Opening the Black Box: How Enterprise Systems Influence
Interprofessional Collaboration
Master thesis, MSc Business Administration, track Change Management
University of Groningen, Faculty of Economics and Business
Date: 20-1-2020
Femmina Petrusma
F.A.Petrusma@student.rug.nl
S2752069
Supervisor: N. Renting
Co-assessor: dr. O. P Roemeling
2 ABSTRACT
Electronic health records (EHR) have been widely adopted to aid information sharing and communication between healthcare professionals. However, it is still unknown how technological implementation has changed collaboration between physicians and nurses, while communication and coordination have great effect on delivering safe healthcare. The goal of this study is to examine this relation by exploring which collaborative IT affordances are actualized and how they influence interprofessional collaboration (IPC). To study this societal matter, this research combined theory of affordance and cultural-historical activity theory, to provide a context-based framework, instead of investigating a phenomenon in its isolation. This qualitative study was conducted at two departments within a Dutch academic hospital that is currently positioned in its post-implementation stage. Guided by grounded theory principles this study was able to iteratively collect and analyse data. Thorough analyses show that EHR poses benefits and constraints to interprofessional collaboration (IPC). Specifically, I demonstrate the existence of three (in) direct collaborative IT affordances, which support the coordination of care, however, simultaneously also decrease the responsibilities of the nurse in the healthcare process, and reinforce the traditional hierarchy between physicians and nurses.
3 TABLE OF CONTENTS ABSTRACT ... 2 TABLE OF CONTENTS ... 3 1. INTRODUCTION ... 5 2. LITERATURE REVIEW ... 8
2.1 Interprofessional collaboration (IPC) ... 8
2.2 Electronic Health Records (EHR) ... 9
2.3 Theory ... 11
2.3.1 Theory of affordances ... 11
2.3.2 Cultural-Historical Activity Theory (CHAT) ... 13
3. METHODOLOGY ... 14
3.1 Theoretical approach ... 15
3.2 Case site and participants ... 15
3.3 Data collection ... 16 3.3.1 Shadowing ... 16 3.3.2 Interviews ... 17 3.5 Ethical considerations ... 20 4. FINDINGS ... 21 4.1 A changed situation ... 21 4.2 IT Affordances ... 21 4.3 Joint registration ... 22 4.4 Shared overview ... 25 4.5. Communication ... 28 5. DISCUSSION ... 34
5.1 Principal findings & meaning ... 34
5.2 Theoretical contribution ... 36
5.3 Strengths & limitations ... 37
5.4 Implications for practice ... 37
5.5 Future research ... 38
6. CONCLUSION ... 39
REFERENCES ... 40
APPENDIX A: Interview guideline ... 44
4
5
1. INTRODUCTION
Information systems (IS), such as electronic health records (EHR), are fundamentally changing healthcare in order to increase the efficiency and effectiveness of its processes (Hagland, 2016; Kruse et al., 2018). Currently, most preventable errors originate from interaction- and communication activities between health professions (Leonard, et al., 2004; Pain, et al., 2017). For example, errors appear during communication due to language barriers or limited contact between professionals (Junaid et al., 2019). Interprofessional collaboration (IPC), in which interaction and communication occur, is seen as a crucial component that is needed to provide safe, effective and optimal health care (WHO, 2010, Van Leijen-Zeelenberg, 2015; Reeves et al., 2017).
Research has already shown that IPC is extremely necessary in healthcare environments since one profession can only achieve a proportion of all diverse patient needs that must be met within limited time (Matziou, et al., 2014; Collette, et al., 2017). Where the nurse observes the medical and social condition, the physician is focused on the treatment plan of the patient (Foronda et al., 2016). Information and joint coordination and communication of both health professions is needed to create a shared understanding about the patient and deliver safe and excellent quality patient care (Chao, 2016). Effective information sharing and communication is more complex due to the different perspectives and roles (Pain et al., 2016).
The implementation of new information-technology changes work practices (Korpela, 2002) and brings health professionals new possibilities, also known as affordances, to collaborate with each other. While the implementation stage is crucial to derive benefit from these possibilities, mainly in the post-implementation stage we can understand its influence (Ahmadi et al., 2018). However, research has mostly overlooked the post-implementation stage (Ahmadi et al., 2018; Grabski et al., 2011; Ha & Jun Ahn, 2014), while only in this stage we are able to create an overarching understanding of the influence of a technological change (Ahmadi et al., 2018).
6
(Collette et al., 2017) because the use of social-historical context, which influences the fit between technology and people, is overlooked most of the time (Bardach et al., 2017).
Research shows that both the implementation and post-implementation stages are a critical change process (Boonstra et al., 2014; Jasperson, et al., 2005) where change managers must be involved with managing both implications and its overall process. However, a deficit of knowledge regarding the influence of EHR on IPC hampers the ability to coordinate the work of the change agent (Weber, 2003). To obtain insight in this relationship and support change managers to understand the post-implementation process, I will develop a context-specific framework which will allow us to look at the features of IS and its social-historical contexts (Orlikowski & Lacono, 2001). By following this contextual approach, I will also contribute to the academic discussion to deliver more context-specific theories (Burton-Jones and Straub 2006; Hong et al. 2013). Specifically, I do so by combining theory of affordance and cultural-history activity theory (CHAT) as lenses that explain what happens within the post-implementation stage of EHR.
