Before taking on an assignment, you will review the content and only use the attached references and resources. No plagiarism or original work is to be done. NO OUTSIDE SOURCES ALLOWED!!

Journal of Biomedical Informatics 53 (2015) 342–354

Contents lists available at ScienceDirect

Journal of Biomedical Informatics

journal homepage: www.elsevier .com/locate /y jb in

Clinical information systems end user satisfaction: The expectations and needs congruencies effects

http://dx.doi.org/10.1016/j.jbi.2014.12.008 1532-0464/� 2014 Elsevier Inc. All rights reserved.

⇑ Corresponding author. Fax: +61 293518642. E-mail addresses: [email protected] (F. Karimi), [email protected].

edu.sg (D.C.C. Poo), [email protected] (Y.M. Tan). 1 Current address: Complex Systems Research Group, Project Management Pro-

gram, Faculty of Engineering & IT, The University of Sydney, NSW 2006, Australia. 2 Fax: +65 67791610.

Faezeh Karimi a,⇑,1, Danny C.C. Poo a,2, Yung Ming Tan b

a School of Computing, National University of Singapore, 13 Computing Drive, Singapore 117417, Singapore b Khoo Teck Puat Hospital, Singapore

a r t i c l e i n f o a b s t r a c t

Article history: Received 3 May 2014 Accepted 15 December 2014 Available online 24 December 2014

Keywords: Clinical information systems End user satisfaction Expectations congruency Needs congruency

Prior research on information systems (IS) shows that users’ attitudes and continuance intentions are associated with their satisfaction with information systems. As such, the increasing amount of invest- ments in clinical information systems (CIS) signifies the importance of understanding CIS end users’ (i.e., clinicians) satisfaction. In this study, we develop a conceptual framework to identify the cognitive determinants of clinicians’ satisfaction formation. The disconfirmation paradigm serves as the core of the framework. The expectations and needs congruency models are the two models of this paradigm, and perceived performance is the basis of the comparisons in the models. The needs and expectations associated with the models are also specified. The survey methodology is adopted in this study to empir- ically validate the proposed research model. The survey is conducted at a public hospital and results in 112 and 203 valid responses (56% and 98% response rates) from doctors and nurses respectively. The par- tial least squares (PLS) method is used to analyze the data. The results of the study show that perceived CIS performance is the most influential factor on clinicians’ (i.e., doctors and nurses) satisfaction. Doctors’ expectations congruency is the next significant determinant of their satisfaction. Contrary to most previ- ous findings, nurses’ expectations and expectations congruency do not show a significant effect on their satisfaction. However, the needs congruency is found to significantly affect nurses’ satisfaction.

� 2014 Elsevier Inc. All rights reserved.

1. Introduction

Healthcare delivery systems are attributed with undesirable characteristics such as unsafe practices, treatment variability, and less than desirable quality of care [1]. Hence, there is great pressure to increase healthcare quality and patient safety while also reduc- ing healthcare costs. Healthcare information technologies (HIT), especially clinical information systems (CIS), have been frequently suggested as efficient means of achieving healthcare quality, patient safety, reduced medical errors, and decreased costs [2–4]. However, the realization of all these benefits is contingent upon clinicians’ continuous use of these systems [5]. Prior research on information systems (IS) has found that users’ attitudes and con- tinuance intentions are associated with their satisfaction with information systems [6,7]. Clinicians’ satisfaction has also been

found to explain intended future use of CIS [8]. In addition, user satisfaction has been studied in relation to other important IS con- cepts such as user acceptance, user resistance and system usability [9–11].

From another perspective, with the increasing amount of investments in CIS implementations, the importance of measuring the effectiveness of these investments has also been increasing. User satisfaction as a subjective or perceptual measure of IS suc- cess is probably the most widely employed IS effectiveness indica- tor in the IS success literature [12]. In their review of the literature on inpatient CIS evaluations, van der Meijden et al. [13] found that user satisfaction was evaluated in 46% of the studies. In fact, user satisfaction shows a system from its users’ point of view. When a ‘‘good’’ information system is viewed by its users as a ‘‘poor’’ sys- tem, it is in fact a ‘‘poor’’ system [14]. In other words, it is hard to deny the success of a system that is liked by its users [12].

