clear action-step descriptions in each stage, the existence of a post- implementation stage and its foundation in the healthcare sector made this model appropriate to serve as an theoretical framework for this study.
The implementation model of Rosenheck (2001) can be seen as a life cycle process (Van
de Ven & Poole, 1995). The model depicts processes of change as moving towards a subsequent
end that is prefigured in the present state, namely the fourth stage of transformation or
determination. However, when applying this model to our results, we noticed that this model can
also be interpreted as a teleological process. In a teleological process, the final cause for the
movement of an entity is a particular goal (Van de Ven & Poole, 1995). In this type of process models, development is a repetitive sequence of goal formulation, implementation, evaluation, and modification, based on what was learned or intended. These characteristics fit in the context of an e-health implementation. As described in the results section, the fourth implementation stage did get a richer meaning than just a termination or transformation of the implementation.
Hence, in this study, we interpreted the stage model of Rosenheck (2001) as a teleological process model.
The stage model of Rosenheck (2001) is used to better specify our initial research question. A refined research question was formulated, which is: ‘Which action-steps need to be taken by the implementing unit in which implementation stage(s) to influence the adoption of intramural e-health applications by employees in mental and elderly healthcare organisations?’.
3. METHOD
This chapter reports the research design and activities of the two research cycles. First, the used method for the systematic literature review is described. The five steps of the literature review process are outlined. Then, the second research cycle is described: the qualitative case study. The case context and data collection and analysis of the case study are presented. This chapter ends with addressing the reliability and validity of this study.
3.1 Research Cycle I: Systematic Literature Review
The first research cycle followed the systematic review methodology prescribed by Khan et al.
(2003). This step-by-step review method is straightforward and appropriate for a literature review in the healthcare sector. The method consists of five steps: (1) framing review questions;
(2) identifying relevant work; (3) assessing the quality of studies; (4) summarising the evidence;
and (5) interpreting the findings (Appendix A). Below we described the way in which the five steps were carried out.
3.1.1 Step I: Framing Review Questions
The first stage contains the composition of a clear, unambiguous and structured research question. After some adjustments, the main research question is set as follows: ‘Which implementation factors influence the adoption of intramural e-health applications by employees in mental and elderly healthcare organisations?’.
3.1.2 Step II: Identifying Relevant Work
Step II is about the identification of relevant work and can be subdivided into two stages: the first stage contains three sub-parts and the second stage contains four sub-parts.
3.1.2.1 Stage I
The first stage can be subdivided into three parts: identifying key words for the syntax;
identifying databases; and setting up inclusion criteria. These elements were specified a priori.
First, key words and search terms are identified, which are built from the literature and from
discussions with two senior researchers and three managers of the case organisation. The used categories were as follows: (a) ‘e-health’ or ‘ehealth’ or ‘sensor technology’ or ‘smart home*’ or
‘domotic*’ or ‘smart health’ or ‘smart-health’ or ‘tele-health’ or ‘telehealth’ and (b) ‘employee*’ or
‘worker*’ or ‘team*’ or ‘work team*’ or ‘user*’ or ‘staff’ or ‘organi$ation’ or ‘professional*’ or
‘business’ and (c) ‘implementation*’ or ‘institutionali$ation’ or ‘adoption*’ or ‘use’ or
‘routini$ation’ or ‘initiation’ or ‘acceptance’ or ‘adaptation’ or ‘infusion’ or ‘normali$ation’. The last category was (d) ‘domotic*’ or ‘elderly care’ or ‘mental healthcare’.
Second, we decided to use Web of Science and EBSCO host COMPLETE as primary search engines for subtracting literature. By using EBSCOhost COMPLETE, we were sure that we included healthcare databases, such as AMED, MEDLINE and CINAHL. This database also includes the most used business research database Business Source Premier, which is useful since our research was business-related. Likewise, we chose to use the Web of Science core collection to include 12.000 journals about science & technologies and social sciences.
Third, the articles to be included in this study had to meet the following selection criteria: (1) the papers must be written in English; (2) the study subject is related to work processes; (3) the study is about intramural care; (4) the concerned IT application falls into the category of e-health; (5) the study focuses on the implementation process of the e-health application; and (6) the study concerns adoption or use by healthcare employees. In Appendix B, these selection criteria are specified more comprehensively.