By combining theory of affordances and CHAT I aim to provide a framework to study actions on an activity level of analysis which helps to examine how technological change interacts its context (Sadeghi et al., 2014). Theory of affordances will allow us to identify the possibilities EHR offers and CHAT provides the ability to look at both the impact of technology and the impact of the complex relationship between the social and technical environment (Crawford & Hasan, 2006). Combining CHAT with theory of affordance will assist us to explain tensions and successes of EHR on IPC and ultimately provide practical directions for change. Hence, this study aims to answer the following research question:
How does EHR change interprofessional collaboration in the post-implementation stage?
7
8
2. LITERATURE REVIEW
This chapter starts with a further explanation of interprofessional collaboration (IPC), consequently l review the current state of knowledge and deficits thereof concerning the effect of electronic health records (EHR) on IPC. Finally, I elaborate on both the motivation and explanation of the theory of affordances and CHAT.
2.1 Interprofessional collaboration (IPC)
Research has already shown that IPC is extremely relevant in healthcare environments, since one profession can only achieve a proportion of all diverse patient needs that must be met within a limited amount of time (Matziou, et al., 2014; Collette, et al., 2017). During IPC both knowledge transfer and creation takes place and all involved parties interact with each other to accomplish mutual objectives (Huq et al., 2017;Karam et al., 2017). IPC is seen as a critical component because poor IPC has adverse effects on health care (Zwarenstein, et al., 2009). Poor IPC influences the ability to deliver patient care (Kvarnström, 2008), puts patients at risk (Pain et al., 2017) and is one of the prime causes of errors in health care (Lancaster et al., 2015). Vice versa, effective IPC is associated with mutual goals, reciprocal respect, partnership with interdependent roles and power sharing (Rose et al., 2011), which results in better management of complex problems, coordination, integration of services and providing good quality patient-oriented care (Samuelson et al., 2012). Multiple scholars have focused on IPC which has resulted in plenty of differing definitions (Morgan et al., 2015; Zwarenstein, et al., 2009). Reeves et al. (2011) performed a review among 107 papers to identify the following key concepts of IPC: (1) coordination and communication, (2) knowledge about each other’s roles and (3) working towards mutual goal. The use of these elements will allow us to identify IPC
9
the time a nurse diagnosis patients with a holistic perspective with a high level of emotional intelligence. On the other hand, the role of a physician is to effectively treat patients and thus values an objective approach which is structured and based upon medical data (Foronda et al., 2016; Eggenberger, 2012). However, these differences in practices and values may also provide diversity and depth in perspectives that complement each other (Foronda et al., 2016).
Another important barrier that can occur to achieve IPC is the power dynamics between physicians and nurses (Rose, 2011). The relationship between physicians and nurses has traditionally been hierarchical. Although hierarchy is not surprising within these professions because of the educational background and different range of responsibilities (Reeves et al. 2010). Hierarchy can have a positive influence on IPC since it gives comfort and clarity regarding responsibilities and communication (DiPalma, 2004), where communication and knowledge about each other’s roles are aspects of IPC (Reeves, et al., 2011). However, hierarchy can also negatively influence IPC. Medical dominance of physicians and compliance of nurses can result in errors when subordinates hesitate to communicate with higher positions due to fear or embarrassment (Edmondson, 1999). This lack of communication increases the number of preventable errors (Calhoun et al., 2014).
2.2 Electronic Health Records (EHR)
The introduction of the enterprise system EHR has radically changed healthcare. Before the implementation of EHR, the physicians held a paper medical record which contained medical and treatment history and the nurses had their own nursing records for every patient. The nurses did not have any access to the information of the physicians and the physicians needed to look up the nurses’ information in their binders. Information was exchanged via phone and during set meetings,
10
a general positive effect of EHR on medical outcomes and on healthcare quality (Campanella et al., 2015; Kruse et al., 2018).
The change towards an electronic medical record not only influences medical outcomes but also changes collaborative activities (Bardram & Houben 2018). On the one hand, O’Malley et al. (2015) reveal via qualitative interviews that EHR fosters communication and facilitates task delegation by facilitating improved access to patient information for all team members. Pain (2016) shows that due to allied health notes health professionals have more insight into each other's work and perspectives and this decreases communication errors. Results of Weir et al. (2011) are in line with the above-mentioned outcomes and shows that EHR is an essential information source and channel to create a shared understanding of the patient.
On the other hand, researchers also show that EHR perhaps improves information access, it does not improve communication and coordination. For example, Chao (2016) explains that although EHR facilitates information exchange, it does not improve the ability to create a clinical picture of the patient, and there is a higher possibility of information gaps when forming a clinical picture. Also, Robinson et al., (2010) claim that EHR does not support in communication because it results in fragmented and incomplete information and it is therefore necessary to follow up with verbal contact. Health professionals also prefer synchronous communication because it offers immediately clear reaction. Although direct communication is preferred, a decrease in face-to-face interaction is found because communicating via EHR does not require physical presence which saves time (Taylor et al., 2014). A decline in interaction decreases clarification and verification and increases miscommunication (Robinson et al., 2010). Although a generally positive effect is found regarding EHR and medical outcomes (Van Leijen-Zeelenberg, 2015; Reeves et al., 2017), no general positive effect is found regarding the influence of EHR on IPC. These mixed results imply that there is an incomplete understanding of how EHR influences IPC. It is important to understand this relation since poor IPC is one of the most common errors in patient care (Lancaster et al., 2015).