In ‘‘Behavioral Theory of the Firm’’, Cyert and March [15] were the first to introduce the user satisfaction concept by proposing that the success of an information system at meeting the information needs of its users leads to either reinforcement or frustration [14,16]. Since this study, extensive research has been conducted in IS literature on user satisfaction (for comprehensive reviews, see [17] or [18]). This

F. Karimi et al. / Journal of Biomedical Informatics 53 (2015) 342–354 343

body of research (e.g., [6,12]) partly dealt with the role of user satis- faction in IS success [19]. Many other studies attempted to develop measurements for assessing user satisfaction [14,16,20–22] or pro- vide typologies of information system characteristics that seem to affect user perception of the system [12,23,24]. They offered practi- tioners with helpful tools to improve their systems and chances of success [25]. Nevertheless, this body of research has been frequently criticized because of its overemphasis on IS technical soundness and specific system characteristics, and that it offers only a limited understanding of the psychological processes that convert a sys- tem’s performance into a user’s reaction to the system [26–28]. Fur- thermore, most of the research is based on empirical findings rather than a theoretical background [29].

Moreover, while several IS studies investigated various aspects of IS usage within the specific context of healthcare such as clini- cians’ resistance towards [3,30] or acceptance of healthcare infor- mation technologies [31], the clinicians’ satisfaction with clinical information systems is under-researched in the IS field. Neverthe- less, the medical informatics literature on CIS satisfaction has examined how clinicians’ satisfaction is affected by different fac- tors such as user characteristics [32–34], computer literacy [33,35], CIS quality, CIS usefulness, service quality [34], impact on work, and impact on patient care [32,33]. These studies pro- vide a useful understanding of clinicians’ satisfaction with a vari- ety of CIS functionalities and satisfaction differences among various user groups. However, like IS user satisfaction research, they usually lack a strong theoretical basis and focus mostly on the technical aspect of CIS. The necessity of further investigation of clinicians’ satisfaction is intensified with respect to the find- ings of previous healthcare information technology acceptance studies indicating that clinicians differ from other types of IS users due to their specialized training, autonomous practice and professional work arrangements [31,36,37].

With regard to the importance of clinicians’ satisfaction in suc- cessful CIS implementations and the need for understanding the psychological processes of satisfaction formation, we aim to develop a conceptual framework to identify the cognitive determi- nants of clinicians’ satisfaction formation based on the disconfir- mation paradigm. This paradigm is the major theory utilized in marketing literature to explore consumer satisfaction [28,38,39]. Hence, the research questions for this study are: (1) ’’What are the cognitive determinants of clinicians’ satisfaction formation with CIS?’’ and (2) ‘‘What are the clinicians’ needs and expectations regarding CIS?’’. Arguing that clinicians’ needs (desires) and expec- tations regarding CIS are two separate satisfaction comparison standards, we will examine their effects jointly in the disconfirma- tion paradigm. To identify the clinicians’ needs, we will employ McClelland’s learned needs theory [40]. The extant IS literature on user satisfaction will also be integrated into the research model of the study. As such, various IS attributes including system qual- ity, information quality, and service quality from Delone and McLean’s [23] IS success model will be utilized as the aspects of the system which clinicians may have expectations about. The impact of perceived CIS performance at the functionality level on clinicians’ satisfaction will be examined as well. The variety of information systems and the conflicting human interests may call for different assessments of a system’s impact and effectiveness in order to capture the different stakeholders’ points of view [18,24]. Hence, the proposed model explaining/predicting clinicians’ satis- faction will be tested among two different clinical user groups (i.e., nurses and doctors) to observe any plausible differences in a practical manner.

The rest of this paper is organized as follows. First, the theoret- ical background of the study is explained. Next, the research model and hypotheses are presented. This is followed by an explanation of the research methodology. Then, the data analysis and results

are presented. The study concludes with a discussion of the find- ings and their implications for theory and practice.

2. Theoretical background

2.1. Definitions

Healthcare information technologies (HIT) span various appli- cations to serve different purposes in healthcare settings. Bhatt- acherjee et al. [41] developed a HIT classification based on the primary purposes of different HIT applications. Similar groupings have been extensively validated and utilized in prior relevant research. This categorization includes clinical HIT, administrative HIT, and strategic HIT. In the present study, CIS refers to the clinical cluster of this categorization representing applications designed to improve patient care, such as computerized physician order entry (CPOE) systems, electronic medical records (EMR), and pharmacy information systems.

Briggs et al. [25] identified three categories of user satisfaction definitions in the IS literature including satisfaction as judgment, satisfaction as affect, and a mixed definition with both judgment and affect elements. An example of a definition framing satisfac- tion as a judgment or evaluation is ‘‘the extent to which users believe the information system available to them meets their infor- mation requirements’’ [14, p. 785]. Some examples of definitions framing satisfaction as an affective state include ‘‘the affective atti- tude towards a specific computer application by someone who interacts with the application directly’’ [21, p. 261], and ‘‘an affec- tive state representing an emotional reaction to the entire Web site experience’’ [42, p. 298]. An example of a definition considering both affective and evaluative components for satisfaction is ‘‘the IS end-user’s overall affective and cognitive evaluation of the plea- surable level of consumption related fulfillment experienced with the IS’’ [29, p. 453].