3.1.2.2 Stage II
The second stage of step II was about conducting the review process. This stage is subdivided into four parts: using the syntax in both search engines; screening the abstracts of the selected articles and including the articles that met the inclusion criteria; scanning the included articles and including the articles that still met the inclusion criteria; and adding relevant Dutch literature to the resulting list of articles. Now, we elaborate on these steps in more detail. First, in EBSCOhost COMPLETE, the syntax was used within the field ‘abstract’, which resulted in 66 studies. For Web of Science, the syntax was used in combination with the field tag ‘topic’ and resulted in 256 studies. Hence, we started our search with a list of 322 selected articles.
Second, the abstracts of the 322 articles were screened, which resulted in the inclusion of 27 articles of the Web of Science selection and 9 of the EBSCOhost COMPLETE selection. In Figure 2, the exclusion flowchart can be found. In addition to the exclusion flowchart, we now express the main reasons for exclusion. Most often, studies were excluded because the abstracts did not meet inclusion criteria two, three and six. 91 abstracts were excluded because the focus was not on work processes or business-related outcomes, but instead on client-related outcomes.
47 abstracts were excluded because the studies did not focus on intramural care, but on homecare. 63 abstracts were excluded because the studies focused on the client as the user of the e-health application.
Third, after selecting initially relevant papers by reading abstracts, the 36 articles were
scanned. Primarily, three duplicates of both databases were recognised and excluded. A sample
of 33 articles remained. Then, 24 articles were excluded based on the non-relevance of the study.
Here, most studies were excluded because the research did not meet inclusion criteria four and five. The exclusion was seven times based on the fact that articles did not focus on work process or business-related outcomes. Also, eight articles were excluded because the studies were not about the implementation process or adoption stage of e-health applications. The systematic literature review ended with thirteen included articles.
After the exclusion process, additional papers were searched in Dutch literature. One relevant research conducted by Nictiz & NIVEL (Krijgsman et al., 2016), an independent centre of expertise for standardisation and e-health, was found.
Figure 2. The exclusion flowchart.
This study focused on factors that facilitate healthcare professionals and healthcare managers to stimulate e-health adoption. After adding this study, the selection process ended with a total of fourteen included articles.
Generally, when conducting a systematic literature review, the process of excluding studies is done by more than one reviewer (Tranfield et al., 2003). In this more compact review, the head author of this paper made the main decisions. However, a senior researcher checked the doubtful exclusions.
3.1.3 Step III: Assessing the Quality of Studies
This research step is about the quality assessment of the studies. Considering that we noticed we
226 articles excluded based on
selection criteria 56 articles excluded based on
selection criteria
3 duplicates excluded
24 articles excluded based on selection criteria
Added 1 Dutch research:
‘E-health Monitor 2016’
Web of Science:
256 potentially relevant articles
27 articles included for content search
EBSCOhost Complete:
66 potentially relevant articles
9 articles included for content search
Total of 36 potentially relevant articles
Total of 33 articles included in the content
search
Total of 13 articles included in the study
Total of 14 articles included in the final
anaysis
were searching for quite specific literature and it turned out that our sample consisted of only fourteen articles, we decided to include all papers that seemed relevant based on content. Also, we wanted to include, for example, small-scale studies, because the implementation factors in these more local-oriented studies can be valuable too. Hence, we did not do a specific quality check. However, the content of the articles was evaluated by the face value of the implementation factors mentioned in the studies. Implementation factors were evaluated as valuable when they seemed manageable by the implementing unit. Appendix B further elaborates on this assessment.
3.1.4 Step IV: Summarising the Evidence
This step is about synthesising the data in order to derive insights from the reviewed literature (Khan et al., 2003). Overarching themes and patterns in the dataset become visible by this data synthesis (Tranfield et al., 2003). This step starts with constructing a table of the study characteristics of the included articles (Appendix C). Then, the fourteen articles were thoroughly read. All factors that, according to the included studies, have a facilitating or restraining influence on the adoption of e-health applications by employees in intramural healthcare settings were extracted (see Appendix B for the inclusion assessment of the implementation factors). The product of this step was a list of 33 facilitating implementation factors, which can be found in Table 1. The articles in which the concerning determinants were found are also displayed in this table.
3.1.5 Step V: Interpreting the Findings
The categorisation of Fleuren et al. (2004) was used to group the 33 implementation factors. In Appendix D, the codebook used for this categorisation is presented. In order to further interpret the findings of the systematic literature review, the study was enriched with a qualitative case study. In the next sub-section, the method used in the case study and the further interpretation of the findings of the systematic literature review are specified.
3.2 Research Cycle II: Qualitative Case Study
As Eisenhardt (1989) states, a case study is a research approach that focuses on understanding the dynamics present within a single setting. To examine the phenomenon of an e-health implementation in its real-life context (Yin, 1981), the output of the systematic literature review was translated to the situation of a case organisation. In order to enrich the results of the systematic literature review with context-specific knowledge, the initial research question was refined. The case study attempted to answer the research question: ‘Which action-steps need to be taken by the implementing unit in which implementation stage(s) to influence the adoption of intramural e-health applications by employees in mental and elderly healthcare organisations?’.
