We willen u hartelijk danken voor het invullen van de vragenlijst.

(1)

Running head: eHealth, client service satisfaction

Is eTriage suitable for all types of mental healthcare clients and disorders and how does the role of the therapist contribute to client service satisfaction?

Freark W. Mous (s2035154)

1 Oktober 2015

University of Groningen Faculty of Economics and Business

MSc BA Change Management

Supervisor: Dr. M.A.G. van Offenbeek Co-assessor: M.L. Hage, MSc

Word count: 12.646

(2)

eHealth, client service satisfaction

2 Abstract

Information Technology (IT) within healthcare is becoming more important to ensure that healthcare remains affordable and accessible. One of these IT systems within healthcare is eTriage, which is an online triage tool used within a large mental healthcare organization.

However, eHealth technology is often top-down implemented and seems not to take the clients’ perspective into account to a sufficient extent. This study researches the viewpoint of the clients by researching how the type of triage, age, gender and the type of disorder

influence client service satisfaction. Furthermore, the role of the therapist is taken into account in these relationships. The results of this study (n=135) indicated that clients who were assigned to specialized mental healthcare were more satisfied with eTriage than with the non-digitalized triage. Thus, eTriage suits the clients from specialized mental healthcare better due to their condition, rather than due to their other personal characteristics. This is also supported by that no evidence could be found of other characteristics of the clients

influencing client service satisfaction. Therefore, it seems that most of the client groups could be treated with the same triage process, due to their similarity in client service satisfaction.

This implies that eTriage is a potential option to be used next to or replacing a non-digitalized triage.

Keywords: eTriage, client service satisfaction, age, gender, eHealth Acknowledgements

First and foremost I would like to thank dr. M.A.G. van Offenbeek for her motivating support and the insightful and extensive feedback. Secondly I would like to thank Lentis and my internal supervisor, dr. M.R. Dekker, for the opportunity to conduct research within an

organization and getting the chance to apply theoretical knowledge in practice. Furthermore, I would like to thank all employees from Lentis who helped to send the questionnaires to the clients. Finally, I would like to thank my fellow researchers for constant brainstorming and sharing ideas regarding this study.

(3)

3

Table of content

Introduction ... 5

Literature review ... 8

eHealth ... 8

Client Service Satisfaction ... 9

Role of the Therapist ... 14

Method ... 15

Case Information ... 15

Data Collection ... 20

Questionnaire and procedures. ... 20

Participants. ... 21

Measurements. ... 21

Data Analysis... 22

Factor analysis. ... 22

Exploring data & Correlation. ... 23

Reliability. ... 23

Analysis of variance. ... 24

Qualitative data analysis. ... 24

Results ... 24

Data Reduction ... 24

Descriptive Results ... 25

Open Question... 31

Discussion ... 31

Summary & Interpretation... 32

Theoretical Implications ... 33

Limitations ... 40

Practical Implications... 40

Future Research ... 43

Conclusion ... 44

References ... 46

Appendix A: questionnaire ... 52

Appendix B: data reduction ... 55

Appendix C: testing hypothesis 1 ... 56

(4)

4

Appendix D: testing hypothesis 2 ... 57

Appendix E: testing hypothesis 3 ... 58

Appendix F: testing hypothesis 4 ... 59

Appendix G: testing hypothesis 5 ... 61

Appendix H: answers to open question ... 66

(5)

5 Introduction

Healthcare providers increasingly employ information and communication

technologies (ICT) techniques to improve quality, access, and efficiency of healthcare (Mair, May, O’Donnell, Finch, Sullivan & Murray, 2012). eHealth is classified as one of those techniques. In this research, eHealth is understood as the process of providing healthcare via electronic means (Oh, Rizo, Enkin, & Jadad, 2005). A typical example of the use of eHealth consists of interactive computer systems using client data to generate case specific advice that can support clinical decision making (Mair et al., 2012). However, research has shown that eHealth implementations are not always successful, since eHealth projects regularly fail or only partly deliver the expected outcomes regarding efficiency and quality improvements (Mair et al., 2012). One reason might be that most eHealth interventions are developed from the organization’s perspective and get implemented top-down (Croonen, 2011). Wilson and Lankton (2004) argue however that eHealth interventions should fulfill clients’ needs. When the organizational perspective is used to implement eHealth, the clients’ viewpoint can get less emphasis. This can cause a gap between the needs of the clients and the ambition of the implementing managers (Daansen, 2012). The question arises as to what extent using a top- down implemented eHealth system is more effective than not using an eHealth system. It needs to be researched in more detail whether this system is useful for everybody and whether it leaves the end-users satisfied.

According to Wilson and Lankton (2004), this general lack of success in implementing eHealth may result from the implementers’ lack of awareness of the clients’ needs. Due to the potential lack of awareness from the implementers this might affect clients. The implementers might not overcome the given fact that some clients do not like to handle IT (Information Technology), or that the IT is inefficient or even ineffective from the clients’ perspective, or even worse, that it lacks efficacy. This could induce that the clients could be guided by the

(6)

6

implementer’s false beliefs. Wilson and Lankton (2004) have demonstrated the importance of the role of the healthcare provider in applying eHealth, since the success might depend on that. Therefore it is justified to ask which role the healthcare provider - in this case the psychologists or screeners - play, in relation to client service satisfaction concerning eHealth.

Next to the role of the implementer, personal factors of clients are also an important facet in the relation towards eHealth. Age is one of these factors, since older individuals often find it more difficult to work with new computer software than younger people do (Parsons et al., 1991). Kunze, Boehm and Bruch (2013) have confirmed the stereotype that older workers are more resistant to change. In contrast to these findings, David and Songer (2009) have found no significant relationship between resistance and age. Next to age, gender might also be an influential factor since Yu, Shen and Lewark (2012) showed that both male and female employees’ satisfaction increased during the implementation of an IT system. Furthermore, female students are more satisfied than male students with an electronic version of learning (Gonzalez-Gomez, Guardiola, Rodriguez, & Alonso, 2012). Another personal factor that might influence the client service satisfaction is the type of disorder. Kelstrup, Lund and Bech (1993) have found a difference in user satisfaction with the treatment along different types of disorders. Clients who were diagnosed with affective disorders or psychoses were more satisfied with their treatment than clients with schizophrenia or paranoia. Moreover, clients with an antisocial or borderline disorder were less satisfied than clients with no personality disorder.