11 2.3 Theory
This study will combine theory of affordances and cultural-historical activity theory (CHAT) to build a context-specific framework in order to study the influence of an electronic system on interprofessional collaboration (IPC). Theory of affordances provides to us a suitable theoretical lens to explain technological artefacts and its relation with its social actors (Leonardi, 2011; Pozzi et al., 2014; Stendal et al., 2016;). Adopting the lens of affordances forces us to analyse the capabilities of technology (EHR) relating to the action taken by individual actors (health professions). However, to understand its influence, the affordances need to be placed within the social-historical context. CHAT both incorporates and specifies situational elements to study how these affordances influence elements within collaboration during activities. It reveals both the overarching activity and underlying activities that are involved in working towards the mutual goal (Greig et al., 2012). By combining both theories into one framework, it can specify how technological affordances influence the current social-historical context. An explanation of both theories is given below.
2.3.1 Theory of affordances
In the last two decades there has been renewed attention towards the use of affordances in IS research, mainly due to its ability to explain the consequences of the implementation of IT artefacts (Markus and Silver, 2008; Pozzi et al., 2014). Theory of affordances originates from the ecological psychology wherein the founder Gibson (1977) states that a goal-directed actor perceives an object in terms of action possibilities (‘affordances’) that the object provides. After its value within the ecological approach, Norman (1988) introduced the concept in the human-computer interaction (HCI) domain (Kaptelinin & Nardi, 2012). Norman (1988) deviates from the original concept of Gibson by describing that the ability to perceive the affordance determines the existence of the affordance. Norman includes the mental and perceptual capabilities of the actor to perceive the affordance. According to Norman (1988), an affordance provides the actor with clues on how to use the artefact.
12
(Pozzi et al., 2014) and goal-directed actions are taken by individuals to realize a perceived affordance (Strong et al., 2014). For example, EHR has a mail function to communicate with the nurse, but only if the physicians actualize the mail function of EHR to exchange medical information with the nurse, it derives benefits from the affordances (figure 2). The effect of affordance can be beneficial but also constraining, it depends on the characteristics of the actor and object (Strong & Volkoff, 2013; Leonardi, 2011).
Figure 2: Visual representation of Theory of Affordances adapted from Pozzi et al., 2014
Within the IS and HCI domain, theory of affordances is mainly focused on the individual potential the affordances provide (Bardram & Houden, 2018). This is not remarkable since the founder Gibson (1977) also focused on individual perception. Moreover, the HCI domain is mostly focused on individual interaction with the computer. When trying to explain how enterprise systems affect collaboration, we need to focus on collaborative potentials. Bardram & Houden (2018) showed that the focus on collaborative action is possible, by studying how technological artefacts afford collaboration, whereby the focus is on the joint action which the affordance creates. Following the definition of Bardram & Houden (2018) a collaborative affordance is: ‘A relation between a (physical and/or digital) artifact and a set of human actors, that affords the opportunity for these actors to perform a collaborative action within a specific social context.’ (Bardram & Houden, 2018, p. 8). Bardram & Houden (2018) found four examples of collaborative affordances within EHR and paper files: portability, collected access, shared overview and mutual awareness. Bardram & Houden (2018) focused only on direct collaborative activities, which implies, physical presence of both professions. I extend the concept of collaborative affordances to both direct and indirect collaborative activities as the introduction of EHR has significantly increased the possibility of indirect collaborative activities. For example, digital information exchange.
13
and that their minds and actions are mediated by social-historical tools. To specify this mediating context, we will utilize cultural-historical activity theory (CHAT) as a data analytical tool (Igira & Gregory, 2009).
2.3.2 Cultural-Historical Activity Theory (CHAT)
CHAT has previously been used in hospital settings, as it is a framework which can unravel unintended consequences of healthcare services (Greig, et al. 2012; Teodorczuk et al., 2015). CHAT is based on the work of a Russian cultural-historical psychologist Vygotsky and according to Vygotsky (1978) we cannot examine an activity without its context. An activity consists of its subject (individual or group), goal, artefacts and sociocultural rules. All these aspects have influence on how the activity will be initiated and executed. Studying the influence of EHR on interprofessional collaborative activities without paying attention to culture, history, roles and work of nurses and physicians, is tampering with the essence of the activity (Kaptelinin et al., 1999).
Engeström (1987) expanded the initial version of Vygotsky (1987) by understanding how people are embedded in the social-cultural interacting context. In Engeströms view, the unit of analysis is the human activity system. This system is object-oriented, tool mediated and culturally mediated (figure 3). Within an activity of delivering health care, the subject (physician or nurse) is mediated by the tool (EHR) to accomplish its object (patient care). This activity is also mediated by its social-historical context, consisting out of
rules, community and division of labour. Division of labour refers to the interactions among members but also the division of power, tasks, roles resources, hierarchy and status. Community refers to the participants who share the same object. Every community has rules. Whereby rules consist out of both formal rules such as protocols and informal rules such as norms and values. These rules can afford or constrain the activity (Igira & Gregory, 2009).
14
Tertiary contradictions emerge when a new activity is developed, and the old activity is hindering the new activity. For example, when EHR changes the work procedures of nurses but nurses want to keep pursuing their established work procedures. The fourth and last contradiction is when there are tensions between different activities which share components. For example, the general practitioner collaborates with the physician, but the physician uses a different system (Engeström, 1987; Mursu et al., 2007). These four contradictions ultimately lead to development and change.
15
3. METHODOLOGY
This chapter elaborates on the apprehended theoretical approach, case site and participants, data collection and data analysis to ensure the controllability of this study (Van Aken et al., 2012).