Through the answers of the refined research question, the research goal of creating an action- based e-health implementation strategy was achieved.
In order to understand the dynamics of e-health implementation, and more specifically
the way these impact adoption by healthcare employees, I conducted nine interviews at three
locations of the case organisation.
3.2.1 Research Context
At a large Dutch mental and elderly healthcare provider, the implementation adoption of one particular e-health application was analysed. This study focused on three intramural locations where care employees work with a pervasive sensing application, which is known as the SOS: the Smart Optical Sensor (J. Waalkens, private mail, May 24 2016). At the moment of conducting this study, a total of 71 sensors had already been in use for more than a year, distributed over the three locations. The settings of the sensors are personally adjustable for every client. Possible sensor settings for the SOS are the detection of restlessness in bed, leaving the bed, leaving the room, and the client calling for help. When the sensor detects one of those events, the care deliverer receives an alarm on his or her SOS-handset. In reaction to this alarm, the carer knows where he or she is needed and can take a specific action to monitor or help the client. With the use of the SOS, caregivers do not need to make anticipative checks every hour when the clients are in their own rooms, like at night-time. When using the SOS, they only have to respond to the alarms on their SOS-handset, which increases clients’ privacy and autonomy.
By conducting interviews at three locations of the case organisation, we investigated in which way the 33 implementation factors found by the systematic literature review had been experienced to contribute to the adoption process of the SOS.
3.2.2 Interviews
During two weeks, I conducted nine interviews at three locations of the case organisation. Each interview took about one hour. The interviewees were asked for permission to record the interviews with QuickTime Player, where upon all participants agreed. The interviewees were employees working at one of the three locations of the organisation. They were already working at these locations during the implementation of the SOS. When selecting the interviewees, their positions in the organisation and their role during the implementation were taken into account.
Therefore, we ended with a diverse sample of interviewees. For instance, I spoke with people who had an additional role during the implementation and with people who did not have an additional role.
The 33 implementation factors found in the fourteen articles functioned as input for the
interview template (Appendix E). In this template, the implementation factors were coupled to a
5-point Likert scale (Likert, 1932). By using this scale, the interview participants had to focus
their assessment on each of the 33 items. We were mainly interested in ‘how’ and ‘why’ the
implementation factors were influential in the adoption of the e-health application. The Likert
scale was used as probe to gain this information. For each factor, the employees were asked to
which extent the factor was present at the implementation of the SOS. By asking this, we were
able to make an exhaustive evaluation of the implementation process of the SOS. This was
valuable information for the management of the case organisation. Employees were also asked
which implementation actions were accompanied by the concerning implementation factors. For
instance, they were asked how the user training looked like. They got questions about which
actions worked, which actions did not work and why they did not work. I also asked the
interviewees about ways to improve the implementation actions. For example, they were asked
how employees could be trained best. This ‘how’-type of questions was most important in this study, since we wanted to create an action-based e-health implementation strategy.
3.2.3 Data Analysis
All recorded interviews were transcribed in detail by using ExpressScribe. The interview transcripts were coded in ATLAS.ti, using deductive and inductive coding strategies (Miles and Huberman, 1994). The product of the data analysis was twofold. First, an evaluation and recommendations report for each of the three locations of the case organisation was created.
Second, a conceptual framework for an action-based implementation strategy was built. Below, the method used to come to both products is reported.
3.2.3.1 Evaluation and Recommendations Reports
For each of the three locations of the case organisation, an evaluation and recommendations report was written (Appendix F shows one of the three reports). In each report, a description of the current implementation situation and recommendations for what can be improved concerning the implementation was given. These recommendations can be used in future e-health implementations. The reports are additional to this study and do not have a function in our main research.
The reports are divided into three stages: the pre-implementations stage, the implementation stage and the post-implementation stage. For each stage, we first gave an overview of the situation. To evaluate the situation of the SOS-implementation against the factors found in the literature, we used deductive codes (Miles and Huberman, 1994). These codes were derived from the output of our systematic literature review. So we used each of the 33 implementation factors as a deductive code. Then, for each of the three stages, recommendations for a subsequent e-health implementation were given. For these recommendations, inductive codes (Miles and Huberman, 1994) were developed out of the action-based ideas suggested by the interviewees. These are suggestions for carrying out an e-health implementation in order to reach adoption. For example, when reading the transcripts, it became clear that practical training before implementing the SOS was needed to provide appropriate use of the application. This led to the inductive code practical pre-implementation training (See appendix G for the complete list of inductive codes).