The effect of using an electronic diagnosis tool in healthcare has shown promising results in the sense that the clients and users were satisfied with this diagnosis tool (Dijksman, Dinant & Spigt, 2013). The downsides of implementing eHealth systems, in terms of the frequent failing of these projects or not delivering the expected outcomes (Mair et al., 2012), it is interesting to look at the results especially from the clients’ point of view since this

(7)

7

viewpoint is not often taken into account. Even though eHealth implementations often seem to fail, governments and organizations still stress the importance of the use of eHealth, which is an interesting contradictory. Next, the results from the clients’ characteristics on the effects of such a specific system have scarcely been studied in mental healthcare. Besides addressing the abovementioned contradictions, this study aims to extending the literature with extra insight of the role of the practitioner in the success of sustaining eHealth implementation. The aim of this study is consequently to answer the following question:

“Is eTriage suitable for all types of mental healthcare clients and disorders and how does the role of the therapist contribute to client service satisfaction?”.

In order to answer this research question, one eHealth implementation is analyzed within a mental healthcare organization. More specifically, it concerns an online triage tool, called eTriage. This organization started to pilot test the online triage tool in 2013 and it is now regularly used to assign clients with mental health problems faster to the right service. It should be noted that this is the step before intake and treatment. This tool is used to guarantee both the quality and speed of assigning clients to the right service. In this study the effects of this recently introduced triage tool are analyzed by using the data available at the healthcare organization together with data retrieved from questionnaires conducted with clients.

The remainder of this thesis is as follows. The part following this chapter discusses literature regarding acceptance and resistance towards information technology. That part is followed by section Three, where the methodology of this study is described. Section Four presents the results obtained from this study. The Fifth section discusses the findings and gives theoretical and practical implications. Finally, the conclusion will be drawn in the Sixth section of this paper.

(8)

8

Literature review eHealth

The most used definition of eHealth, according to Oh et al. (2005), is the definition from Eysenbach (2001), which read as follows:

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. (p.20)

Furthermore, according to Kwankam (2004) eHealth is necessary to keep up with the enormous growth of health information. Additionally, Kwankam (2004) stresses that this technology allows to use medical resources more efficiently and to reduce administration costs.

According to Austin and Boxerman (2003), the goals of eHealth are the increased efficiency, the improved quality, the increased commitment to evidence-based medicine, the empowerment of clients and the development of new relationships between clients and health professionals. These goals play an important role in the strategic plan of mental healthcare in the Netherlands (GGZ Nederland, 2013). Because of the previously named goals and the rising costs of mental healthcare (CBS, 2014), eHealth has become one of the most important technologies that can keep the healthcare affordable and accessible for everybody in the Netherlands (GGZ Nederland, 2013). That is why eHealth is part of the strategic plan in the Dutch mental healthcare system, although eHealth projects often seem to fail (Mair et al., 2012). This seems contradictory, which is why it is interesting to further explore this subject.

(9)

9 Client Service Satisfaction

Technology seems to be necessary to keep healthcare affordable and is also required in order to improve the quality of healthcare (Austin & Boxerman, 2003). This implies that the clients have to start using these technologies, since eHealth is not limited to the employees only. With eTriage the clients notice much more from this technology since they are the end- users of the specific system. It is not only the system that the employees use in their work that changes, but the clients are the main subject of the change, since it is their intake measure.

This is different than with eHealth implementations directed at the employees, who have to use new IT systems, whereby the clients do not have to notice anything, aside from improved outcomes. This means that the employees and clients have something in common in eTriage implementation, since they are both end-users of the IT system. Maillet, Mathieu and Sicotte (2015) showed that eHealth technology goes together with increased user satisfaction. In their research the use of an Electronic Patient Record (EPR) explained more than 50% of the variation in the satisfaction of the end-users, in this case nurses, since the system supported their decision making process, collaborative work and nursing care overall. These results account for the nurses, since they are the end-users of the EPR-system. Therefore a positive relationship between the use of a digitalized triage and client service satisfaction could also apply on the clients who use eTriage, since they are the end-users of this eHealth system.

The quality improvement of healthcare due to technology is supported by Wang, Cheng and Huang (2013). Additionally, due to the technology within healthcare, they also found improvements in satisfaction of the clients. Furthermore, clients making use of

telehealth, in the sense that they are receiving home monitoring services, were more likely to report higher user satisfaction than those receiving usual care (Grant, Rockwood & Stennes, 2015). However, no difference in client satisfaction was found by Tousignant et al. (2011),

(10)

10

who compared clients who got in-home telerehabilitation with clients who got face-to-face treatments.

In the Netherlands the healthcare organizations are assessed with the opinion of clients’ satisfaction with the provided healthcare (Zorginstituut Nederland, n.d.). Furthermore, customer satisfaction results from customers’ positive experiences of using a product or service (K. H. Kim et al., 2008). Significant positive relationships within healthcare between service quality and patient satisfaction have been found (e.g., Andre et al., 2008;

McAlexander, Kaldenberg & Koenig 1994; O’Connor, Shewchuk, & Bowers, 1991).

Furthermore, client satisfaction is very important as it is related to motivation and treatment compliance (Piron et al., 2008). Therefore the client service satisfaction is important to study, since the quality and the clients’ satisfaction are closely related to each other. For this specific study the combination of client service satisfaction with eHealth is also important. Chang and Chang (2008) have indicated that the positive effects of the consumers’ positive evaluations of the use of IT-based tools are related to patient satisfaction through improved service efficiency, increased effectiveness of data communications, and reduced frequency of errors.

Next to the use of eHealth, gender and age might also play a role in the satisfaction of the clients, since Scott (2013) found an indication that these variables might affect the uptake of e-learning. This is relevant since e-learning and eTriage both use a digitalized environment to, respectively, learn and fulfil a triage. Kummervold et al. (2008) found that the overall use of internet for health purposes is growing in all age groups. There was especially a strong growth among young women, which might indicate an overall improvement in the familiarity with technology and consequently increase the user satisfaction as well, since the actual use of technology is positively related with user satisfaction (Maillet, Mathieu & Sicotte, 2015).

On that account the following hypothesis is composed:

(11)

11

H1: Clients who used eTriage are more satisfied with the intake process than clients who did not use eTriage.

Next to the use of the eTriage, the focus of this research will be on the following

characteristics of the client to measure client service satisfaction with; namely age, gender and type of disorder.