3.1 Theoretical approach
Aim of this study is to explain the influence of an information system on interprofessional collaboration within the post implementation stage. This study assumes that data is constructed by the interaction between participants and their environment, where patterns evolve through the interpretation of the researcher (Charmaz, 2000). Following this viewpoint, qualitative research methods are needed to discover these patterns. Moreover, little is known about the influence of EHR in the post implementation stage. Therefore, this study tries to obtain insights on how EHR influences IPC where qualitative methods are appropriate to enhance the understanding of this topic (Wiesche 2017; Edmondson & McManus, 2007). I used informed grounded theory as a suitable way to systematically and flexibly collect and analyse qualitative data (Charmaz, 2006). Grounded theory provides a practical way to gather and analyse data. Methods of grounded theory are appropriate for descriptive studies since it focuses on analysing social processes within a context and allows to generate theories. Moreover, grounded theory supports the iterative process of discovering new insights and linkages (Charmaz, 2006). Although traditional grounded theory does not use preconceived theories upon data (Glaser & Strauss, 1967), informed grounded theory can be used within different theoretical frameworks. Using preconceived theories contribute to analysing the multi-layered nature of individual, cultural and historical change in contexts. For example, cultural-historical activity theory adds an up-close analytic procedure (Seaman, 2008) and thereby highlights and explores perceptions regarding phenomena (Khan, 2014, Seaman, 2008) and theory of affordance contributes by analysing technological artefacts and its relation with social actors (Stendal et al., 2016; Leonardi, 2011). Grounded theory places high responsibility on the researcher, which makes this study more vulnerable for subjectivity (Hall & Callery, 2001). The change management background of the researcher contributes by increased attention to the broader organisational context within this study. However, the researcher had limited medical educational background, therefore initial shadow sessions took place.
3.2 Case site and participants
16
reliability, data collection has been done within one surgical (trauma surgery) and one medical (internal) department (Van Aken et al., 2012), which also served as the unit of analysis. Two distinctive departments have been chosen due to the main difference in providing care. As the medical internal department analyses the human body and tries to discover the problem, is the problem at the surgical department familiar. Because this study examines how EHR influences interprofessional collaboration, both nurses and specialist at the medical and surgical ward have been interviewed and observed. Participants have been selected via purposive sampling to make sure different opinions and professions are represented in the final sample (Robinson, 2014). During initial interviews, the dimensions, boundaries and scope of this research were established. After these interviews the official sampling criteria were established. Respondents were selected based on their diverse opinions on the information system to gain a maximum variation in the data. Moreover, criteria apprehended to select respondents were occupation and department (Bryant & Charmaz, 2007). To better understand the patterns and interactions, after every interview data has been transcribed and analysed. The process of scheduling interviews, conducting interviews and data analysis has been an iterative process to pursue a better understanding of the activity at hand (Gill et al., 2014).
3.3 Data collection
In order to gather primary data, a qualitative approach has been apprehended. Specifically, multiple data collection methods incorporating shadowing, field interviews and semi-structured interviews have been used. During shadowing sessions, functionalities of EHR could be studied. Moreover, specific background information has been collected by looking into previous research done by the examined hospital. This research was executed regarding the implementation of EHR. When valuing multiple realities, combining data collection methods (triangulation) provides a stronger validation of concepts and propositions (Jonsen, 2009).
3.3.1 Shadowing
17
change since the implementation of Epic?’’. Moreover, it allows individuals to be researched as an embedded part of the social and organizational environment and it is able to observe connections across work teams (McDonald, 2005). Following the advice from Gill et al. (2014) and McDonald (2005), all relevant observations were written down in a small hard-back notebook. After a session of data collection, written notes were converted to Microsoft Word and analysed. In total, six shadows have been conducted, ranging from 1.5 to 8.5 hours, with a total of 23 hours. The first shadow session took place on the 30th of September and the last session took place on the 19th of December. The shadowing sessions started at the medical department, where the research mainly focused on the roles of the participants and the organisational context, which assisted to create an initial image of the context and jobs of health professions. After these initial shadowing sessions, the research focused more on both direct and indirect collaborative moments including and excluding EHR. At the surgical department the researcher already had background information, due to previous shadowing sessions and interviews, resulting in fewer shadow sessions necessary to reach saturation. Gained knowledge from observations acted as input in interviews. Moreover, shadowing sessions assisted the researcher in acquiring respondents for formal interviews. The table below shows an overview of the executed shadowing sessions (table 1). Majority of the shadowed participants have also been interviewed. To ensure the privacy of the (observed) respondents, actual names have been replaced by fictitious names throughout the remaining of this paper.
Name Department Occupation Setting Observation
duration in hours
Thomas Medical Medical specialist Consult 4
James Medical Medical specialist Ward round 3
Damian Medical Medical Resident Multidisciplinary consultation 1.5
Sarah Medical Nurse Ward round, Multidisciplinary
consultation, medicine rounds 8.5
Amy Surgical Surgical Resident Ward round 3
Zoë Surgical Nurse Ward round 3
Table 1: Overview observations
3.3.2 Interviews
18
Word and analysed together with the shadowing notes. 4 informal interviews took place, which are included in the table below (table 2).
