Electronic health records adoption in South African Healthcare Providers : a case of North West Province

iJNwu

Electronic Health Records Adoption

in South African Healthcare Providers: A

Case of North West Province

TK Modise

G

orcid.org/ 0000-0001-8743-1724

Dissertation submitted in fulfilment of the requirements for the

degree

Master of Commerce in Computer Science and Information

Systems

at the

North West University

Supervisor:

Co-supervisor:

Dr. M.E. Jantjies

Prof N Mavetera

Graduation ceremony: October 2019

Student number: 23255897

2020 -01- 0 8

ABSTRACT

Electronic Health Records (EHRs) are a technology that is being implemented in healthcare institutions around the globe. This technology enables the healthcare sector to enjoy increased efficiency and throughput, whilst cutting overhead costs in healthcare centres. The aim of this study is to investigate the adoption of these EHRs by identifying the factors that influence their rate of adoption within healthcare institutions. Diffusion of Innovation (Doi) Theory and a derivative of the Technology Acceptance Model, (TAM2) are used as the theoretical lenses through which this problem was viewed.

Respondents participated in a mixed methods study that evaluated the effect which the factors identified by the above theories, had on the adoption of EHRs in their workplace. The sample for this study comprised of public and private primary healthcare facilities from the North West Province. Three participants from 3 different healthcare facilities participated in the qualitative iteration of the study, while 56 participants took part in the quantitative iteration of the study. Interviews and an electronic questionnaire were utilized to collect data required for analysis in this study. Data collected was analysed using Descriptive Statistics and Correlation Analysis.

Results showed support for some of the factors of TAM2 and Doi, namely Relative Advantage, Output Quality, Result Demonstrability, Computer Self-Efficacy, System Complexity and Enjoyment/Job Satisfaction. A new variable - Patient Safety Endangerment - was also found to have a significant influence on the healthcare worker's decision to use a particular EHR. It is with this information that a deeper understanding of how EHRs are used in the North West Province can be established and this information can be used by decision makers when implementing similar systems within the province to maximise their adoption.

Keywords

Electronic Health Records; Technology Acceptance Model; Diffusion of Innovations Theory; National Health Insurance; South African Healthcare.

TABLE OF

CO

NTENTS

ABSTRACT .............................................................. 1

ACKNOWLEDGEMENTS ................................. IX CHAPTER 1 OVERVIEW OF THE STUDY ....................... 1

1.1. 1.2. 1.3. 1.3.1. 1.3.2. 1.4. 1.5. 1.6. 1.7. 1.8. 1.9. 1.10. 1.11. 1.11.1. 1.11.2. 1.11.3. 1.12. 1.12.1. 1.13. 1.14. Background of the study ........................ 1

Problem statement .................... 2

Brief Literature Review & Theoretical Perspective ............ 3

Comparative literature ... 3

Theoretical Literature ... 3

Importance of the study ....... 4

Research Design ............................... 5

Research Questions ...... 6

Major research questions ... 6

Research Aim and Objectives ...... 6

Major Objectives ...... 6

Research Hypothesis ............................ 6

Ethical Considerations ...... 7

Quality & Integrity and Independence & lmpartialness ... 7

Consent, Voluntary Participation & Participant Safety ... 7

Confidentiality and Anonymity ... 7

Research Scope and Delimitation ......... 8

Use of theory in predicting adoption ... 8

Summary ... 8

Provisional Chapter Outline ... 8

CHAPTER 2 LITERATURE REVIEW ...... 10

2.1. Electronic Health Records (EHRs) ......... 10

2.2.1. 2.2.2. 2.3. 2.3.1. 2.3.2. 2.3.3. 2.3.4. 2.3.5. 2.3.5.1. 2.3.5.2. 2.5.2.3. 2.5.2.4. 2.5.2.5. 2.5.2.6. 2.3.6. 2.4. 2.5. 2.6. 2.7. EHR Implementations Around the World ... 11

EHR Implementations in South Africa ... 13

The concept of User Acceptance .......................... 16

The Theory of Reasoned Action (TRA) ... 17

The Technology Acceptance Model (TAM) ... 19

Theory of Planned Behaviour (TPB) ... 21

TAM Extensions ... 27

Diffusion of Innovations (Doi) ... 22

Explaining Rate of Adoption - A Doi Approach ... 24

Perceived Attributes of Innovation ... 24

Type of Innovation-Decision ... 25

Communication Channels ... 26

Nature of the Social System ... 26

Extent of Change Agents' Promotion Efforts ... 26

EHR Implementation barriers ... 27

EHR implementation .................................. 27

Choice of Theoretical Model ...... 32

Conceptual Model ...... 33

Summary ... 35

CHAPTER 3 RESEARCH DESIGN & METHODOLOGY ......... 36

3.1. Research Paradigm ....... 36

3.2. Research Methodology ......... 37

3.2.1. Qualitative Research Methodology ... 37

3.2.1.1. 3.2.2. 3.2.2.1. 3.2.3. 3.3. 3.3.1. 3.3.2. 3.3.3. 3.3.3.1. 3.3.4. 3.3.4.1. 3.4. 3.4.1. 3.4.2. 3.4.3. 3.5. 3.6.

Characteristics of Qualitative Research ... 37

Quantitative Research Methodology ... 37

Characteristics of Quantitative Research ... 38

Research Strategy Used ... 38

Data Collection and Analysis Methods ......... 39

Interviews ... 39

Questionnaire ... 40

Qualitative Data Analysis Methods ... 40

Thematic Analysis of Data ... 40

Quantitative Data Analysis Methods ... 41

Multivariate Correlational Analysis ... 41

Population and Sampling .................................. 41

Sample vs. Population ... 41

Population ... 42

Sampling ... 42

Trustworthiness and Authenticity ................ 42

Summary .................. 43

CHAPTER 4 RESEARCH FINDINGS AND ANALYSIS ................ 44

4.1. Research Findings ............ 44

4.1.1. Research Participants ... 44

4.2. Qualitative Analysis -Interview Findings ................... 44

4.3. Factors from the qualitative phase ...... 54

4.3.1. Importance of Patient Safety ... 55

4.3.2. Family, Co-workers, and Supervisor Influence ... 56

4.3.3. Incomplete Functionality, Electricity and Network Outages ... 56

4.3.4. Change Management. ... 56

4.4. Quantitative Analysis - Questionnaire Results ........... 56

4.4.1. Gender ... 57

4.4.2. Age ... 58

4.4.3. Level of Education ... 58

4.4.4. Instrument Reliability ... 59 4.4.5. Descriptive Statistics for all Measures ... 61

4.4.6. Conceptual Model Testing ... 67

4.5. Summary ............................ 70

CHAPTER 5 FINDINGS DISCUSSION ... 72

5.1. 5.2. 5.2.1. 5.2.2. -- - ,

i·

>-

.