Increasing age and technology seem not to go together that well since for example Arning and Ziefle (2007) have shown that older adults have a more pragmatic view or even a rejecting attitude towards technology. However, in the case of healthcare, research has shown that older adults assess eHealth technologies more positively than younger adults (Arning &

Ziefle, 2009). This was assumed to be the case because when the adults’ age increases, they tend to acknowledge the benefits of eHealth technologies better. The positive view towards eHealth might also appear due to the fact that these technologies might influence their health and are thus more worthy of trying. However, the positive view by older clients is not shared by the professionals; Hing and Hsiao (2010) showed that the use of Electronic Medical Record (EMR) was inversely associated with the physicians’ age. Nevertheless, it could be argued that younger people are raised with much more technology than older adults, which could mean that they also have higher chances to get more familiarized with various

technologies. As mentioned before, Maillet, Mathieu and Sicotte (2015) have shown that the actual use of a healthcare technology is positively related with user satisfaction, which could mean that younger people have indeed a more positive attitude towards the use of technology than older adults, since they have used the technology more often. The following hypotheses are composed:

H2a: Younger adults, compared to older adults, are more satisfied when using eTriage.

H2b: Younger adults, compared to older adults are less satisfied when eTriage is not used.

(12)

12

Taking the next personal characteristic, namely gender, into account, this should be done with caution. This is because most of the nurses, often the end-users in the literature, are female; for instance only one fifth of the nurses in the Netherlands are male (CBS, 2012).

Research from Arning and Ziefle (2009) has shown that females were less likely to use technology, in the case of their research Personal Digital Assistant (PDA). Furthermore, Davis and Songer (2009) have shown that females show more resistance to implementation of information technologies. Yu, Shen and Lewark (2012) showed that both male and female employees’ satisfaction increased during the implementation of an IT system. For the male employees this was because it improved the communication and the relationship between the managerial and the non-managerial levels, and for the women the higher user satisfaction was due to the successful implementation of the technological innovation. Concerning e-learning, female students are more satisfied than male students with this electronic version of learning (Gonzalez-Gomez, Guardiola, Rodriguez, & Alonso, 2012). This could indicate that women are more into technology than men, hence the following hypothesis is formed:

H3a: Female clients, compared to male clients, are more satisfied when using eTriage.

H3a: Female clients, compared to male clients, are less satisfied when not using eTriage.

Furthermore, as mentioned before, a difference in user satisfaction with their treatment along different types of disorders has been found by Kelstrup, Lund and Bech (1993).

Therefore this is an interesting characteristic to take into account. Kelstrup, Lund and Bech (1993) found that clients who were diagnosed with affective disorders or psychoses where more satisfied with their treatment than clients with schizophrenia or paranoia. Additionally, clients with an antisocial or borderline disorder were less satisfied than clients with no personality disorder.

(13)

13

Moreover, eTriage influences the client also by the reduced amount of steps in the intake process and less travel time, since the information can be gained online instead of on- site. On the other side, eTriage also limits the contact with healthcare providers. Especially in the case of mental healthcare, the contact might have impact on the client service satisfaction.

This would be the case since the disorders the clients are facing can bring the need of contact.

This could be the first step in the treatment process, because a listening ear can already comfort the client. The process of creating a bond with the client is the start of the treatment process according to Rogers (1951).

The mental healthcare in the Netherlands differentiates two types of mental healthcare, namely basic and specialized mental healthcare. These categories are based on the Diagnostic and Statistical Manual of Mental Disorders (DSM). According to the Nederlandse

Zorgautoriteit (2013), clients are appointed to the specialized mental healthcare when there is suspicion of a DSM disorder with high risk and/or high complexity. The clients with

suspicion of a DSM disorder that have less high risk and/or complexity are assigned to the basic mental healthcare (Nederlandse Zorgautoriteit, 2013). The difference between the two types of categories, the higher the risk/complexity, probably determines the need for treatment due to the height of the risk and the complexity. Hence compared to basic mental healthcare, specialized mental healthcare has increased need for treatment and it could be argued that the need for personal contact at the start of the treatment is higher since this important according to Rogers (1951). He states that creating a bond of trust is the start of the treatment process and this bond cannot be created when there is no personal contact. This can also be seen in the crisis route for mental healthcare, since the clients are assigned to a crisis team that takes care of these clients immediately, since the risk is too high and action should be undertaken

directly. Here the contact is also important, thus the crisis route requires a personal route, with personal contact. On that account the following hypothesis is composed:

(14)

14

H4a: Clients who are assigned to specialized mental healthcare, compared to clients who are assigned to basic mental healthcare, are less satisfied when using eTriage.

H4b: Clients who are assigned to specialized mental healthcare, compared to clients who are assigned to basic mental healthcare, are more satisfied when not using eTriage.

Role of the Therapist

Implementing a new IT system brings major risks according to Markus (2004), starting with the fact that users will not use the technology, followed by the risk of them misusing it or thirdly that they will use the technology without returning the expected

benefits. In addition to that, Markus (1983) explains user resistance in terms of the interaction between system characteristics and the social context of its use. Gibson (2003) also

demonstrates the importance of social systems next to the technical system that changes during an implementation. The changes in these systems can also be the cause for user resistance in IT implementations. This brings up the importance of the social context when using technology. Thus, it brings up the social context when using an eHealth system, since multiple parties are involved. In the case of this research the employee closest to the client would be the therapist who performs the treatment. Accordingly, this would be the person who has most chance to influence the client. Social influence plays a role in the study from Maillet, Mathieu and Sicotte (2015). In this study an ERP system is implemented and social influence was positively related to the satisfaction of the end-users, who in this specific research were nurses. However, most eHealth interventions are developed from the

organization’s perspective and get implemented top-down (Croonen, 2011), which can cause a gap between the needs of the users of the system and the initiative of the implementers (Daansen, 2012). eHealth implementations often lack success (Mair et al., 2012), which according to Wilson and Lankton (2004) results from the implementers’ lack of awareness of

(15)

15

the clients’ needs. The implementers in this case would be the therapists, since they stand the closest to the client. Contrasting this, in the research from Liu et al. (2015), the relationship between social influence and user satisfaction has not been found, since the only important factor that played a role in this research was how the technology can help the therapist. Thus, the following hypothesis is constructed:

H5a: The resistance of therapists towards the use of eTriage will weaken the positive relationship between client service satisfaction and the clients’ characteristics.