In this study, we adopt the second definition and conceptualize clinician satisfaction with CIS as an affective state representing an emotional reaction to a CIS that a clinician directly interacts with. The reason for this choice lies in the fact that a user with a positive evaluation of an IS might still not feel satisfied with it [25]. Fur- thermore, the evaluation component will be captured by the two determinants of satisfaction (i.e., needs and expectations congru- encies) which will be discussed later in the paper.

2.2. Disconfirmation paradigm

In order to identify the cognitive determinants of clinician sat- isfaction formation we draw on the disconfirmation paradigm. The disconfirmation paradigm is the dominant framework for explain- ing consumer satisfaction in marketing literature [28,38,39]. This paradigm considers satisfaction the result of an evaluative judg- ment between the perceived performance of a product and a pre- consumption comparison standard. Three different states can occur after this comparison process: (1) positive disconfirmation: when the perceived performance is above the comparison stan- dard, (2) confirmation: when the perceived performance meets the comparison standard, or (3) negative disconfirmation: when the perceived performance is below the comparison standard. Sat- isfaction is more likely when positive disconfirmation or confirma- tion occurs, while dissatisfaction is expected when negative disconfirmation occurs [38,39].

2.2.1. Expectations congruency (disconfirmation) model Different studies have examined the effects of various compar-

ison standards such as predictive expectations [43,44], desires and needs [27,45], and experience-based norms [46,47]. However, the

344 F. Karimi et al. / Journal of Biomedical Informatics 53 (2015) 342–354

most common pre-experience comparison standard in the discon- firmation paradigm is expectations [38,46].

Based on the expectations congruency model3, satisfaction is the result of comparing the perceived performance of a product to the expectations about that product’s performance. When the perceived performance exceeds expectations satisfaction arises, while the per- ceived performance falling behind expectations result in dissatisfac- tion [27,39]. Marketing literature provides substantial empirical support for the influence of expectations congruency on consumer satisfaction [27]. Hence, one part of our research model examines the effect of clinicians’ expectations and expectations congruency on their satisfaction with CIS.

2.2.1.1. Definition of expectations. Despite the wide inclusion of expectations in consumer satisfaction research, there is no consen- sus on the conceptual definition of this construct [27]. Spreng et al. [27] identified two different conceptualizations for expectations. The first view (i.e., predictive expectations) defines expectations as ‘‘primarily perceptions of the likelihood (or probability of occur- rence) of some event’’ (p. 16). The second view (i.e., evaluative expectations) adds another component to this likelihood estima- tion that is ‘‘an evaluation about the goodness or badness of the event’’ (p. 16). This judgment component can be misleading and bias the effect of expectations on satisfaction. Similarly, in the IS literature, Szajna and Scamell’s [48] review of expectations defini- tions in social psychology and organizational behavior research recognized two components of expectations: (1) ‘‘a future time perspective’’ and (2) ‘‘a degree of uncertainty’’ (p. 494). Following the predictive conceptualization, we define expectations as an information system end user’s set of pretrial beliefs about the eventual performance and attributes of the information system. To identify these attributes we rely on the prior IS user satisfaction literature, specifically Delone and McLean’s [23] IS success model. This model provides three categories of IS attributes including information quality, system quality, and service quality. They will serve as the aspects of a system that clinicians may have expecta- tions about.

2.2.2. Needs congruency (disconfirmation) model Although the expectations congruency model is widely

employed in the consumer behavior research and in some IS end user satisfaction studies, the model has its shortcomings. First, it fails to explain dissatisfaction when low expectations are con- firmed [27]. Second, the disconfirmation effect can only account for the aspects of the product for which consumers hold prior expectations, although consumers may also be dissatisfied with the unexpected aspects after consumption [28]. One suggested method for addressing these shortcomings is to use perceived actual performance as an additional antecedent of satisfaction [44], while another is to utilize a different comparison standard, i.e., one’s values (or needs, desires, wants) [45]. Therefore, another part of this study investigates the effects of both the needs congru- ency model and perceived CIS performance on clinicians’ satisfaction.