3.2.3.2 The Conceptual Framework for an Action-Based E-health Implementation
Strategy
The second part of the data analysis was constructing a conceptual framework for an
action-based e-health implementation strategy. Through the development of inductive codes as
mentioned above, categories of implementation actions were developed. Now, we will further
specify the development process of the inductive codes. When analysing the transcripts of the
case study to develop inductive codes, we focused on three specific research criteria. First, we
focused on action-based implementation factors. These were factors that could be coupled to
specific implementation actions. Second, the study focused on factors that can be influenced by
the implementing unit. Third, we focused on the implementation process, and not on the factors
concerning the e-health application itself. Subsequently, application specific implementation
factors were not included in the conceptual framework. However, if implementing units are interested in recommendations about implementation factors concerning the application itself, they can take the implementation factors found by our systematic literature review into consideration (Table 1, row 4).
As Thomas (2006) states, “The outcome of an inductive analysis is the development of categories into a model or framework that summarises the raw data and conveys key themes and processes” (p. 240). The categorisation of the inductive codes led to 26 implementation action- steps. These action-steps are the core of the resulting action-based e-health implementation strategy. The implementation categories of Fleuren et al. (2004) are used to group the 26 implementation action-steps. Simultaneously, the four implementation stages of Rosenheck (2001) function as a foundation for the conceptual framework for this e-health implementation strategy. The action-steps derived from the case study are divided into these four implementation stages (Appendix H). In the next section, we further elaborate on the reliability of this study.
3.3 Reliability and Validity of the Study
In this study, reliability is covered in two different ways. First, as Armstrong et al. (1997) state, there is a notion of reliability within the call for transparency of technique. “By keeping all collected data in well-organised, retrievable form, researchers can make them available easily if the findings are challenged or if another researcher wants to reanalyse the data” (Marshall &
Rossman, 1989, p. 146). In this study, all research steps in the systematic literature review and case study are exhaustively documented. For the systematic literature review, the reasons for exclusion or inclusion of the 322 studies were documented. In the case study process, the interviews were recorded and transcribed in detail and the added inductive codes are clearly displayed. Also, for all types of selection or classification in this study, codebooks were created (see Appendix B, D, & H). By this transparent documentation, other researchers are better able to replicate the study.
Beside the transparency of technique, this study meets inter-rater reliability. As Mays &
Pope (1995) state, by organising an independent assessment of qualitative data by additional skilled qualitative researchers and comparing agreement between the raters, the reliable analysis of qualitative data can be enhanced. In the systematic literature review process, the head author, two senior researchers and three managers of the case organisation checked the syntax.
Also, one senior researcher verified the doubtful exclusion cases. The head author, one senior researcher and three managers of the case organisation checked the interview template.
Likewise, in the case study process, two senior researchers were involved in the classification of the resulted action-steps. The division of the action-steps into the four implementation stages of Rosenheck (2001) was independently done by the head author and two senior researchers. In this way, inter-rater reliability was met.
Next to these methods to cover reliability, we used the validity procedure of
triangulation, “were researchers search for convergence among multiple and different sources of
information to form themes or categories in a study” (Creswell & Miller, 2000, p. 126). Hence, to improve validity and reliability of findings (Golafshani, 2003), this study used multiple methods of data collection and analysis.
4. RESULTS
The results of the systematic literature review and qualitative case study are presented in this chapter. In the first section of this chapter, the descriptive results are described. Here, we first present a sub-section with findings of the systematic literature review. A list of 33 implementation factors that facilitate the adoption process of an e-health implementation in intramural mental and elderly healthcare is depicted. In the next sub-section, findings of the case study in combination with the systematic literature review are reported. By means of pattern coding, five implementation factors that are evaluated as most important by our interview participants were found. These factors were (1) the need for user involvement; (2) the need for theoretical and practical training; (3) the need for super users; (4) the need for evaluation; and (5) the need for post-implementation user support. In this sub-section, the five factors are described by the output of both the two research cycles: the systematic literature review and the case study. First, for each factor, an overview of the literature found in the sample of our literature review was given. Second, related findings of the case study are presented.
After the descriptive results, the second section of this chapter describes the interpretive results. In this section, the conceptual framework for an action-based e-health implementation strategy is presented. The implementation stages of Rosenheck (2001) and the categories of Fleuren et al. (2004) are used to structure the conceptual framework and to give meaning to the findings in relation to the extant literature.