H5a: The resistance of therapists towards the use of eTriage will reinforce the negative relationship between client service satisfaction and the clients’ characteristics.

Method

Questionnaires have been used in order to analyze the effects of eTriage, an intake tool used at a large mental healthcare organization, on the satisfaction of clients. Quantitative

information provided by the mental healthcare organization was used to study the possible influence of the role of the therapist. Quantitative data was the starting point of this study. In addition, qualitative data provided the necessary further insights in the quantitative data. The next section will start with background information of the case, followed by the explanation of the quantitative and qualitative research methods.

Case Information

The data used in this research has been collected from a large mental healthcare organization in the Netherlands. The website of this mental healthcare organization

(http://www.lentis.nl) provided together with a conversation with dr. M.R. Dekker (personal communication, 10^th of April, 2015) the background information for this study. The

organization where this research has been conducted delivers basic and specialized mental healthcare to more than 25.000 children, adolescents, adults and elderly in the three northern

(16)

16

provinces in the Netherlands. This organization employs over 4.300 people across more than 100 locations.

The first step in treating the clients is the triage. The triage can be started after General Practitioner (GP) has referred the client to the mental healthcare organization. The mental healthcare organization started in the summer of 2013 to offer a digitalized type of triage, named eTriage, to improve the access of new clients to best matching care. eTriage should also improve the safety for the client, since it can recognize risk signals at an early stage, meaning that the healthcare organization can act quickly in case of emergency.

Before the healthcare organization started to use eTriage, there was a non-digitalized triage process, where a psychiatrist and a nurse did the screening. Now a digitalized triage process is used in parallel to a non-digitalized triage. In the digitalized triage the client has gotten a more important role and a psychologist starts working with the information provided by the client. This change means that roles have changed for many stakeholders in the mental healthcare organization due to eTriage (see Table 1). As mentioned before, the client has a more important role in the triage process, since he provides the necessary information online.

Based on the information provided by the client, a psychologist will create a working hypothesis, wherein the psychologist did not have a role before the implementation of eTriage. The role of the registration service, Lentis Aanmeld Service (LAS), has changed from an information gathering role towards an information managing role. The screeners (psychiatrists and nurses) have lost their role of assigning clients towards the right

department, since the client gives this information with the help of eTriage. GPs, intakers and therapist have kept their role after the implementation of eTriage, but for the intakers and therapist there is more (digital) information available.

Table 1

Stakeholders’ roles before and after the implementation of eTriage

Old role New role

(17)

17 Client - Fulfil triage on site with a

screening

- Online triage

GP - Sending clients to mental

healthcare

- Sending clients to mental healthcare

LAS - Sign up of the client - Gathering information of

clients

- Sign up of the client - Check clients’ abilities for

eTriage

- Managing information of clients

Screener - Assign client towards right department based on given information

No role

Psychologist No Role - Testing and clarifying working hypothesis Intaker - Determines the clients’

treatment trajectory

- Determine the clients’

treatment trajectory Therapist - Determines the treatment

plan together with the client

- Determines the treatment plan together with the client The focus of this research was on the effects of the implementation of the eHealth triage tool, called eTriage, compared to the non-digitalized version of triage. This specific tool falls within the eHealth category since it supports the healthcare delivery with the use of IT (Eysenbach, 2001). The case is, as said, about the triage process that brings the clients to the right service within the mental health organization (see Figure 1 for a flow chart of eTriage and the previous triage process). The triage process has changed from a manual to a

digitalized form of triage. The process starts with the GPs, who send clients to the mental healthcare organization. This information is received by the registration service (LAS). With this information, the clients will be called to confirm their email address and to inform them about the triage process. It should be noted that for eTriage the Dutch language skills should be sufficient and the client needs to have a working internet connection. When this step is

(18)

18

completed, the client will receive automatically an email with the link to fill in the triage online. When this online triage has been completed, the outcomes will be reviewed by a psychologist. Together with the triage information and the professional knowledge of the psychologist, the latter will form a working hypothesis. But in case of risk, the system alarms the supervisors and the client will continue with the crisis service, in which case a working hypothesis will not be made by the psychologist. When there is no risk detected, the working hypothesis is tested and if necessary clarified in a ten-minute phone call between the

psychologist and the client. After this phone call the client will be addressed to a specific department within the mental healthcare. This decision information will also flow back to the LAS, where the client will be referred to the right department. The client will be placed on a treatment waiting list of the specific department. Subsequently, the client will undergo an intake to determine the treatment trajectory. Finally, the therapist will determine together with the client a treatment plan so that the treatment can start. If other issues show up during the treatment process, the client can also be addressed to another department. From there onwards the process will repeat itself, starting with placing the client on the waiting list of the new department (see Figure 1).

(19)

Figure 1

Flowchart of the triage process

(20)

Data Collection

In the following section the data collection and the structure of the questionnaire will be described, and information about the respondents will be provided.

Questionnaire and procedures. In this research quantitative data was gained with the help of questionnaires since the available database was missing information about the

dependent variable. The questionnaire was administered by post with the extra opportunity to complete the questions online. In the questionnaire the participants could find a link so they could answer the questions online. The online questionnaire was made with the online program ‘Qualtrics Survey Software’. The reasoning behind delivering the questionnaire at first by post and per by email is that sending the questionnaire per email would create a bias towards people who could not use eTriage, with the reason they did not have a working internet connection or an email address. To make it as easy as possible for the respondents to complete and return the questionnaire a return envelope was included, so that it did not bring delivery costs for the participants.

The questionnaire consisted of 17 self-constructed items to assess client service

satisfaction concerning the triage process. Besides that, an item concerning the clients’ idea of which triage they did, to test the assigned condition. For both age and gender, one item was included. The items were controlled by both the university supervisor and the mental

healthcare organization supervisor of this thesis. The questionnaire was tested on five fellow students, who provided feedback. All the feedback contributed to the improvement of the formulation the questions and of the introduction text. The introduction text was added to clarify what this research was about. The final introduction text and questionnaire can be found in Appendix A.

Participants were asked to answer the question on a 7-point Likert scale, the options being completely dissatisfied/disagree (1), largely dissatisfied/disagree (2), somewhat

(21)

21

dissatisfied/disagree (3), neither satisfied/agree nor dissatisfied/disagree (4), somewhat satisfied/agree (5), largely satisfied/agree (6) to completely satisfied/agree (7). The choice to use a 7-point Likert scale was made to prevent the participants from being too neutral in their answers (Coleman, Morris & Preston, 1997). Furthermore, a 7-point scale tends to be more reliable and more valid than smaller Likert scales (Alwin, 1997).