2.2.2.1. Definition of needs. There is also no overall accepted defini- tion for the desires construct in the marketing literature. Spreng et al. [27] attributed this lack of consensus to the various possible levels of abstraction for conceptualizing desires. In a means-end framework, Spreng et al. [27] explain that desires ’’can be defined abstractly in terms of the most basic and fundamental needs, life goals, or desired end-states or more concretely in terms of the

3 In the rest of this study, when a specific comparison standard is used in the disconfirmation paradigm, we refer to it as the standard congruency (disconfirma- tion) model. This is a common practice in the relevant literature.

means that a person believes will lead to the attainment of the desired end-states’’ (p. 16). In the context of end user satisfaction of information systems, it will be more useful to explore the influ- ence of higher-level desires on end user satisfaction because there has already been extensive research on the various attributes of information systems that can affect end user satisfaction. Further- more, prior IS studies [19] have shown that the conversion of higher-level desires into concrete product attributes (especially in the case of complex IS such as CIS) is not easy or straightforward for different IS stakeholders.

To identify these higher-level needs (desires), we reviewed needs theories in the psychology literature. Major needs theories include Maslow’s needs-hierarchy theory [49,50], Alderfer’s ERG theory [51,52], Murray’s manifest needs theory [53], and McClel- land’s learned needs theory [40]. McClelland’s learned needs the- ory is one of the most popularly accepted theories of motivation [40,54] and has been extensively employed in research in various job settings including healthcare environments [55,56]. This theory recognizes individual differences but also has specified content (specific needs categories). Hence, it offers considerable promise of explanatory power [57]. For this reason, we utilize it in our framework as the theoretical guide for identifying clinicians’ higher-level needs to explore their satisfaction with clinical infor- mation systems.

2.2.2.2. McClelland’s learned needs theory. Building on Murray’s manifest needs theory [53], McClelland [40] proposed the learned needs theory. According to this theory, individuals acquire certain needs through their life experiences in their culture. Four of these learned needs are the need for achievement, need for affiliation, need for power, and need for autonomy. Once these needs are acquired, they may be considered as personal predispositions that influence an individual’s perception of work situations and goal setting behavior [58]. These needs are considered important for understanding people at work [59]. Hence, in this study we will examine the role of CIS performance congruency with these needs on clinicians’ satisfaction with CIS.

2.2.2.3. Need for achievement. The need for achievement has been defined as ‘‘behavior towards competition with a standard of excellence’’ [58, p. 11]. It is ‘‘the drive to excel, to achieve in rela- tion to a set of standards and to strive to succeed’’ [60, p. 349]. With the increasing emphasis on patient-centered care, healthcare institutions such as hospitals are striving to achieve the best possi- ble patient outcome. CIS are health information technologies designed for the purpose of improving patient care [41]. Further- more, previous belief elicitation research on physicians about using EMR and CPOE shows that they believe these CIS influence (posi- tively or negatively) their performance, productivity and efficiency, and patient outcomes [61]. Due to the importance of achieving the best possible patient outcomes in the healthcare community, our research model investigates the congruency of the CIS with clini- cians’ need for achievement in terms of patient outcomes. These patient outcomes include (1) patient satisfaction, (2) healthcare quality, and (3) medical error occurrence (patient safety).

2.2.2.4. Need for affiliation. In the ‘‘Handbook of Social Psychology’’ [62], affiliation is defined as ‘‘the act of associating or interacting with one or more other people’’ (p. 465). The multidimensional model of affiliation [63] proposes four primary reasons (or social rewards) for people’s desire to affiliate: ‘‘(1) positive affect or stim- ulation associated with interpersonal closeness and communica- tion, (2) attention or praise, (3) reduction of negative affect through social contact, and (4) social comparison’’ (p. 1008). The reason that is most relevant to the context of CIS use in healthcare is social comparison. Social comparison ‘‘involves the seeking of

F. Karimi et al. / Journal of Biomedical Informatics 53 (2015) 342–354 345

information about a self-relevant issue from others when objective criteria for evaluation are not readily available, particularly with respect to opinions, beliefs, and other socially relevant attributes’’ [63, p. 1009]. This information can be helpful for reducing ambigu- ity, uncertainty and confusion, and improving response to relevant situations [62,63]. Individuals are increasingly dependent on the exchange of information in the workplace to carry out their job-related tasks. For instance, communication takes 80% of a healthcare manager’s time [64]. Moreover, patient–clinician and clinician–clinician communication shapes an essential component of clinicians’ jobs. Prior studies have shown that CIS affects interac- tions and communication among these social actors in healthcare settings [8,65]. In the context of EMR and CPOE, better documenta- tion is believed by physicians to improve the communication between colleagues and nurses [61]. Therefore, the social compar- ison dimension of the need for affiliation among clinicians and the CIS congruency in facilitating the fulfillment of this need will be investigated in this study.