~~

l

5

.

3

.

1

.

3:a:

Zea

:r

Experience with computers ... 72

Answers to Research Questions ........................... 73

Which factors influence adoption of EH Rs in the North West Province? ... 73

Conceptual Model Attributes ... 73

Revised Conceptual model ... 78

How can the adoption of EH Rs in South African healthcare facilities be improved? ... 79

How are EH Rs used in healthcare facilities that have implemented them? ... 79

What can be done to improve EHR adoption in South Africa? ... 80

Future Research .................... 83.

Limitations ......................... 83

Conclusion ..................................................... 84

REFERENCES ... 85 APPENDIX A ... 98 APPENDIX B ... 100 APPENDIX C ... 101 APPENDIX D ... 103 APPENDIX E ... 104 APPENDIX F ... 109 vi

LIST OF FIGURES

Figure 2-1: Summary of common modules within an EHR ... 12

Figure 2-2: Theory of Reasoned Action. Adopted from Fishbein & Ajzen (1975) ... 18

Figure 2-3: The original Technology Acceptance Model adopted from Davis (1986) ... 20

Figure 2-4: The revised Technology Acceptance Model adopted from Davis (1989) ... 21 Figure 2-5: The Theory of Planned Behaviour adopted from Ajzen (1991) ... 22

Figure 2-6: TAM 2 adopted from Venkatesh and Davis (2000) ... 29

Figure 2-7: Extended TAM to include determinants for perceived ease of use. Adopted from Venkatesh and Davis (2000) ... 30

Figure 2-8: Diffusion of Innovations Conceptual Model. Adopted from Rogers (1995) ... 23

Figure 2-9: Variables determining the rate of adoption of innovations. Adopted from Rogers (1995) ... 24

Figure 2-10: Conceptual model of the study ... 34

Figure 3-1: Sequential Exploratory Research Design adopted from Creswell, Plano Clark, Gutmann, & Hanson (2003) ... 39

Figure 4-1: Conceptual model without Image attribute ... 55

Figure 4-2: Gender of respondents ... 57

Figure 4-3: Age Group ... 58

Figure 4-4: Level of Education ... 58

Figure 4-5: Previous working experience with computers ... 59

Figure 5-1: Revised Conceptual Model ... 79

LIST OF TABLES

Table 2-1: Health information systems per province adopted from Department of Health (2012) ... 13

Table 4-1: Summarized Participant Responses ... 45

Table 4-2: Identified Salient Beliefs ... 54

Table 4-3: Gender frequency table ... 57

Table 4-4: Age group frequency table ... 58

Table 4-5: Education level frequency table ... 59

Table 4-6: Instrument Reliability ... 60

Table 4-7: Descriptive Statistics for Innovation Adoption Attributes ... 62

Table 4-8: Hypothesis Correlation Values ... 68

Table 4-9: Correlation Values between Actual System Use and another factor ... 70

Table 5-1: Innovation Attributes to Main Attributes mapping ... 73

ACKNOWLEDGEMENTS

I would like to express great gratitude to my academic supervisors, Dr. M.E. Jantjies and Prof N Mavetera

who through the years of undertaking this research work, have constantly supported and encouraged me

to pursue its completion.

To my late sister, Kenanao Shirley Madise, I dedicate this body of work to you, for I know you would have

been proud. I would also like to extend a special thank you to Patience Mavetera, as well my family for their unwavering love and support throughout this academic journey.

Lastly a big thank you to the healthcare facilities that took part in this study along with the North West Provincial Health Department, all of whom this research study would not have been successful without their participation.

LIST OF ABBREVIATIONS

Council for Scientific and Industrial Research Diffusion of Innovations

Department of Health Electronic Health Record Health Information System

Information Communication Technology Technology Acceptance Model

Theory of Planned Behaviour Theory of Reasoned Action Perceived Usefulness Perceived Ease of Use Subjective Norm Behavioural Intention

CHAPTER 1 OVERVIEW OF THE STUDY

This chapter provides an overview of the study being discussed. The chapter contains, amongst other topics, a description of the problem statement which this study will solve, the aim and objectives of the study as well as the hypotheses tested by said study. A brief literature review is provided to introduce the topic of Electronic Health Records (EHRs) and the concept of technology adoption. The chapter closes with a breakdown of how this study is structured.

1.1. Background of the study

According to the Department of Health (Department of Health, 2012), e-Health is defined as the utilisation of technology to facilitate healthcare activities. In the foreword of the Department of Health's e-Health strategy document, the Minister of Health mentioned that the development of an e-Health system would be the enabling factor of the National Health Insurance project that the nation has embarked on (Department of Health, 2012). E-Health itself, covers a vast domain of Information Communication Technologies (ICTs) that are used to promote, support and strengthen healthcare. EHRs form a subset of e-Health technologies, which in essence, seek to enhance the delivery of medical care to patients. According to Abayomi-Alli et al (2014:22), an EHR is a "A [medical] record that is theoretically capable of being shared across different healthcare settings and includes a range of data like demographics, medical history, medication and allergies, immunization status, laboratory test results, radiology images, vital signs, personal statistics like age and weight, and more". Health Information Systems (HIS) tend to go beyond the scope of a health record and include other areas of managing the healthcare institution's daily operations. These include but are not limited to things like billing of patients, hospital catering and inventory

& purchasing.

The use of EHRs in healthcare practices provides many benefits to healthcare personnel, organizations as well as governmental organizations in comparison to paper-based records. Some benefits of technology can be found in the medical record itself: "increased legibility and comprehensiveness and easier access to information", to name just three. (Atherton, 2011: 186) Although the use of EH Rs brings a lot of benefits with it, healthcare facilities in South Africa and in many other African countries, are still to leverage the technology to its fullest extent.

Documentation from the Department of Health (2012) and Kleynhans (2011) shows that South Africa embarked on a project to implement a national EHR back in the year 2002. This project kicked off with a workshop that was aimed at highlighting the current state of EHRs within the country, together with the challenges that could be faced when establishing a national EHR. The workshop helped identify goals that the EHR project had to achieve. Those goals were as follows:

• To integrate health record systems in the country by bringing together all the different health information systems facilitating access to health records within a province and across provinces. • To develop a population healthcare knowledge base.

• To improve governance, planning administration and management of health systems at both national and provincial level.

• To improve the efficiency of health service delivery through both personal care and public health

services.

• To enable national monitoring and evaluation of health trends within the country.

• Achieve comprehensive privacy and confidentiality requirements protecting the citizens.

What can be taken away from the report is that, health systems within the country, exist as "islands" of

information. That is, each system only contains patient information for the healthcare facilities in its area.