Participants. The questionnaire was sent to all clients of two specific departments which had had a triage in 2015, because they would have the most recent memory of the triage process. The total number of clients who received the questionnaire is 1317 specified to type of healthcare; 817 clients of specialized mental healthcare (PsyQ), 500 clients of the basic mental healthcare (Lentis). Of all the clients who received a questionnaire, 11.5%

started eTriage (152) and 88.5% started the offline triage. Altogether 135 clients responded to the questionnaire, this gives a response rate of 10.3%. Of the participants, 23.0% was male (31) and 77.0% female (104). The age of the participants was between 18 and 89 years, with an average of 40.74 year and a standard deviation of 17.34 year. Of the participants, 42.2%

used the offline triage (57) and 57.8% eTriage (78). The percentage of participants who responded from the basic mental healthcare was 43.0% (58) and subsequently the percentage of participants from the specialized mental healthcare was 57.0% (77).

Measurements. Client service satisfaction. This construct was measured on self- constructed items based on different dimensions from existing measurement scales, which are explained below. These dimensions were selected based on the suitability when evaluating the triage tool, since the dimensions should cover the content of the triage process. The first dimension was reliability and consists of two items were based on Wolfinbarger and Gilly (2002), Yang and Jun (2002) and Madu and Madu (2002). An example of an item used in this scale is “In hoeverre bent u tevreden over de zorgvuldigheid waarmee is omgegaan met uw gegevens?”. The dimension content was also taken into account, four items were created

(22)

22

based on Cox and Dale (2001) and Kaynama and Black (2000), an example item reads as follows “Hoe duidelijk vond u de vragen die u werden gesteld tijdens de aanmelding?”.

Accessibility was also taken into account when measuring client service satisfaction. This dimension was measured on two items based on Cox and Dale (2001) and Yang and Jun (2002). An example item of this scale is “In hoeverre bent u tevreden over het algehele contact met de zorginstelling?”. Based on Cronin, Brady, & Hult (2000) the last dimension sacrifice was measured on two items. One of these items reads as follows “De moeite die de aanmelding u heeft gekost?”.

Personal characteristics. The information about personal characteristics was obtained by asking about the participants’ age and gender. Furthermore, the information about what type of triage and what type of disorder the clients had was available before the questionnaire was sent. Therefore, these characteristics were used to send the participants slightly different questionnaires. The type of disorder was determined based on the logo of the paper that had been used, since the specialized and basic mental healthcare have their own logo on the questionnaire. The triage type was determined by the introduction text since this was different according to the triage type since a different introduction text was used.

Data Analysis

Factor analysis. First it is important to determine whether or not a dataset is suitable for a factor analysis, in order to do that sample size and strength of the relationship among variables should be considered (Pallant, 2013). For the sample size Nunnally (1967) recommends to have at least ten times as many participants as variables. Tabachnick and Fidell (1996) recommend to inspect the correlation matrix for correlations over 0.30 to check relationship between variables. Before extracting the variables the Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy and the Bartlett’s Test of Sphericity should be completed. For a suitable factor analysis the KMO index should be above 0.50 and the

(23)

23

Bartlett’s Test of Sphericity should be significant (p<.05) (Hair, Anderson, Tatham & Black, 1995; Tabachnick & Fidell, 1996). Furthermore, only the items that have a high factorial load, that is 0.6 according to Kline (2014), are added to the factors.

Exploring data & Correlation. To measure the correlation between variables, the type of data is important. In this thesis the data can be categorized as interval, since multi- item scales tend to behave as interval (Boone & Boone, 2012). According to Pallant (2013), Pearson correlation is the best measure in this situation. The correlation is interpreted based on the guidelines of Hemphill (2003), who states that correlations between .10 and .29 are small, correlations between .30 and .49 are medium and correlations ranging from .5 to 1.0 are large. The same accounts for negative correlations.

Reliability. Cronbach’s alpha, a method to measure the internal consistency, is according to Santos (1999) the most commonly used tool to measure reliability. According to Schmitt (1996) reliability of a Cronbach’s alpha coefficient between .7 and .8 is acceptable, between .8 and .9 is good and above .9 is excellent. However, coefficients between .5 and .6 are indicators of poor reliability and coefficients below .5 are identified as unacceptable.

According to Pallant (2013) this means that the Cronbach’s alpha coefficient should be above .7 to ensure sufficient internal consistence. Nonetheless, according to Streiner (2003) a Cronbach’s alpha over .90 is not desirable, because this means that the inter-correlation amongst the items is too high, in the sense that the items have too much overlap and are therefore redundant. Besides that it means that the construct is measured too specific.

Furthermore, a remark should be made concerning the internal consistency of Cronbach’s alpha, since Cortina (1993) states that a high value on the scale is strongly affected by the amount of items of the construct. Therefore a high value of alpha does not necessary mean that there is a high amount of internal consistency.

(24)

24

Analysis of variance. To test the hypotheses one-way and two-way analyses of variance (ANOVA) were performed. Before the analysis could be performed the many assumptions (Pallant, 2013) of ANOVA were checked: level of measurement, random sampling, independence of observations, normal distribution and homogeneity of variance.

The assumption of random sampling was not a problem for this research, since the scope of this research were clients from a specific period in time and all clients from this period were selected. However this fact was taken into account when drawing conclusions broader than this specific situation. The guidelines that Pallant (2013) describes were used to complete the ANOVA analyses. The same pertains to the performed moderator analysis.

Qualitative data analysis. The questionnaire finished with an open question, in which the participants could leave their remarks and/or suggestions. The question was: “Heeft u nog opmerkingen of suggesties voor verbetering van de aanmeldprocedure?”. The results were analyzed by using inductive reasoning, based on repetition of certain terms (see Appendix H).