4.1 Descriptive Results: Systematic Literature Review
This part of the results section views the descriptive analysis of the systematic literature review.
The outcomes of the systematic literature review are presented in Table 1. This table depicts a
list of 33 implementations factors that can facilitate adoption of e-health implementations in
intramural mental and elderly healthcare. These factors give an overview of the current state of
the literature about adoption of e-health implementations in the intramural mental and elderly
healthcare sector. The classification of Fleuren et al. (2004) was used to categorise the
implementation factors. We do not dwell upon the 33 implementation factors too long, whereas
we immediately refined and contextualised them in a subsequent case study. The factors were
used as input for the case study. In the next section, the five most influential implementation
factors, according to the interviewees of the case study, are described.
Table 1. List of 33 implementation factors necessary for e-health adoption in intramural mental and elderly healthcare settings.
1. Determinants related to the socio-political
context 1.3 The e-health application has a positive impact on the interaction with clients 9
2.1 The workflows are redesigned 54, 73, 151, E53, D1
2.2 Technical experts are involved during the implementation of the e-health
application 132, E3
2.4 The technical and social consequences of the e-health application are
investigated before the start of the implementation 132, E53,
2.5 There is a collaboration between technical staff and clinical users during the
implementation 65, E50,
2.6 The e-health application matches with the goals of the organization 9
Factors that can be directly influenced by the implementing unit
3.1 There is training in the use of the e-health application 9, 65, 102, 132, 157, E3, E50, E53, D1 3.2 The management encourages de use of the e-health application 9, 102, 251 , E3, D1
3.3 There is timely user support when this is needed 54, 251, D1
3.4 There is no pressure on the employees to use the system (voluntariness) 102, 251,
3.5 It is possible for users to make local adaptations in the use of the e-health
application 73
3.6 The rationale for the e-health implementation is quality improvement (and not
cost cutting) 102
3.7 The e-health application is compatible with the values, culture and practices of
the users 251
Factors inherent to the individual
3.8 Users' percieve the e-health application as having relative advantages compared
to the situation it replaces * 9, 65, 251, E3, E53,
3.9 Adoption depends on the phenomenon of social influence 9, 251, 3.10 Users need to gain acceptance of the new e-health application 65, 157, 3.11 Users need to make a shift in attitude towards the work processes 65
3.12 Users need to be aware of the e-health implementation 102
2. Determinants related to the organization
Factor included in article:
Category of determinants (Fleuren et al., 2004) Factor found by the systematic literature review (Assen, 2017)
3. Determinants related to the adopting user (health professional)
#
Table 1 (continued). List of 33 implementation factors necessary for e-health adoption in intramural mental and elderly healthcare settings.
4.2 Descriptive results: Case Study x Systematic Literature Review
In this section, the five most influential implementation factors, as resulted from the case study, are described. For each of the five factors, an overview of the insights found in the articles of the systematic literature review is given. Also, related findings of the case study are described. Note that the suggestions put forward in the section of the case study cycle are the voices of the interviewees. The sequence of describing the five implementation factors is based on their occurrence in the implementation process, as can be seen in the conceptual framework that will be presented in the interpretive results section.
4.1 The e-health application is interoperable with the existing ICT infrastructure 132, 151, E50, E53, D1 4.2 The e-health application is flexible and responsive to patients’ needs ** 54, 102, D1 4.3 The use of the e-health application is reliable and safe 102, 157, D1 4.4 The e-health application is visible in the organisation 251, D1 4.5 It is possible to try the application before it goes live (Trialability) 251, D1 4.6 In the e-health application, the privacy of the clients is protected E53, D1
4.7 The e-health application is a supplement; it complements and does not replace
face-to-face service ** 102
4.8 The quality and availability of the information produced by the e-health
application is good 251
4.9 The e-health application is not complex in its use 251
4.10 There is a protocol which can be followed by the users in case of
problems/defects 132
5.1 There is made use of super users 23, 54, 132, 157, 73, D1
5.2 Clinicians, staff and other users are involved in the development and planning of
the implementation 65, 132, 251, E3, D1
5.3 There is a technology & community e-platform 23, 157, E3,
5.4 There is an evaluation of the e-health implementation 23, E3
5.5 The implementation approach is iterative, rather than in sequence E50
* This determinant can also be classified as a characteristic of the innovation
** This determinant can also be classified as a characteristic of the socio-political context 5. Determinants related to the facilities needed
to implement the innovation
Category of determinants (Fleuren et al., 2004) # Factor found by the systematic literature review (Assen, 2017) Factor included in article:
4. Determinants related to the innovation