This data makes it hard to construct a comprehensive profile of a patient's complete medical history, as

their information is stored in a doctor's filing cabinet or system that might be in another province/area. This

disparate data is one of the many reasons why said technology is not being utilized to its fullest potential, as systems from two different vendors most likely will not be able to communicate with each other very easily.

1.2. Problem statement

The challenge of adoption is experienced on a daily basis, be it the clothes we wear to work and school or

the applications we choose to install & use on our smartphones. There has been a significant proliferation

of ICT throughout the 21st Century (Caspary and Boothe, 2015). This proliferation of technology has

brought about a plethora of technologies unleashed into the market for consumption. This makes the

choice to use certain software over others, harder as there are so many options to choose from. Although

some of these technologies and software applications go on to do well in their respective markets, fall

away due to underutilization.

Technology is only effective when its intended users actually make use of it and can exploit the benefits

that it brings. Certain factors prohibit individuals from using a particular technology, even if that technology

is beneficial to them. In the instance of the smartphone, factors such as the individual's socio-economic

status i.e. whether they can afford the cell phone or not - could qualify as one of these factors.

Beforementioned factors which contribute towards the adoption of these ICT technologies are often one

component that is overlooked when developing a new technology or software application.

Considering that the South African Department of Health (DoH) has outlined plans to implement a national

health insurance fund which also aims to implement a new eHealth system, the focus of this study is to

assess the adoption of the current health recording systems, in order to provide input into the oncoming

system.

1.3. Brief Literature Review & Theoretical Perspective 1.3.1. Comparative literature

The slow adoption of EHRs has prompted researchers to try to determine, if ever, worldwide acceptance of this technology will occur. The use of theoretical models to try and explain the adoption or usage of technology is a popular route for researchers when attempting to clarify whether a given technology will be used as it should or is being used at all.

Ford et al (2006) reviewed historic literature regarding the adoption rates of EHRs among physicians in the United States. It was with this data that he predicted the adoption of EHRs using the Technology Diffusion Model. The result was that under the conditions at that time, EHR adoption would reach its maximum market share in 2024 in the small practice setting.

Nevhutalu (2013) investigated the current processes of patient record keeping, management and the referral process of patients within three public hospitals in the Limpopo province with the aim of using the findings to establish an EHR framework to facilitate the e-referral process. Nevhutalu found that all three public hospitals used a health system (MEDICOM) to record patient data, although this system required

that some information be hand written and stored in filing cabinets.

Studies by Kleynhans (2011) & Weeks (2014) looked at barriers that impeded the implementation/adoption

of EH Rs within South Africa. The implementation barriers that were identified by Kleynhans (2011) included amongst others, the costs of implementing an EHR. Findings of this nature are more commonly faced by the organization as a whole and are not the responsibility of the individual using the system.

Weeks (2014:142) found that "paper-persistence" amongst the older generation of respondents was one

of the hindrances in the adoption of electronic means of recording patient healthcare data in their case

study.

In another study, Wilkens (2009) aimed to discover the factors that influenced the adoption of electronic health records by 94 hospital managers in Arkansas, USA. The quantitative study used TAM constructs, to evaluate the hospital manager's intention to adopting EHR technology in their facility. The results

showed that respondents that had adopted EHR components felt that EHRs would be beneficial to their

work.

1.3.2. Theoretical Literature

Theoretical lenses will be used in this study to identify the factors that influence an individual to use EHR

systems in a South African Healthcare setting. This study will employ constructs from two adoption theories to establish a conceptual model that identifies the influences on EHR technology usage. Initially TAM and the Theory of Planned Behaviour (TPB) are identified as the theories to be used as the theoretical lenses used to view the problem under investigation.

TPB and TAM are both revisions of by Ajzen & Fishbein (1980). According to TRA, the intention of an individual to perform a certain behaviour can be attributed to their attitude towards performing that

behaviour and subjective norms Ajzen & Fishbein, (1980). As opposed to TAM, TPB included an extra construct namely Perceived Behavioural Control (PBC) which is based on two beliefs: control beliefs and their perceived facilitation. TAM alludes to perceived ease of use as opposed to subjective norms which is like perceived behavioural control in the sense that these two attributes determine the relative difficulty an individual would face when trying to perform a behaviour.

According to Chuttur, (2009), TPB provides more details that can be used to explain the intention of participants to use a computer application when compared to TAM. A study by Mathieson, (1991) revealed that the construct subjective norms allowed for the TPB model to identify significant groups whose opinions about a technology were important to the user and hence influenced the user positively or negatively to use the technology. The added predictive power of TPB has also been linked to the addition of the construct PBC. This construct is what allows for the identification of barriers that impede system use such as limitations of user's skills. It is these two constructs that allow TPB to overcome the disadvantages that TAM faces as mentioned by Levine & Pauls (1996) and Levine, Little & Mills (1997).

TAM2, a derivative of TAM identifies several variables that increase the predictive power of TAM. TAM2 takes into consideration the individual as well their organizational context. TAM2 is thus selected to be the theoretical lens through which adoption of EH Rs will be viewed because of this added predictive power. TAM2 variables are categorised into Perceived Usefulness (PU) and Perceived Ease of Use (PEOU) Venkatesh & Davis, (2000). Determinants of PU include Subjective Norm, Image, Job Relevance, Output Quality, and Result Demonstrability. PEOU is influenced by anchor variables (Computer Self-Efficacy, Perceptions of External Control, Computer Anxiety, and Computer Playfulness) and adjustment variables (Perceived Enjoyment and Objective Usability). Experience and Voluntariness act as mediators of Behavioural Intention and Subjective Norms. Descriptions of TAM2 variables will be provided in Chapter 2.

Doi also provides an account of user adoption and paints a picture of the entire life cycle of the decision process that goes into adopting an innovation. The theory identifies five (5) innovation attributes, namely: Relative Advantage, Compatibility, Complexity, Trialability, and Observability that, according to the theory have an influence on whether an innovation is adopted and how quickly its adoption will spread. These attributes from Doi and TAM2 have all been brought together to form a conceptual model that tries to explain adoption of EHRs from an individual, organizational and innovation perspective. This conceptual model derived from these adoption models will be discussed in greater detail in Chapter 2.

1.4. Importance of the study

ICT in healthcare plays a significant role in making healthcare more affordable and accessible. These are the sentiments of the current South African National Department of Health (NDoH) minister Dr. Aaron Motsoaledi (Chowles, 2015). In an effort to increase access to quality healthcare to its citizens, a NHI fund is to be established (Department of Health, 2012; Mayosi & Benatar, 2014; Mayosi et al., 2009). At the heart of this fund is a national e-Health system that will handle all patient information and reporting

(Department of Health, 2012). In an article by Chowles (2015), the minister of health emphasizes "the importance of a functional eHealth system, if public and private sector resources are to be merged under a single [National Health Insurance] (NHI) fund to pay for all healthcare services." This highlights the importance of having a working e-Health solution when the NHI fund rolls out.