Name Department Occupation
Sarah Medical Nurse
Emma Medical Head of nurse
Boris Surgical Surgical resident
Anna Surgical Head of nurse
Table 2: Overview of informal interviews
In total 14formal interviews have been conducted until theoretical saturation was reached. Semi-structured interviews were conducted to give insights how physicians and nurses think about the influence of EHR on their collaboration. The interview protocol (Appendix A) provided a structure to the semi-structured interviews, while simultaneously allowed room for probing questions (Charmaz, 2006). Location of the formal interviews was arranged by the respondents, which made the respondents feel more at ease (Jacob and Furgerson, 2012). The goal of initial interviews was to understand their roles and how these have changed, perspective towards EHR and how the professions collaborated. After initial interviews, the interview protocol was collaboratively assessed and restructured by the researcher and expert in the medical field. The interview protocol starts with introductory questions regarding their occupation and how EHR influences their daily work routines. The core questions were based on the influence of information systems regarding information exchange, collaboration activities, communication and perspectives of each other’s tasks and responsibilities. Iteratively, adaptions were made when emerging insights developed. All interviews started with a similar introduction and closing remarks (appendix A) to inform the respondent about the goal of the interview, the interest of the researcher in their perspective, the confidentially of the data and to sign the informed consent. The interviews vary from 19 till 60 minutes with an average of 33 minutes. The first interview has been conducted on the 16th of October and the last interview was on the 16th of December. With permission of the respondents all interviews were audio taped. An overview of the participants is shown in table 3.
Name Department Occupation Gender Interview duration in minutes
Years of experience as practitioner
Oliver Medical Medical
specialist
M 60 33
Jack Medical Medical
specialist
19
Thomas Medical Medical
specialist
M 36 13
James Medical Medical
specialist
M 54 21
Damian Medical Medical
resident
M 27 2
Jennifer Medical Nurse F 32 19
Sarah Medical Nurse F 19 2
Emily Medical Nurse F 25 6
Liam Surgical Surgeon M 24 6
Robert Surgical Surgeon M 22 27
Lisa Surgical Surgeon F 30 7
Michael Surgical Nurse
specialist
M 46 34
Lily Surgical Nurse F 24 3
Lauren Surgical Nurse F 21 1
Table 3: Overview of formal interviews
3.4 Data analysis
20
identify groups among the codes and cluster these in axial codes. Theory of affordance helped with identifying IT affordances. For example, it became visible EHR assisted in a shared overview, due to the possibility of nurses to read the notes of the physicians. CHAT helped to structure the social-historical context. For example, vertical division of labour combined shift in roles and nurse dependent on physician. Third, selective coding has been performed, which applies overall codes to relevant categories and allows to move to a more abstract level of conceptualisation (Strong & Volkhoff, 2010). The categories have been based on technological and social-historical contextual features. Where theory of affordances has mainly been applied to the technological features and CHAT to social-historical contextual features. The overall coding process is performed in a qualitative data analysis software program Atlas.ti 8.4 and is displayed within the codebook, including explanation of the code and quotes (appendix C)
3.5 Ethical considerations
21 4. FINDINGS
This section describes and analyses how EHR can positively but also negatively influence IPC and the roles of health professionals. This chapter will start with a short illustration describing the changed situation. By analysing the data in light of theory of affordances and CHAT, this study found three actualized (in) direct collaborative affordances. These affordances and the effect of these affordances on IPC will be presented, including extractions from observations and interviews. Differences between physician and nurses, the two focus departments and data collection methods are discussed when relevant.
4.1 A changed situation
As mentioned in the previous chapter, the information system was implemented two years ago. Before the implementation, physicians worked in a registration system which was not integrated with other medical employees. ‘It is quite interesting that before we had Epic, the nurses did everything on paper. I think our nurses were the very last who worked with paper files and task list, which had to be approved by physicians. Consequently, we had to walk to these paper files to check and approve medication’’ [Jack, medical specialist]. Moreover, before the implementation of EHR, nurses could not access the information of physicians ‘Physicians had their own electronic registration, we were sometimes allowed to look at it for a second, but no more’’ [Jennifer, medical nurse]. With the implementation of EHR, physicians switched to a different information system and nurses switched from paper files to electronic health records. As such, dossiers of physicians and nurses were integrated into one system. EHR system plays a huge role in the daily work routine of physicians and nurses. ‘It is our patient information system. So, all patient information is placed and stored in the system. Lab results, tests, yes everything is listed and we also need to gather our information from it’’ [Liam, surgeon].
4.2 IT Affordances
22 4.3 Joint registration
The first IT affordance, joint registration, implies the ability to register patient information into a shared system. This affordance is constructed between the interaction of the system, the actors (physicians and nurses) and the goal to register relevant information. Abstracted from shadow sessions and interviews, it can be implied that physicians mainly report their findings and treatment plan under ‘notes’. Nurses mainly register medical information under ‘functional status’, where medical information contains checks regarding blood pressure, temperature, and heart rate. Within the surgical department nurses also report under ‘care plan’, which includes mobility and pain. Extraordinary symptoms can be written in notes. ‘If patients show remarkable symptoms, we will register this extra information by making a note in the system.’’ [Emily, medical nurse].
When actualizing the affordance joint registration, it appears that it helps physicians and nurses to understand each other’s jobs and responsibilities. ‘When both professions register information into one information system, it decreases the confusion we experience as nurses. It gives so much more insight into the decisions made and the treatment plans followed by physicians. This way, we can understand each other better, have information about the other person’s tasks and anticipate on this information´´ [Jennifer, medical nurse]. However, respondents also indicate that EHR limits in registration. ‘We have the functional status, where you can write something down, but the space is limited. You can probably write no more than one sentence within this functional status. Sometimes you want to write just a little bit more about a subject, for example when someone is nauseous and which actions we took’’ [Lily, surgical nurse].
23
joint registration is to gather information and have information about each other’s tasks. This activity is visualized in figure 5.