Results Data Reduction

To test whether or not the hypothesized factors came to the fore a factor analysis was performed (see Appendix B: results, Table 2). The assumptions were checked with the KMO Measure of Sampeling Adequacy (.893) and Bartlett’s test of sphericity (p = .000) and

therefore supported the use of a factor analysis. The results from this analysis revealed three instead of four factors. Moreover, these factors covered different items than anticipated and are therefore named after the contents of the items in their specific factor. Factor 1

(confidence/clarity) exists out of eight items, with an eigenvalue of 9.493, this factor was accountable for 55.8% of the variance. Factor 2 (relevance) exists out of four items, with an eigenvalue of 1.239, this factor was accountable for 7.3% of the variance. Factor 3

(usability/method) exists out of three items, with an eigenvalue of 1.042, this factor was

(25)

25

accountable for 6.1% of the variance. Two items were not discriminating, which is why they were not placed within one of the factors. Based on the factor analysis the dimensions are measured on a reduced number of items, since some of the items produced too much distortion. The Cronbach’s Alpha for the three factors, confidence/clarity, relevance and usability/method were respectively .92, .78 and .84 (see Appendix B).

Descriptive Results

Table 2 shows the mean scores, standard deviations, correlations and Cronbach’s Alpha’s of the variables that were measured. The dimension confidence/clarity was positively related to relevance, r = .76, p < .01, usability/method, r = .59, p < .01 and type of healthcare, r = -.20, p = .03 but not to age, r = .15, p = .09, gender, r = -.13, p = .16 and triage-type, r = - .12, p = .17. This means that high client service satisfaction for the dimension

confidence/clarity goes along with high client service satisfaction in the relevance and usability/method dimension. Furthermore, when a client makes use of specialized mental healthcare, this goes along with client service satisfaction in the dimension confidence/clarity.

The dimension relevance was positively related to usability/method, r = .55, p < .01, but not to age, r = .11, p = .22, gender, r = -.16, p = .70, triage-type, r = -.11, p = .21 and type of healthcare, r = -.17, p = .06. This means that high client service satisfaction for the dimension relevance goes along with high client service satisfaction for the dimension usability/method.

The dimension usability/method was positively related to age, r = .21, p = .01 but not to gender, r = -.05, p = .54, triage-type, r = .06, p = .48 and type of healthcare, r = .01, p = .89.

This means that when age increases, the client service satisfaction for the dimension

usability/method also increases. Furthermore, age was negatively related to type of healthcare, r = -.18, p = .04, but not to gender, r = -.14, p = .11 and triage-type, r = -.14, p = .10. This means that albeit the slightly weak effect, older adults are less likely to be treated within the specialized mental healthcare.

(26)

26

Additionally, gender was not related to triage-type, r = -.02, p = .82 and type of healthcare, r

= -.11, p = .22. Finally, triage-type was positively related to type of healthcare, r = .64, p <

.01. This means that there is an increased amount of clients who used eTriage in specialized mental healthcare.

Table 2

Exploring data, correlations and Cronbach’s alpha (n = 126)

Note. Cronbach’s alpha is on the diagonal in parentheses.

*p < .05, **p < .01

1: male = 0, female = 1

2: Basic mental healthcare = 0, Specialized mental healthcare = 1

3: non-eTriage = 0, eTriage = 1

Analytic Results

To test the hypotheses, multiple one-way and two-way analyses of variance were performed (see Appendix C-G). The first hypothesis was tested by comparing the client service satisfaction of the clients who did use eTriage with the clients who did not use eTriage. The analysis showed non-significant differences between them on client service

Variable m SD 1a 1b 1c 2 3 4 5

1. Client service satisfaction

1a. Confidence/Clarity 5.93 1.11 (.92)

1b. Relevance 5.31 1.24 .76^** (.78)

1c. Usability/Method 5.52 1.21 .59^** .55^** (.84)

2. Age 40.74 17.34 .15 .11 .21^* -

3. Gender¹ - - -.13 -.16 -.05 -.14 -

4. Triage-type² - - -.12 -.11 .06 -.14 -.02 -

5. Type of healthcare³ - - -.20^* -.17 .01 -.18^* .11 .64^** -

(27)

27

satisfaction. The results can be specified in the underlying dimensions: confidence/clarity, F(1, 128) = 1.91, p = .17, relevance, F(1, 127) =1.61, p = .21, usability/method, F(1, 132) = .74, p = .48 (see Appendix C). Thus, the first hypothesis is rejected.

The second hypothesis was tested by comparing the client service satisfaction of older and younger adult clients within both the groups that did and that did not use eTriage. Two groups were created because the assumption of equality of error variances (Levene’s test) was not met when these age groups were not made. The two groups were separated based on the mean and median age (respectively 40.70 and 40.50), therefore the lower age group included ages from 18 to 39 (N=66) and the higher age group included ages between 40 and 89 (N=68). The lower age group consisted of 16.7% men (11) and 83.3% women (55), 39.4%

used offline triage (26) and 60.6% used eTriage (40), finally, 39.4% of the clients (26) made use of basic mental healthcare and 60.4% made use of specialized mental healthcare. The percentages for the triage-type and type of healthcare are the same, but when examined, the similarity of the age groups is a coincidence, since not all eTriage clients used specialized mental healthcare. The higher age group consisted of 20.4% men (20) and 70.6% women (48), 48.5% used offline triage (33) and 51.5% used eTriage (35) and finally, 51.5% of the clients (35) made use of basic mental healthcare and 48.5% of clients (35) made use of specialized mental healthcare. The analysis showed non-significant differences in client service satisfaction with the triage procedure between the two age groups. The results can be specified in the underlying dimensions: confidence/clarity, F(1, 123) = 1.50, p = .22,

relevance, F(1, 122) =1.13, p = .29, usability/method, F(1, 127) = 2.96, p = .09. Furthermore, the analysis showed non-significant interaction-effects between age and type of triage. The results can be specified in the following interaction-effects: confidence/clarity (F(1, 123) = .23, p = .64), relevance (F(1, 122) = .07, p = .80) and usability/method (F(1, 127) = 1.93, p = .17) (see Appendix D). Thus, the second hypothesis is rejected.

(28)

28

The third hypothesis was tested by comparing client service satisfaction between male and female clients within both the group who did and who did not use eTriage. The analysis showed non-significant differences in client service satisfaction between the two gender groups within the group who used eTriage. The results can be specified in the underlying dimensions: confidence/clarity, F(1, 124) = 1.91, p = .17, relevance, F(1, 123) = 3.65, p = .06, usability/method, F(1, 128) = 0.44, p = .51. Additionally, the analysis showed non-significant interaction-effects between age and type of triage. The interaction-effects can be specified into the underlying dimensions: confidence/clarity (F(1, 124) = .16, p = .69), relevance (F(1, 123) = .09, p = .77) and usability/method (F(1, 128) = .22, p = .64) (see Appendix E).