In order to ensure the successful implementation of a new national e-Health system, an assessment of the factors that influence acceptance or adoption of these systems will help gain valuable insight from potential/current adopters. With this information, the developers of such technologies will -going forward -be able to continuously "evaluate the ability of these interventions [ connecting] with their target audiences throughout development as well as their efficacy and effectiveness" (Atkinson, 2007:612). Facilities that already have these systems in place or plan to implement these systems will be able to use the proposed model presented in this study to maximize usage of the system in their facility.

Furthermore, this study also seeks to add to the South African scientific community's body of knowledge through the manner in which the conceptual model of the study is put together. Researchers have used theoretical models such as TAM, TPB, TRA and Doi before to study user adoption, however there are relatively few studies that have considered both the individual and organizational perceptions specifically among healthcare professionals (Putzer & Park, 2010).

1.5. Research Design

This research study will make use of a pragmatic research design paradigm. That is to say that the study will not be committed to any one system of philosophy or reality. According to Creswell (2003) the pragmatists look to the 'what' and 'how' of research based on its intended consequences and where they want to go with it. Instead of a method being the important aspect, the problem receives most of the attention, and the researcher uses all approaches to understand the problem.

The data collection for this study will follow a sequential exploratory strategy. The first phase is qualitative in nature and utilises semi-structured interviews with three individuals from healthcare facilities in the North West province. The nature of these interviews was such that they allowed the researcher to record the general perspectives and opinions of the participants on the subject of EHR usage as well as verify themes that were identified from the literature. The findings of this phase went towards constructing the survey that would be used in the quantitative phase.

The second phase - the quantitative phase - collected data through an online survey constructed using the findings of the qualitative phase. The questionnaire was given to respondents to complete and their responses exported to a spreadsheet. The main attribute of the study was the actual system usage of the EHR in the healthcare centre. Actual system usage was determined by behavioural intent which in turn was influenced by various PU and PEOU attributes. The qualitative findings analysis was done using thematic analysis and the quantitative findings were analysed using descriptive statistical analysis in order to accept or reject the hypotheses put forward in below sections.

Ten (10) healthcare facilities (public and private) have been identified through a desktop search to participate in this study. From these 10, a target of five (5) participants from each facility will be utilised for the qualitative aspect of the study. The questionnaire will be distributed across any of the 10 facilities which agree with conducting the research at their facility.

1.6. Research Questions

This section presents the study research questions and related objectives which shaped the subsequent hypothesis of the study.

1.7. Major research questions

i. Which factors influence adoption of EHRs in the North West Province?

ii. How can the adoption of EHRs in South African healthcare facilities be improved? 1.8. Research Aim and Objectives

The aim of this study is to investigate the adoption of EH Rs by identifying the factors that influence its rate of adoption within healthcare institutions in the North West province.

1.9. Major Objectives

i. To identify the factors that influence the individual's decision to adopt EHR systems in South Africa. ii. To determine how the adoption of EHRs in South African healthcare facilities can be improved.

1.10. Research Hypothesis

The conceptual model of this study is made up of a number of factors identified from prominent adoption theories. From each of the identified factors, a hypothesis on the effect of that particular factor on the adoption of electronic health records was constructed.

Hypothesis 1: Behavioural Intent is positively correlated with Subjective Norm through Voluntariness. Hypothesis 2: Behavioural Intent is positively correlated with Subjective Norm through Experience.

Hypothesis 3: Perceived Usefulness is positively correlated with Subjective Norm through Experience.

Hypothesis 4: The Nature of the Social System within the workplace is positively correlated with Subjective Norm.

Hypothesis 5: Change Agent Promoters are positively correlated with Subjective Norms.

Hypothesis 6: Relative Advantage is positively correlated with Perceived Usefulness. Hypothesis 7: Job Relevance is positively correlated with Perceived Usefulness.

Hypothesis 8: Output Quality is positively correlated with Perceived Usefulness.

Hypothesis 9: Result Demonstrability/Objective Usability is positively correlated with Perceived Usefulness.

Hypothesis 10: Complexity is negatively correlated with Perceived Ease of Use.

Hypothesis 11: Trialability is positively correlated with Perceived Usefulness.

Hypothesis 12: Communication Channels are positively correlated with Subjective Norm.

Hypothesis 13: Computer Self-Efficacy is positively correlated with Perceived Ease of Use.

Hypothesis 14: Perceptions of External Control are positively correlated with Perceived Ease of Use. Hypothesis 15: Perceived Enjoyment is positively correlated with Perceived Ease of Use.

Hypothesis 16: Computer Playfulness is positively correlated with Perceived Ease of Use. Hypothesis 17: Computer Anxiety is negatively correlated with Perceived Ease of Use. Hypothesis 18: Perceived Ease of Use is positively correlated with Actual System Use.

Hypothesis 19: Perceived Usefulness is positively correlated with Perceived Ease of Use.

Hypothesis 20: Behavioural Intention is negatively correlated with Patient Safety Endangerment. Hypothesis 21: Perceived Ease of Use is positively correlated with Behavioural Intention.

Hypothesis 22: Perceived Usefulness is positively correlated with Behavioural Intention. Hypothesis 23: Behavioural Intention is positively correlated with Actual System Use. 1.11. Ethical Considerations

1.11.1. Quality & Integrity and Independence & lmpartialness

The researcher will ensure that the study is of a standard that is acceptable to the scientific community and will present the findings as accurately as possible, with no prejudice or biases. The measures taken to ensure the above-mentioned will be described in more detail in chapter 3 of this study.

1.11.2. Consent, Voluntary Participation & Participant Safety

Permission to conduct this study has been applied for and obtained from the North West University Ethical

clearance committee before following a similar process with the provincial department of health. The participants that are involved in this research study have been informed of their choice to participate and their choice on terminating their participation during the process. Furthermore, the results of the study can be freely shared with the participants upon request. All the facility employees that are to participate in this study will do so at the discretion of the facility management.

1.11.3. Confidentiality and Anonymity

The privacy of the respondents will be handled with the utmost of regard and will not divulged in any way, whether directly or indirectly, unless otherwise stated by the respondent.

The researcher will adhere to the North West University research ethics policies further by using pseudonyms where necessary to protect the identity of participants.

1.12. Research Scope and Delimitation 1.12.1. Use of theory in predicting adoption

The use of theory to describe the psychological factors that determine user acceptance is not the only measure of acceptance/adoption. There are many other factors apart from the psychology of a user that can equally, if not greater, predict the use of technology i.e. corporate culture of an organization which can be a factor for the acceptance/adoption of technology in an organization. The use of theory shall only be utilized to predict usage of EHRs. This study shall not in any way try to disprove or create a new theory but just apply existing theory in the context of healthcare settings in the North West province.