When analysing division of labour and the affordance, it appears that the effect of limited registration mainly applies to the nurse. According to the division of labour, the nurses checks and registers both medical and psychological-social information, where medical information consists out of blood pressure, heart rate, urinary consistency e.g. and psychological-social information consists out of implications with family, state of mind e.g. However, in the functional status there is no room for psychological-social information. During the shadowing session Anna mentioned: ´The progress notes are medical-related most of the time, I would prefer if some information would be human-related because that kind of information is most often missing. There is no real space in the current system to register this kind of information’’ [Anna, surgical nurse]. Moreover, EHR restrict the nurse in writing her own patient summary. ‘In the early days the nurse wrote her own status. She wrote information and the next day during the transition of the nurses this information was red. In this way, the ‘new’ nurse could create an image of the patient. It was a yellow paper with a kind of summary of the information of the patient, why the patient was in the hospital, what are the problems. This kind of summary does not exist anymore’’ [James, medical specialist].
The absence of this summary restrains the nurses to obtain a good image of the patient. ‘There is not really a place where you can directly read about the patient. You need to check many notes of the physicians before you understand what is going on with the patient’’ [Lily, surgical nurse].
24
information. The surgical resident listens to the nurse and registers in their progress report under anamnesis the mobility and night rest of the patient. However, information regarding the psychological situation and the necessity to call in a psychologist is not registered [fragment observation Amy]. Anna complies with the observation: ‘Often our information is also regarding the social or psychological situation of the patient, however the physician does not dictate this information’’. [Anna, surgical nurse]. However, the limited role of psychological-social in the treatment plan is not shared by all interviewees. Some physicians from the medical department indicate that they find it a pity that psychological-social information has vanished. ‘The nurse needs to decide whether they find it worth it to rapport. In the old situation, they had to sit down and register the information of the day. Doing good observations, about what is the patient is doing today, did he come out of the bed, yes that is part of the nurse’s diagnosis. These are the things you want to have systematically reported and since Epic is introduced, this is not the case anymore’’ [Oliver, medical specialist].
25 4.4 Shared overview
The second IT affordance, shared overview, means the ability to easily share all available information via EHR. By making the nurse’s information easily accessible for physicians and making physician’s information accessible for nurses, the health profession can share their information and work together towards a mutual goal. Following fragment shows how shared overview is actualized: ‘James opens EHR and looks at the registered blood pressure of his patient. Based on this information, he discusses with the physician about the medication and they decide to decrease the amount of medication’’ [fragment observation James]. This illustrates that the affordance is actualized when they read each other’s information, analyse this information and maybe use this information.
All interviewees agree that a shared overview is an enrichment to their jobs. Although they could already access the information, a shared system saves time and provides a more comprehensive overview of the patient. ‘Epic consists of all the information of a patient, and not only the doctor’s information but also information from other people, for example nurses who are also involved with the patient. I do not have to walk to the department and check the files of the nurses. I can just easily look it up online. It helps us as physicians to work together on the same information towards a mutual goal. And do I like it? Yes, I do.’’ [Thomas, medical specialist]. In previous situations nurses did not had the opportunity to access the information of physicians. The ability of the nurse to access the information of the doctor gives nurses the possibility to create a better medical picture of patients and better understand the analysis of the physician and provide useful input in the treatment plan. ‘Before we almost did not have any medical background. To have this extra information is such an enrichment to our jobs. We have the possibility to look into the notes of the physicians and other paramedics. We can check pictures, blood results. We have way better access to all the information and that allows us to better interpret situations, provide valuable input and make a well thought-about decision to call or not to call the doctor’’ [Jennifer, medical nurse]. It seems that the affordance positively influences the ability to share information in order to work towards a mutual goal.
26
medical test takes a couple of days, sometimes weeks, before we have the results. And every time, you are really searching for a result in a long list of notes, and you have no idea which are known results and unknown results.’’ [Damian, medical specialist].
When actualizing this affordance in its social-historical context, it appears that the system influences information sharing activities. A striking difference is detected between the surgical and medical department. The performed shadowing sessions showed that in the medical department physicians and nurses directly collaborate by exchanging information during ward rounds, which are held every day. During this collaborative activity the physician visits the appointed patients on the department. Before the patient visit, the physician talks with the nurse to create an up-to-date image of the patient since the nurse sees the patient more often and can provide a respectable image of the patient. This activity takes place in the office of the physician. The nurse is called into the office and brings the computer on wheels which displays EHR. EHR (tool) supports this activity by providing a shared overview. The physician leads the meeting by reading out all relevant information from EHR. The nurse can read along from her computer on wheels (Division of Labour). The activity is visualised in figure 7.
27
during the morning because they cannot give any input and their role is very small. So yes, that is something that has changed. Before Epic, we could not easily access this kind of information, so we were more dependent on the input of the nurse and there was more interaction. I also find it difficult, but I am very happy that I have easy access to all the information’’ [James, medical specialist]. Also, some nurses mention that they do not see the added value anymore of attending this collaborative activity. ‘Honestly, I find it quite useless. You need to stop your work activities and most of the time the physician almost asks nothing. I have something better to do than being there’’. [Sarah, medical nurse]
This leads to a secondary contradiction between the affordance of shared overview and the role of the nurse
who used to provide input, during this collaborative activity (division of labour) (figure 8). Our data indicate that the shared overview makes the presence of the nurse
unnecessary.
28
have to search for this information in Epic, I sometimes can’t find the correct information.’’[ Boris, surgical resident].
Moreover, the outline of the ward round at surgical differs compared to the medical department. The performed shadowing sessions show that during this activity the physician walks on the hallway with the computer on wheels which displays EHR. The nurse also has their own computer on wheels. The nurse who is responsible for the patient joins the physician and leads the meeting by starting with summarising the background information of the patient and its condition. The physician listens and asks explanatory questions regarding how the patient slept and its mobilisation (division of labour). The physician dictates the given relevant information under amnesia within the EHR function ‘progress notes’. EHR is used in this activity to clarify, to register the information and as background information. The nurse leads the meeting which makes the nurse feel valued during the meeting. The role of the tool EHR is more facilitating than the situation within the other department and it does not overrule the input of the nurse. The different division of labour and rules results in a harmonised activity where all parties feel valued.