Consequently, the third hypothesis is rejected.

The fourth hypothesis was tested by comparing client service satisfaction based on the type of mental healthcare the clients got. The clients who were assigned to specialized mental healthcare were compared with the clients who were assigned to basic mental healthcare within both the eTriage and the non-eTriage group. The analysis showed non-significant differences in client service satisfaction between the two mental healthcare groups. The results can be specified in the underlying dimensions: confidence/clarity, F(1, 126) = 3.45, p

= .07, relevance, F(1, 125) = 1.95, p = .17, usability/method, F(1, 130)= 0.50, p = .48.

However, the analysis showed a significant interaction-effect between type of mental healthcare and type of triage, for the client service satisfaction dimension usability/method (F(1, 130) = 6.21, p = .01). The other interaction-effects can be specified into the underlying dimensions: confidence/clarity (F(1, 126) = 1.01, p = .32) and relevance (F(1, 125) = .09, p = .77). The significant interaction-effect between type of mental healthcare and type of triage, for the client service satisfaction dimension usability/method is further examined in a table and plot (see table 3 and figure 2). This revealed that basic mental healthcare clients who used eTriage were less satisfied (usability/method) (m = 5.17) than specialized mental healthcare

(29)

29

clients (m = 5.67). This in contrast to basic mental healthcare clients who did not use eTriage, they were more satisfied (m = 5.60) than the clients of specialized mental healthcare (m = 4.70) (see table 3 and figure 2). When these differences were further examined with a split file (separation based on Type of Healthcare) One-Way ANOVA, the results showed a significant difference in client service satisfaction for specialized mental healthcare between the two triage types, F(1, 74) = 8.10, p = .03. However, for basic mental healthcare no significant difference in client service satisfaction was found between both triage types, F(1, 58) = 1.53, p = .22 (see Appendix F).

Table 3

Interaction effect between Type of Healthcare and Triage-type

Type of Healthcare (I) Triage-type (J) 95% Confidence Interval

n Mean Std. Deviation Lower Bound Upper Bound

Basis GGZ (Lentis) Offline 47 5.596 .174 5.251 5.940

Online 12 5.167 .345 4.485 5.849

Specialistische GGZ (PsyQ)

Offline 10 4.700 .378 3.953 5.447

Online 65 5.667 .148 5.374 5.960

(30)

30 Figure 2

Interaction effect between type of disorder and triage-type on client service satisfaction (usability/method)

The fifth and last hypothesis was tested by comparing client service satisfaction for each group of clients who were treated by one specific therapist within all (separate) clients’

characteristics groups. The analysis showed a non-significant difference in client service satisfaction between all therapists. The results can be specified in the underlying dimensions:

confidence/clarity, F(43, 86) = 1.07, p = .38, relevance, F(43, 85) = 0.84, p = .73,

usability/method, F(45, 88)= 1.18, p = .25. Furthermore, between triage-type and resistance non-significant interaction-effects were found, and can be specified in the following

underlying dimensions: confidence/clarity, F(9, 76) = 0.85, p = .58, relevance, F(10, 74) = 0.83, p = .61, usability/method, F(10, 77)= 1.31, p = .24. Between age and resistance non- significant interaction-effects were found, and can be specified in the following underlying dimensions: confidence/clarity, F(14, 70) = 0.67, p = .79, relevance, F(15, 68) = 0.57, p = .88, usability/method, F(15, 71)= 0.53, p = .91. Between gender and resistance non-significant interaction-effects were found, and can be specified in the following underlying dimensions:

4,6 4,8 5 5,2 5,4 5,6 5,8

Basis GGZ (Lentis) Specialistische GGZ (PsyQ)

Satisfaction (usability/method)

Offline Online

(31)

31

confidence/clarity, F(8, 77) = 0.42, p = .91, relevance, F(8, 76) = 0.75, p = .65,

usability/method, F(8, 79)= 0.32, p = .96. Finally the analysis could not show interaction- effects between type of disorder and resistance, because no therapist treated clients in both basic and specialized mental healthcare (see Appendix G). According to the results of this thesis, the viewpoint of a therapist towards eTriage does not play a role. This is because there was no difference in the client service satisfaction between the groups of clients from

different therapists. Therefore, even if different therapists would have different levels of resistance (not researched), there is no difference between client service satisfaction of the clients of the different therapists. Consequently, the fifth hypothesis is rejected.

Open Question

In the open question where clients could leave remarks and suggestions for improving the registration process, 57 clients (42.2%) answered this question (see Appendix H). Two main issues were raised. The first issue was about the waiting time between the intake and the start of the treatment itself. In 42.1% of the cases, the clients indicated that the waiting times were too long or they recommended to shorten these waiting times. The second issue

concerned clarification of the registration process. Over ten percent of the clients indicated that they did not have enough information about the procedure and the goal of the procedure, or thought the procedure and/or questionnaire was not clear enough.

Discussion

The next section discusses the findings of this research. Thereafter, theoretical implications and limitations of this study are presented, followed by the practical

implications. Subsequently, the end of this section includes suggestions for future research and closes with a conclusion.

(32)

32 Summary & Interpretation

By studying the client service satisfaction concerning the registration process at two mental healthcare organizations, this study aimed at answering the following research question: “Is eTriage suitable for all types of mental healthcare clients and disorders and how does the role of the therapist contribute to client service satisfaction?”. The clients who used eTriage were expected to be more satisfied with the intake process as compared with the clients who did not use eTriage. Furthermore, a difference between age groups was also expected. More precisely, younger adults were expected to be more satisfied when using eTriage; and when eTriage was not used, older adults were expected to be more satisfied than younger adults. Concerning gender it was expected that female clients, compared to male clients, were more satisfied when using eTriage and less satisfied when not using eTriage.

Clients who were assigned to specialized mental healthcare were expected to be less satisfied when using eTriage, and more satisfied when not using eTriage, when compared to basic mental healthcare. Finally resistance of therapists towards the use of eTriage was expected to weaken the positive relationship and reinforce the negative relationship between client service satisfaction and the clients’ characteristics.

The results did not show significant differences in client service satisfaction between the clients that used eTriage and the clients who did not use eTriage, which means that client groups researched in this study are equally satisfied with the non-digitalized triage service and with eTriage. Consequently, eTriage is a good option for the non-digitalized triage, and has potential to be used either next to or as replacement for the non-digitalized triage.