1.13. Summary

This study looks at various attributes that used to predict the intentions of a healthcare employee to use the EHR system at their workplace. The above-mentioned attributes, took into consideration the psychological state of the individual as well as their organizational context. The research study was divided into a qualitative phase and a quantitative phase. The qualitative phase of the study looked at uncovering attributes that were not already identified by TAM2 and Doi. The quantitative phase verified and measured the type of relationship between the identified attributes and the main variables of the study.

Analysis of the data collected involved descriptive statistics, bivariate as well as partial correlations. Several attributes that displayed weak relationships with the main variables were removed to leave only those attributes that displayed a strong relationship with adoption of EHRs. Stakeholders and decision makers of healthcare facilities can look into strengthening these aspects surrounding the use of EHRs. 1.14. Provisional Chapter Outline

The results of the above-mentioned research are presented in five main chapters. A brief overview of these chapters and general structure of this research is summarized in the following paragraphs.

• Chapter 1: Overview of Study

An introduction of the research topic is given in this first chapter. The problem statement and motivation for the research are discussed, as well as its aims and objectives, hypotheses, questions and how the research was carried out.

• Chapter 2: Literature Review

The second chapter comprises of an account of the concept of user acceptance. A comparison of the models that are mentioned in the brief literature review is given as well as an evaluation of the literature that exists on the use of TPB, TAM (and its derivative TAM2) and Doi on technology acceptance.

• Chapter 3: Research Method and Design

Chapter three discusses the research methodology employed within this study. This covers the collection and analysis of the research data. Those methods applicable to the research were chosen based on their suitability to achieving the research objectives.

• Chapter 4: Data Analysis and Results

This chapter covers the results of the data collection. The collected data is transformed to meaningful information that is useful to this study using methods of analysis mentioned in the preceding chapter.

• Chapter 5: Research Conclusion and Future Work

This chapter concludes the research study and it shows how the problem addressed, was solved. Research contributions, future work and conclusions are also discussed.

CHAPTER 2 LITERATURE REVIEW

In this chapter, EHRs are discussed in greater detail. The discussion begins by defining what EHRs are and how they differ from Electronic Medical Records (EM Rs). The discussion continues by outlining how various countries have implemented EH Rs which are reflected on as case studies in order to identify what constitutes the EHR systems in those nations. Towards the end of the chapter, the thesis places the focus on Technology Acceptance which is discussed and an account of the models that try to predict it is given.

2.1. Electronic Health Records (EHRs)

The term Electronic Health Record is frequently confused with other related terms that refer to a digitized patient medical chart. "The type and extent of electronic health records vary and often at times, what one country calls an EHR may not be the same as that developed in another country" (Kleynhans, 2011: 10).

Kleynhans highlights the realization that some authors use the term EHRs while others refer to EMRs. There is often confusion over whether these terms can be used interchangeably or not. The contrast

between the two terms allows the decision makers looking to adopt this technology in their facility, to know exactly what distinguishes an EHR from an EMR in order to avoid procuring the incorrect system.

Abayomi-Alli et al (2014:22) defines an electronic medical record as "A record in digital format that is theoretically capable of being shared across different health care settings and it includes a range of data like demographics, medical history, medication and allergies, immunization status, laboratory test results,

radiology images, vital signs, personal statistics like age and weight, and more". According to the National Alliance for Health Information Technology (NAHIT), this electronic record is created and managed by licensed clinical staff from an organization that is involved in the individual's healthcare (Neal, 2008). NAHIT defines EHRs as follows: "The aggregate electronic record of health-related information on an individual that is created and gathered cumulatively across more than one health care organization and is managed and consulted by licensed clinicians and staff involved in the individual's health and care." (Neal,

2008:1).

From the above definitions, the following information regarding EHRs and EMRs can be deduced.

Characteristics of the EMR include:

• The record is generated when a patient is visiting a physician or any healthcare facility.

• It captures all medical information from that visit; all clinical units, for example casualties and radiology.

• It is created and consulted by licensed clinicians.

Characteristics of an EHR include:

• It is the summation of all the individual patient medical records.

• Includes records from all health care facilities visited in a patient's lifetime.

Distinguishing EMRs from EHRs is important because it brings our attention to what functionality is expected from each of them. The confusion over the definitions of EHRs and EMRs has been cited by Bagley (2012) as one of the many reasons which have led to this technology having low adoption rates. According to Bagley (2012), the confusion between EMR and EHR has resulted in incorrect interoperability specifications of these systems.

In the study by Bagley (2012) the author used Google Search trends to identify which of the two terms (EHR or EMR) was used predominantly between 2008 and 2012. The logic behind this was that, the term that was being searched for the most, reflected a more informed understanding of said term and thus, an increase in its adoption. Upon completion of the study, the conclusion was that EHR was the term that vendors, governments and medical practitioners were leaning towards at that time as opposed to EMRs.

Software implementations of an EHR are called EHR systems. EHR system implementations vary from system to system based on the needs of the healthcare facility they serve. This along with different technology standards, is the cause of the interoperability issue cited Bagley (2012). The systems currently being used in South African facilities are tailored to the healthcare facility or group of facilities they serve and as such do not all follow a set standard.

2.2.1. EHR Implementations Around the World

In this section EHR implementations from around the world are examined in order to be able to examine how EHRs in South Africa fair against their global counterparts. It is important to note however, that the functionality of these EHRs might evolve as time progresses, with new functionality being added as and when needed by the respective nations. The functionality which makes up an EHR today, might only refer to a subset of what is makes up an EHR in future (Kuhn et al., 2015).

EHR implementations around the world are composed of different modules which contain certain information that contributes towards the patient's EHR. These modules might have different names per country, but their functionality remains the same with slight differences here and there. A review of the Health systems in Australia (Canada Health lnfoway, 2011 ), Belize (Belize Ministry of Health, 2009), Canada (Canada Health lnfoway, 2008), Denmark (Gray et al. 2010), Estonia (Tiik & Ross, 201 0; E-Health, 2010), Netherlands, New Zealand (Gray et al. 2010) and Sweden (Gray et al. 2010) was conducted. These systems appeared to possess common functionality/modules between the various systems. These modules are summarized in the image below.

Shared Health Summary Drug Monitoring System e-Referral s Laboratory and Testing Electronic Heath Record System Pub Ii c Health Information Computerized Provider Order Entry (CPOE)

Maternal Chi Id Health

Figure 2-1: Summary of common modules within an EHR The above image consists of the following:

• A Shared Health Summary

This part of the system handles the patient's health status and includes, but is not limited to, information regarding; medication lists (current or previous medications), details of allergies, immunisations, notes made previously regarding the patient by physicians and/or nursing staff, problem lists detailing

administrative classification (such as ICD-9-CM) of previously diagnosed conditions, and discharge

summaries from encounters at previous facilities. • Laboratory and Testing

This module records all data captured and obtained from laboratory tests, vitals and radiology. These include, but are not limited to; laboratory reports, radiology reports, radiology images, diagnostic-test results, and diagnostic-test images.