4.5. Communication
29
fragment: Sarah opens her work list and noticed that another order appeared on the screen. It indicates that she needs to supply medication to one of her patients [fragment observation Sarah]. Physicians need to order actions for nurses, for example take a blood test. Nurses can find this order in their worklist and then the nurse can carry out this order.
Almost all physicians and nurses indicate that communicating via EHR increases the coordination of care. When communicating via the system, you do not have to find that one person, but you can just ask/order something via the system. ‘Because they are mostly behind their computer, they frequently check their worklist. They can see, hey this is a new activity. Nurses spot this request of order sooner and understand what they have to do, I think that is a good development’’ [Damian, medical resident]. Moreover, physicians can check online whether an action is performed by a nurse. ‘The physician asks the nurse why an order is not executed. The nurse had no idea the order was not performed. The physician shows his screen to the nurse where it indicates that an order was placed, and it has not been executed’’ [observation fragment Zoë]. With the possibility to order digitally and the ability to check if it happened, it increases the coordination of care which has a positive effect on IPC.
However, respondents mention that the possibility to communicate with the other professions has a negative effect on face-to-face communication and direct feedback on actions taken. Almost all interviewees indicate that communication via EHR replaces face-to-face communication. ‘Where before the information system, physicians communicated with the nurses. Now, you can just order the medication. You indicate that it needs to be given right away, and it will appear on the worklist of the nurse. This makes the direct communicative part minimal’’ [Jack, medical specialist]. Lisa even describes the current collaboration cold due to the increase in interaction. ‘EHR made collaboration between us, physicians and nurses, cold. Previously, we discussed the treatment plan and course of action in the hallway. Now, you see the patient, order relevant medicine from your computer, after which the nurse executes the order’’ [Lisa, surgeon]. Besides face-to-face communication, feedback on action taken is also decreased. ‘Sometimes the physician prescribes medicine and we look into EHR and find this prescription. However, he did not give any feedback that he performed this action. Sometimes physicians immediately assume that the nurses read it anyway. I do not like that. Just tell me when and why something needs to happen. This makes me also more able to explain to the patient what I am doing’’ [Emily, medical nurse].
30
be drawn that the aim to communicate via worklist is to efficiently coordinate the care process (object). The physician needs to place orders which include the particular action and the time of this action. Subsequently the nurse can carry out the order. However, communication via this tool leads to confusion at the side of nurses. ‘Sometimes the physician does not tell anything and suddenly you see that there is an order of medication in the system and then I think, okay? What do I have to do with that?’’ [Lauren, surgical nurse]
Moreover, nurses have their own standard routines when to give medication (division of labour). However, the physician needs to set the order in the system including the time of the medication, the nurse is not allowed to do that (rules). However, physicians have limited knowledge concerning the medication rounds of nurses. Consequently, the nurse needs to tell the physician which time they want to have the medication in the system. Often the physician does something wrong. Extracted from all the interviews with nurses this leads to irritation from nurse to physician. ‘You always have to call again and say; hey, this order is not correct, you have to order it likes this. Sometimes, another mistake is made. It keeps you busy all the time and that does not work’’ [Sarah, medical nurse]. Vice versa, almost all physicians indicate that they find it annoying that nurses complain about the wrong orders ‘And then another comment about the times that do not match, and this is not correct concerning the amount. It is only grizzle and as a physician you do not want to spend your time on this task. I do understand that it should be correct because you do not want to make any mistake, but it does overrule in the collaboration’’ [James, medical specialist].
31
32
33
34 5. DISCUSSION
The aim of this study was to examine the influence of the technological change of health records on interprofessional collaboration in the post-implementation stage. The corresponding research question was: ‘’How does EHR change interprofessional collaboration in the post-implementation stage?’’ This chapter examines the findings and impact of this study by looking at the principal findings and meanings. After which the theoretical contribution, strengths and limitations, and implications for practice and future research will be discussed.
5.1 Principal findings & meaning
The findings have shown that EHR influences IPC in a very different way previous literature made me expected. Derived from the literature I expected that EHR mainly has a positive effect on information sharing and slightly more negative effect on communication and coordination. However, the findings including the social-historical context suggest that EHR mainly influences IPC by increasing the coordination and decreasing the role of the nurse. Ironically, the implementation of EHR, which increased the information access of the nurse, made the nursing role more reliant on the physician. I observed that the social-historical context mediated the (in) direct collaborative IT affordances which affected the nursing role by troubling the ability to gain a clinical picture of the patient, decreasing their authority, possibly decreasing the input during information sharing activities and limiting their role regarding register of information. These limitations for the nurse to perform his or her job are in contrast with the current expanding role of nurses and the shift towards becoming more autonomous (Matziou et al., 2014). This study showed that collaborating via EHR, resulted in a shift in autonomy regarding the treatment plan. The physician (dominance) could digitally order it after which the nurse executes the order (compliance) with minimal communication. Although it appeared communication is becoming more important with EHR. Due to minimal communication and difficulty to register extra information, barriers exist to provide input. This change reinforces traditional hierarchy, where nurses were framed as ‘handmaidens’ to physicians (Matziou et al., 2014). This revises IPC into a more cold, hierarchical, arm-length relationship between physician and nurses. Although hierarchy can increase the clarity regarding responsibilities and tasks (DiPalma, 2004), it puts emphasis on the autonomy and expertise of the physician, while undermining the role of the nurse.