Furthermore, no differences were found between younger and older clients in client service satisfaction, and the same accounts for male and female clients. This implies that the gender and age groups can be treated in a similar way and no major adjustments in the triage process need to be made according to the age and gender of the clients. Additionally, when the type of

(33)

33

disorder was taken into account as well, difference in one of the underlying dimensions of client service satisfaction (usability/method) was found. This means that the clients who were assigned to the specialized mental healthcare were more satisfied with the triage process when the registration was taken online instead of offline. The degree to which the clients need help is dependent on the type of disorder, since there is a higher need within the specialized mental healthcare, which could have a connection with their increased satisfaction with eTriage. This might apply since the clients with a higher need of help might have the idea that they are helped on a short notice with eTriage and therefore are more satisfied about eTriage. Which could mean that eTriage suits clients in the specialized mental healthcare better than the non- digitalized version. More specifically it could suit the clients from specialized mental

healthcare better due to the fact that they can fulfill the triage process at a location of their own choice. The results showed no difference in client service satisfaction between the groups of clients who were treated by different therapists, which implies that the therapist does not influence the clients’ satisfaction. Therefore no additional attention is needed to improve the clients’ satisfaction due to the viewpoint of the therapists on eTriage. It was not researched in this study whether the therapists had a positive or negative attitude towards eTriage.

However, there was no difference in the client service satisfaction between the groups of clients that had been treated by different therapists.

Theoretical Implications

In this section the theoretical implication of the findings of this study will be explained per hypothesis. Firstly, the increase in client service satisfaction due to the use of eHealth technology that was described by Maillet, Mathieu and Sicotte (2015) could not be confirmed in this study. This could be the case since the clients are satisfied with the fact that they are getting the help they need and see it as a necessity whether or not via it happens eTriage, which could be a reason for why no client service satisfaction difference between clients who

(34)

34

used eTriage and who did not was found. Next to that it could be of importance that the clients are not provided with different registration processes, and therefore they remain uninformed and unaware of the options they could have. This might mean that they have no prior knowledge of other methods to complete the registration and because of that it did not influence the client service satisfaction. For example, had a client from the group who did not use eTriage, who had to take public transport for more than two hours, known it would have been possible to complete the registration online, he could have been less satisfied with the original registration process. An other difference between this study and the research of Maillet, Mathieu and Sicotte (2015) is that in the latter the results showed an increase in user satisfaction over time when eHealth technology was used. However, in this study it was not possible to track the client service satisfaction over time at clients who did not use eTriage before and then switched to using eTriage.

Different groups of end-users seem to react differently to eHealth. For instance in the study of Maillet, Mathieu and Sicotte (2015) the nurses, who are the end-users of eHealth, showed an increase in satisfaction when using eHealth. However, the results of this paper showed that clients, although they are also end-users of eHealth, were not more or less satisfied when eHealth was used. This might be because even though they are end-users, clients do not need to use eTriage every day. They will probably use eTriage only once, whereas nurses have to use this system more often. Furthermore, due to the fact that clients do not use eTriage regularly, the positive effects of such an eHealth technology do not show up over such a short period of time. Therefore, based on the results of this study it can be concluded that not all end-users will react in a similar way on eHealth. More specifically, clients who used eTriage do not show an increase in client service satisfaction probably due to their limited use of the eHealth system. For instance, since the clients only use eTriage once,

(35)

35

they have no additional knowledge of other systems and therefore cannot compare the systems.

The results concerning age and triage-type were not in line with the hypothesis. The reason for this might be that the literature does not show a clear answer regarding age and technology. Arning and Ziefle (2007) conclude that older adults have a more pragmatic view or even a rejecting attitude towards technology. However, the same authors found later that older adults assess eHealth technologies more positively than younger adults, since with increasing age adults tend to acknowledge the benefits of eHealth (Arning & Ziefle, 2009).

Nonetheless, Hing and Hsiao (2010) contradict these findings again by stating that the use of eHealth facilities is inversely associated with the physicians’ age. Additionally, Maillet, Mathieu and Sicotte (2015) have shown that the actual use of eHealth is positively related to end-user satisfaction. However, as described before these results apply for nurses, who are the end-users just like clients in this study, but the nurses and the clients do not seem to react the same way to eHealth, although both are end-users. This might be because they use eHealth in different ways.

These findings suggest that clients in mental healthcare cannot be easily compared to other end-user groups of eHealth (e.g. physicians, employees). Again this might be because of the limited amount of time that clients spend on using eTriage. This could also mean that there is also limited amount of time that can constraining factors might bother the triage process and therefore do not play a major role, this in comparison to the everyday use of eHealth technology. Furthermore, the relationship between age and technology, and therefore eHealth, could also be changing these days, since technology continues to play a larger role in our lives and therefore requires people to function and cooperate with it. Maillet, Mathieu and Sicotte (2015) have shown that the actual use of a healthcare technology is positively related with user satisfaction, but since it could be that all age categories are getting to the same level

(36)

36

in their use of technology, for instance in how often they use technology, which could be the reason why the expected results were not found, specifically there was no difference in client service satisfaction for both age groups.

In this study no difference has been found between the age of the clients on client service satisfaction. The reason might be that again in research the clients within the mental healthcare cannot be compared with other end-users, due to their different roles. For instance the actual use of eTriage could be different, since clients rely on eTriage to get rid of their complaints. Furthermore the clients who are not able to use eTriage, since they do not have the resources to use it, will not be assigned to eTriage. This is different to a work site were you have to use eHealth systems and often cannot reject it as an end-user. In the research of Arning and Ziefle (2009) the older participants were relatively young compared to the actual average age of the clients in mental healthcare. For the literature this could mean that different age groups in mental healthcare can be treated in a similar way, since no difference in client service satisfaction according to age has been found in this study. This is because it could be that both age groups see it as a necessity, clients do not experience the negative consequences since they have to use it only for a limited amount of time, and the clients who use eTriage also have the resources to do so.

Additionally the results and the literature concerning gender were not in line with each other either. Females were expected to be more into technology and therefore also into

eHealth applications. Yu, Shen and Lewark (2012) showed that both male and female

employees’ satisfaction increased during the implementation of an IT system, but for women it was due to successful implementation of the technological innovation. Furthermore, female students were more satisfied than male students with an electronic version of learning

(Gonzalez-Gomez, Guardiola, Rodriguez, & Alonso, 2012). However, the literature is not unanimous concerning gender, because in an earlier study females were less like to use