• Public Health Information

This module tracks those health activities that are not really related to a single individual. Essentially it keeps track of health activities that affect communities at large such as poisoning, domestic abuse, animal bites, reportable infectious diseases and disease outbreaks. This information can be used by local governments for statistical purposes and provides decision support. The Public Health Information module also includes information on HIV/AIDS, including pre/post-test counselling data as well as data from clinic visits.

• Computerized Provider-Order Entry

This is the part of the EHR system concerned with creating, dispensing, cancelling and administering

prescriptions to patients. Not only can this module order medications, but the physician is also able to order tests such as laboratory tests, radiologic tests, consultation requests and nursing orders.

• Maternal Child Health

The module that is responsible for recording all pregnancy and infant delivery information.

• Drug Monitoring System

This is the part of the EHR system that ensures that the physician or licensed healthcare personnel

accurately records patient safety critical information such as dosages and guidelines that apply to

healthcare professionals. Features of this module include: clinical guidelines, clinical reminders,

drug-allergy alerts, drug-drug interaction alerts, drug-laboratory interaction alerts (e.g., digoxin and low level of serum potassium) and drug-dose support (e.g. renal dose guidance).

• e-Referrals

E-Referrals allow clinicians to electronically request a referral from other healthcare providers.

From the EHR implementations studied, it is important to note that not a single system possessed the complete set of functionalities depicted in Figure 2-1. Some implementations had more or less modules,

with some going beyond the scope of EHR's and including hospital administration functions such as facilitating medical scheme claims and scheduling of appointments with patients. This brings into light the observation that the various countries of the world have a non-standardized view of the functionality of an

EHR system.

2.2.2. EHR Implementations in South Africa

According to Wright, O'Mahony and Cilliers (2017:5) the Council for Scientific and Industrial Research (CSIR) and Department of Health "have reported at least 42 different Health Information Systems (HIS),

i.e. systems that recorded transactions specifically in support of patient administration and care, already

in operation in the public sector in 2013". Of these systems, the following are in use in provincial healthcare facilities from the various provinces (see Table 2-1 ). Private healthcare facilities (whose systems are not documented in Table 2-1) are developed and maintained by the private healthcare groups for which they fall under, and are not shown in this table.

Table 2-1: Health information systems per province adopted from Department of Health (2012).

Province Health Information System

Eastern Cape Delta 9

Free State Meditech; PADS

Medicom; Soarian MedSuite; PharmAssist;

Gauteng

Province Health Information System

KwaZulu-Natal Medicom; Meditech; PALS; Pro-Clin;

ReMed

Limpopo Medicom

Mpumalanga PAAB

North West PAAB

Northern Cape Nootroclin

Western Cape Clinicom; Delta 9; PHCIS; JAC Pharmacy

The systems mentioned above are described below. All the information is from the websites of the

respective vendors unless otherwise stated in the descriptions. A desktop search could not provide details

for all the systems mentioned in the table above and as such those systems are not described below. Other systems that are operational within the country but are not present here can be studies in the article by Wright, O'Mahony and Cilliers (2017).

i. Delta 9

Delta 9 TM provides a solution for healthcare institutions called UniCare TM. According to Ethniks (2013),

UniCare TM is currently in use in 108 hospitals and clinics throughout Southern Africa. Eighty (80) of these

hospitals are in the public sector and 28 in the private sector. The solution provides for the effective

management of management of: Admissions; Pre-admissions; Billing; Credit control; Reporting;

Dispensing; Stock control; Retail interface; Electronic claim submissions (EDI) and Management

information;

Modules supported by UniCare ™ include Patient Registration; Master Patient Index & Records; Patient

Record; HIV/AIDS Management; Appointments; Order Entry and Results Reporting; Laboratory &

Radiology; Operating Theatre; Accident & Emergency; Accounts; Pharmacy Management; Dispensing

Function; Stock Control; Purchase Administration; Interfaces to other Software.

ii. Meditech

According to Meditech (2018:1), "they have provided integrated software solutions to healthcare organisations throughout Africa and the Middle East since 1982". Meditech claims that their system provides the following benefits: maximum productivity through intuitive navigation and instinctive,

specialty-specific functionality; Evidence and Expert-Based Decision Making and Mobile Solutions

including Physician-Driven Implementation as well. iii. Pro-Clin

Pro-Clin is a system that has been in production since 1988 and was acquired by a private company in 2002. This system is implemented in the Kwa-Zulu Natal (KZN) province and boasts modules such as: Out-patienUClinic Administration; Occupational Health; HAART Management; In-patient Administration;

Ward Administration; Theatre Administration; Account Administration; Dispensary Administration; 1.O.D. Administration; Diary Menu; Access Security; and Integrations & Interfaces (DigiData, 2017).

iv. ReMed

According to the KZN Department of Health (2008: 1 ), "ReMed Chronic Dispensing Programme is web -based system, and is designed by pharmacists from Pharmaceutical System Development". The application has been developed with speed and reliability being two of the main areas of focus. The database system can be loaded onto the host institution's file servers and can thus accommodate multiple users.

Both thermal transfer printing and standard A4 laser label printing is available. Other features include the setting up of medicine groups with commonly used regimens grouped for rapid dispensing of for example ARVs, patient "search" functions, "queued prescription" functioning allowing for clinic filter and date range specifications, together with password access control and audit reports. REM ED back-up support features a Service Level Agreement with the developers, with a response time of a maximum of 24 hours in most instances for queries or problems.

v. PAAB

The Patient Administration and Billing System (PAAB) is run by private company but is owned by the Department of Health. "While mainly used for administration, a clinical data-recording module has been added but lacks the functionality to enable the data to be used in an integrated manner. Furthermore, the system does not currently support electronic linkage to a pharmacy system, direct importing of laboratory or radiology results, and decision support" (Wright, O'Mahony and Cilliers, 2017:7).

vi. Nootroclin

Nootroclin is a HIS implemented throughout the Northern Cape Province and has been functional since January 2000. According to MindMatter (2018: 1 ), Nootroclin is described as a "database agnostic system and is currently able to operate on Oracle, lnformix, Cache or lnterbase". Nootroclin has the following modules: Master Patient Index; Patient Registration; Inpatient Admissions; Outpatient Bookings; UPFS Billing; Debtors Management; Clinical Checklists; Pharmacy Management (NootroPharm Module); Order Entry; Results; ART monitoring; Diets; Theatre; and Management Information.