2.2.2.5. Need for autonomy. In his ‘‘Explorations in Personality’’ book, Murray [53] states that the five sets of needs, including the need for dominance, need for autonomy, need for aggression, need for deference, and need for abasement, can be taken together. The common concept among these needs is the element of control. The need for power and need for autonomy are of interest to this study. While the need for power is concerned with controlling other peo- ple, the need for autonomy is about controlling one’s way of work- ing. The system investigated in our study did not impose any changes in the power distribution in the hospital for either nurses or doctors. Hence, the need for power is not included in the survey.

In contrast, the system did change how doctors deal with order- ing laboratory or radiology tests and consequent tasks. That is, it tampered with their autonomy. For nurses, the system increased the convenience of carrying out investigation orders, but no work control restrictions were inflicted. Hence, the need for autonomy is only included in the doctors’ questionnaire. Before proceeding to the next section, it is worth clarifying the distinction between expectations and desires.

2.2.3. Expectations and needs distinction Expectations and needs (desires) are conceptually different. The

simple distinction between them identified in the information sys- tems and marketing research literature is that ‘‘expectations state what the individual thinks will happen, while desires represent what the individual would like to happen’’ [19, p. 300]. Spreng et al. [27] further elaborated that ‘‘expectations are beliefs about the likelihood that a product is associated with certain attributes, benefits, or outcomes, whereas desires are evaluations of the extent to which those attributes, benefits, or outcomes lead to the attainment of a person’s value. Expectations are future-ori- ented and relatively malleable, whereas desires are present-ori- ented and relatively stable’’ (p. 17). Our theoretical framework therefore recognizes expectations and needs (desires) as two dis- tinct comparison standards in the disconfirmation paradigm.

3. Research model

This section presents a research model based on the theoretical framework discussed in the previous chapter. Table 1 defines each construct and Fig. 1 depicts their relationships.

3.1. Expectations and expectations congruency

Expectations have long been the dominant comparison stan- dard in the disconfirmation paradigm in the marketing literature

[46], and satisfaction has been considered to result from a low dis- crepancy between the pretrial expectations and the post hoc per- ceptions [39]. Similar to Spreng et al. [27], this study defines clinician expectations congruency as a clinician’s subjective assess- ment of the comparison between his or her expectations of a CIS and the CIS performance received. In turn, clinician expectations are conceptualized as a clinician’s set of pretrial beliefs about the eventual performance and attributes of a CIS. The positive associa- tion of expectations congruency and satisfaction has received con- siderable empirical support in the marketing literature [27]. Such an association has also been hypothesized in some IS satisfaction research [19,66]. We hypothesize a similar positive effect of clini- cians’ expectations congruency on their satisfaction with CIS:

H1. Clinician expectations congruency is positively related to clinician satisfaction with CIS.

The extant marketing literature also proposes a negative rela- tionship between expectations and expectations congruency. This is because high expectations are more likely to be negatively dis- confirmed and low expectations are more likely to be positively disconfirmed [27,39,67]. This association has been included in a number of IS studies in different contexts such as satisfaction with knowledge management systems [19], end user computing satis- faction [66], and measuring Web-customer satisfaction [42]. Simi- larly, we hypothesize that:

H2. Clinician expectations are negatively related to clinician expectations congruency.

In addition to the indirect effect of expectations on satisfaction through expectations congruency, a direct positive impact of expecta- tions on satisfaction has been postulated in prior marketing research [44]. The results of a number of IS studies have provided empirical support for the effect of IS users’ expectations on their overall satisfac- tion with the IS [68,69]. A longitudinal experiment by Szajna and Scamell [48] showed an association between the realism of IS users’ expectations and their satisfaction with the IS. A meta-analysis by Mahmood et al. [70] also found a strong correlation between the two constructs. On these grounds, the next hypothesis is:

H3. Clinician expectations are positively related to clinician satis- faction with CIS.

3.2. Needs and needs congruency

Collepals.com Plagiarism Free Papers

Are you looking for custom essay writing service or even dissertation writing services? Just request for our write my paper service, and we'll match you with the best essay writer in your subject! With an exceptional team of professional academic experts in a wide range of subjects, we can guarantee you an unrivaled quality of custom-written papers.

Get ZERO PLAGIARISM, HUMAN WRITTEN ESSAYS

Why Hire Collepals.com writers to do your paper?

Quality- We are experienced and have access to ample research materials.

We write plagiarism Free Content

Confidential- We never share or sell your personal information to third parties.

Support-Chat with us today! We are always waiting to answer all your questions.