NootroClin supports standards such as ICD-10 for diagnosis, NSN for pharmaceutical stock, and HL7 for interaction with other systems (Disa*Lab).

vii. Clinicom

Clinicom is a system used in nearly all hospitals in the Western Cape (Wright, O'Mahony and Cilliers, 2017). According to Western Cape Government (2016: 1 ), "the Clinicom system offers a number of benefits to patients, doctors and administration staff, which includes but is not restricted to the following: The same patient number will be used for a patient at all Western Cape hospitals and Primary Healthcare Services that utilise the central Western Cape Patient Master Index; Patient detail and history can be viewed across institutions in developing a single patient record; More efficient management of outpatient visits, including an electronic outpatient appointment booking system."

viii. PHCIS

The Primary Health Care Information System (PHCIS) was implemented in 2006 and now "connects 176

primary care facilities across the Western Cape province administering EHRs of over 5.2 million patients"

(Chowles, 2014:1). The Western Cape Government (2016:1) states that ever since being implemented,

benefits experienced by facilities include: "Coordinated and streamlined patient administration; Minimal

duplication of patient information; Access to individual electronic patient records (EPR) via a unique patient

number; Provision of an automated and more reliable headcount statistics; Improved service to patients; Reduced waiting times and time for patient admission; Improved communication between facilities; Improved patient flow; Easier locations of patient medical records regardless of where they were previously registered; and Sites and Dates of PHCIS rollout to Community Healthcare Centres (CHC's)".

ix. JAC Pharmacy

Another system widely implemented in the Western Cape, JAC Pharmacy provides "pharmacy stock control, e-prescribing and medicines administration as a single integrated solution" (Mills, 2014:1). JAC is currently deployed at more than 30 hospitals in the Western Cape and is due to be deployed to the major

clinics in the province (Mills, 2014:1). 2.3. The concept of User Acceptance

Over the course of time, many innovations have come and gone or rather, evolved from their primitive

state until a certain maturity state where they could not be upgraded anymore; or they totally get replaced

by a new innovation that performs the same function more efficiently. The same phenomenon can be seen

in hospital record keeping practices. These practices initially emerged as paper-based practices where patient information was captured on paper and stored in filing cabinets (Atherton, 2011; Mcclanahan, 2012). With the evolution of technology came new ways that made the record keeping process more efficient. Hospital record keeping practices have thus evolved to the state in which they currently exist in.

Evolution occurs naturally and it is hard for man to resist it. However, for technological innovations, this point is not as simple. After the birth of an innovation, it can quickly cease to exist, not because it ceased

to function the way it was designed, but because it was not well received by society. This can be referred to as Technology/Innovation adoption.

Dillon (1996: 1) defines technology user acceptance as "the demonstrable willingness within a user group to employ information technology for the tasks it is designed to support". Within the context of this study, technology adoption refers to the use of EHRs (if any) within healthcare settings. The lack of acceptance could nullify the potential benefits of this technology. It therefore becomes important to evaluate the use

of this technology and put in measures (if need be) on how to effectively use it.

The past decades produced an increase in studies and research that has looked at the acceptance of various technologies by individuals, groups of individuals and organizations (Calisir et al., 2013; Lu et al., 2014; Grover, 2015). These studies have looked into which factors have proven to be significant in predicting the use of technology. Dillon (1996:1) highlights that the concept of acceptance has not been

applied to instances whereby "users claim they will employ it without providing evidence of actual use, or using the technology for purposes unintended by the designers or procurers".

From these studies, three major theoretical approaches have seemed to be dominant namely: TRA, TPB and TAM. The rest of these studies are enhancements of existing theories and "identify intrinsic and extrinsic factors involved in decisions, intentions and individual's satisfaction about the acceptance and the use of information technology" (Silva and Dias, 2007:69). This point is demonstrable by the fact that the two latter theories mentioned above (TPB and TAM) are derivatives of TRA.

2.3.1. The Theory of Reasoned Action (TRA)

From as early as 1862 psychologists began investigating how attitudes affected behaviour. Two early psychologists, Thomas and Znaniecki (Levine & Pauls, 1996; Levine, Little & Mills, 1997: 1) viewed attitudes as "individual mental processes that determine[d] a person's actual and potential responses".

Upon investigating the relationship between attitudes and behaviour, theorists found a relatively low correspondence between these two variables (Wicker, 1969; Abelson, 1972). They found that when it comes to certain behaviours, people would say one thing, and do the other - in essence becoming hypocrites. For example, there are a number of warnings regarding the dangers behind smoking, or the implications of speeding but people still engage in such activities despite the warnings.

The low correspondence between attitudes and behaviour led social psychologist Martin Fishbein to distinguish between attitudes towards an object and attitude towards behaviour in relation to that object. (Glanz, Rimer, & Viswanath, 2015) Fishbein (1975) and Ajzen (1982) pointed out that for an attitude to be a better predictor of a behaviour, the attitude would have to be more pertinent to the situation or behaviour. For example, a general attitude - say, an attitude toward Asians - and the behaviour - say, a decision whether to help a particular Asian couple-would not yield a close correspondence between the attitude and the behaviour i.e. people could claim to not have a problem with Asians, but in a situation to assist an Asian couple the same people could be reluctant towards performing the behaviour (Myers, 2010). An experiment by Richard Page along with Harold Sigall (1971) devised a supposed device to determine the attitudes of white university students about black people. The experiment involved one group of students taking a typical pen and paper questionnaire to determine their attitudes, while another group was subjected to questions regarding black people while being plugged into the supposed device. Before the students were split, they had been asked questions regarding their beliefs before the experiment commenced.

The students were then made to believe that the machine worked by asking them questions that they had previously inadvertently given the answers to. Once the students were under the impression that the machine worked, the machine was hidden and the questioning commenced. The experiment revealed that compared to the students that took the pen and paper questionnaire, those on the fake machine displayed more negative beliefs regarding black people. Supposedly out of fear of being judged by the experimenter,

the respondents that had been plugged on the fake machine reversed their initial judgements. These two

scenarios respectively demonstrate how attitudes towards a behaviour and societal influences have an impact on the behaviour of a person.

TRA employs the principle of aggregation to predict user behaviour. This principle entails that the effects

of a certain attitude on a particular behaviour become more apparent when the person's aggregate or

average behaviour is observed, rather than at isolated acts (Myers, 2010). This highlights the fact that

people actually contemplate "the consequences of their actions before they decide to engage or not

engage in a given behaviour'' (Fishbein & Ajzen, 1980:5). When formulating TRA, the major assumption

was that the best predictor of behaviour is Behavioural Intention, which is modelled as the combination of

the attitudes an individual holds towards a behaviour and the influences they receive from society (Glanz,

Rimer, & Viswanath, 2015). Behavioural intention indicates how much effort an individual would like to

commit to perform such behaviour. Higher commitment is more likely to mean that behaviour would be

performed (Fishbein & Ajzen 1975).

Beliefs and Attitui:te toward

Evaluations Behavior

-Behavioral Actual

Intention Behavior

Normative

Beliefs and Subjective

-Motivation to Norm

copy

Figure 2-2: Theory of Reasoned Action. Adopted from Fishbein & Ajzen (1975)

As is depicted in Figure 2-2, the intentions of an individual to perform a behaviour is attributed to their

attitude with respect to performing that behaviour as well as any subjective norms. The focus of the

theoretical model is understanding the determinants of specific human behaviour by looking at the

individual motivational factors that lead to the undertaking of the behaviour in question (Glanz, Rimer, &

Viswanath, 2015).

The theory draws on two concepts, namely; the Principle of Compatibility and the concept of behavioural

intention. The Principle of Compatibility specifies that in order to predict a specific behaviour directed to a

specific target in a given context and time, specific attitudes that correspond to the specific target, time

and context should be assessed (Ajzen 1988; Fishbein & Ajzen 1979). The concept of behavioural

intention states that an individual's motivation to engage in behaviour is defined by the attitudes that

influence the behaviour (Fishbein & Ajzen 1975).

The attitude of the individual with respect to performing the behaviour is based on the outcomes of

performing that behaviour. This is composed of the behavioural belief itself and weighted by evaluations

of the behavioural outcome. Therefore, if an individual holds a strong belief regarding the outcomes of

performing certain behaviour, then their attitude towards their behaviour will also be positive. Mathematically this is represented as:

A=

Lbiei

Where A: represents the attitude,

b;: represents the individual's salient beliefs e;: represents the individual's evaluation of the

outcomes of that behaviour

Subjective norm is the perception that the person performing the behaviour the people who are important to them. Subjective norm is composed of; normative beliefs i.e. the approval or disapproval of the performed behaviour by people that the individual views as important to them, as well as the motivation to comply with those people. Therefore, if an individual believes that certain people which they value would

approve of them performing a behaviour and the individual is willing/motivated to comply with these people,

then their subjective norm would be positive. Mathematically this can be represented as:

SN=

Lnb

imci

Where SN: represents the subjective norm,

nb;: represents the individual's normative beliefs me;: represents the individual's motivation to comply

Thus, according to TRA, the behavioural intention (Bl) of a person to perform a behaviour can be determined using the following formula.

BI= A+ SN

Where A is the measure of the attitudes that a person holds about performing certain behaviour and SN is the measure of subjective norms associated with performing that behaviour.

2.3.2. The Technology Acceptance Model (TAM)

TAM is a derivative of TRA which according to Holden & Ben-Tzion (2009:2), "was developed in the 1980's, in light of concerns that workers were not using IT made available to them". TRA and TPB (discussed later) have been used to predict adoption from individual perspectives regarding activities such smoking, exercise, condom use and moral behaviour (Yzer et al., 2015; Kangas et al., 2015; Paul, Modi & Patel, 2016). TAM on the other hand was created solely to predict usage in the IT context.

The technology boom brought with it an introduction of information systems into organizations. As these systems gained popularity, "researchers and practitioners devoted substantial research effort into determining which factors affect users' beliefs and attitudes on the IS acceptance decision, and what factors contribute to user resistance" (Lee, Kozar, and Larsen, 2003:754). The output of such efforts was

the adaptation of TRA by Davis (1986) called TAM. Figure 2-3 below gives a diagrammatic representation

of TAM as formulated by Davis.

.

Extemal

Variab

les

Perceived

Ease

o

f U

se

(EOU)

Attitude

Tovmd

Using (A)

Behavioral

+ - -

, .

1

Intention

U

se

A

c

tu

al

(BQ

Syst

em

U

se

Figure 2-3: The original Technology Acceptance Model adopted from Davis (1986)

According to the TAM theory, a person's acceptance of an Information System (IS) mainly lies in their evaluation of PU as well as the PEOU of the innovation. According to Yucel & Gulbahar (2013:96), PU "is defined as being the extent at which a person believes that the use of a system will improve his or her performance", while PEOU "refers to the extent at which a person believes that using a given application is free of effort" Yucel & Gulbahar (2013:96).

The behavioural intention in turn, can be determined by considering a person's attitude toward the IS in question and their perceptions concerning its usefulness. Attitudes towards the IS are a result of the beliefs that a person holds about the IS. These beliefs are based on whether the individual perceives the system as useful to them as well as the relative ease that goes into using the system. Matthews (2012: 1) further elaborates that "External variables, such as the task, user characteristics, political influences, organizational factors, and the development process, are expected to influence technology acceptance behaviour indirectly by affecting beliefs, attitudes, or intentions."

After the birth of TAM, the author Davis (1989) tested his theory empirically and concluded that the theory was more accurate at predicting and explaining behaviour through the use of only three constructs. He therefore eliminated the effects of the attitudes towards the behaviour and the effect of the external variables on intentions to use the system and ultimately actual system use. The revised TAM proposed by Davis (1989) is presented in figure 2-4 below.

Perceived

U~fulness

(U)

Pe~ived

Ease of Use

(EOU)

Intentions

to

Use

Actual System Use

(Usage)

Figure 2-4: The revised Technology Acceptance Model adopted from Davis (1989) 2.3.3. Theory of Planned Behaviour (TPB)

As TRA gained traction in the psychology community, limitations in the theory were identified. Among

these were criticisms that the theory neglected other factors which in real life could be used to predict

individual behaviour (Grandon & Peter P. Mykytyn, 2004; Werner 2004) as well as "the assumption that

intention directly led to action without limitations" (Truyong, 2009: 178). Another pitfall of TRA is that it fails

to predict the behaviour of "people who have little or feel they have little power over their behaviours and

attitudes" (Levine & Pauls, 1996: 1; Levine, Little & Mills, 1997: 1 ). To overcome this limitation, the TRA was

modified to include a third antecedent - Perceived Behavioural Control (PBC).

PBC refers to a person's perception of the level of difficulty when performing a behaviour of interest.

(Chestnutt, 2016). The addition of this construct gave birth to the Theory of Planned Behaviour (TPB).

Perceived Behavioural Control is based on control beliefs and their perceived facilitation. "Control beliefs

include the perceived availability of skills, resources, and opportunities, whereas perceived facilitation [or

influence of control beliefs (Ajzen, 1991 )] is the individual's assessment of available resources to the

achievement of a given set of outcomes" Chuttur. (2009: 12).

As shown in Figure 2.5 below, PBC also has a direct influence on the actual behaviour being performed.

That is, even if the individual has a positive attitude and positive views regarding their subjective beliefs, if

they believe that their perceived behavioural control over the behaviour is low, the probability of them

performing the behaviour is also low.