35
Moreover, it makes it more difficult to create a full image of the patient. This finding corresponds with the study of Varpio et al. (2015). They found that EHR does not leave room for what they call the patient story. The limited role of psychological-social information can be explained by the fact that healthcare needs to become more efficient. When the healthcare focus is mainly on medical information, it can potentially be faster and cheaper. However, it undermines the attention for the patient as a human being which is an important component in delivering personalised care (Goldfarb et al., 2017) (figure 13).
Figure 13 Changed situation
36
Taken together, this study investigated what possibilities EHR offers and how they change IPC by opening the black box through specifying the influence of the social-historical context. By doing this, this research revealed that EHR changed the roles of the health professionals, in which it does not support the role of the nurse and it strengthened hierarchical collaboration.
5.2 Theoretical contribution
In line with the research of Bardram & Houden (2018) I argue that digital properties can offer not only individual activities, but also collaborative activities. Additionally, I add that digital properties can also afford indirect collaborative activities, as joint registration, shared overview, and communication does not require physical presence of both actors. Collaborative activities can be actualized simultaneously or separated from the other actors. All three (in) direct collaborative IT affordances support the idea of affordances as a relational concept, meaning they are neither property of the actors (physician and nurse) nor property of EHR (Strong & Volkoff, 2013; Leonardi, 2011). The found affordances are actualized when an actor collaborates via EHR during an activity. For example, during consult when a physician communicates with the nurse via EHR because the physician wants to ask something regarding the patient. When the physician asks the nurse something regarding the patient in person, the affordance is not utilised and therefore not actualized.
In combination with theory of affordances, this research applied CHAT as a data analytical tool to study the influence of the affordances on IPC. As mentioned, CHAT can create a deeper understanding of the relation between people and technology within its social-historical context (Igira & Gregory, 2009; Kaptelin & Nardi, 2018). For example, by using CHAT I found that social-historical context determines how the system is used and how this influences the collaboration. For example, at the surgical department the shared overview did not increase the input of the nurse because the nurse leads the ward round. This indicates that the influence of an enterprise system is mediated by its social-historical context and should not be studied separately (Kaptelinin & Nardi, 2012).
37 5.3 Strengths & limitations
The strengths of this study can be presented in twofold. First, the position of the researcher within this process. The researcher has a change management background which gave attention towards the organisational aspects. Moreover, the researcher positions itself next to the participants during data collection methods instead of an objective analyst. Moreover, during the research process the researcher had bi-weekly meetings with a medical research professional working within the researched hospital who provided the researcher with additional medical background information. This allowed the researcher to truly understand the motives and interaction of the participants and discover underlying motives and behaviours. Second, the iterative process of data collection and analysis consisting out of observations, interviews, informal interviews and secondary data provided a stronger validation of concepts and propositions. This iterative process allowed to verify input from shadowing sessions during interviews and vice versa which helped to recognise patterns and evaluate and test these with new data during our data collection.
Also, several limitations within this study need to be considered. First, there are many different system designs in different hospitals. The collection and analysis of data from only one hospital may compromise the transferability of this study. It is very likely other hospitals have a different social-historical context, design and usage of EHR. To overcome this limitation, this study collected and analysed data from two distinctive departments. Within these departments, health professionals already had a different arrangement of care and design and use of EHR. Second, although this study analysed the influence of EHR, participants sometimes compared the communication or information sharing with the previous situation from two years ago. Since this has already been some time ago, the accuracy of these answers can be called into question. I found that psychological-social information has a decreasing role in the EHR system and in the treatment of the patient. However, the role of this information has not been analysed before the information system. It is possible that the limited function of psychological-social information within the treatment plan only became obvious through EHR. I have tried to narrow these limitations by comparing my qualitative data with quantitative data collected by the university hospital itself with respect to the situation before and during the implementation of EHR. Nonetheless, these limitations imply that the research may not be fully comprehensive.
5.4 Implications for practice
38
influence on the influences of the affordances. Therefore, practitioners are advised to discuss with both physicians and nurses which affordances they want to actualize and how they want to structure the healthcare process. For example, more direct face-to-face communication moments to increase the interaction between the two health professionals and thereby prevent the decreasing role of the nurse. Second, our study detected some contradictions with the current activity systems. For example, decreased authority of the nurse in the system comparing to its former responsibilities and tasks. This contradiction leads to irritation from both sides. Although some nurses are maybe not authorised, I suggest providing additional education to motivated nurses, who will afterwards be authorised to adjust these orders. This will increase the role of some nurses on the department and decrease the administrative burden of physicians. Another contradiction has come to the surface regarding the input of the nurse. As our study showed contextual factors have a mediating influence. Changing the division of labour, for example by increasing the role of the nurse during ward rounds, can enlarge the role of the and make the nurse feel valued. I advise practitioners to invite change managers to handle above implications during the post-implementation stage, since change managers are trained in coping with organizational challenges, and understanding the perspectives of organisational members, and bringing them together to ultimately increase the efficiency and effectiveness of the organisation (Cawsey, 2015). It is advised to change managers to use the CHAT framework to look for new contradictions. CHAT it is an appropriate framework to detect contradictions in an early stage, which are ultimately sources of change.
Lastly, I would like to point out that an electronic healthcare system is more than a digitalisation of records, and that the way of designing an enterprise system reflects the vision of the organisation. This study showed that EHR is mainly focused on medical information, while a patient also has psychological-social information. This general philosophy of the organization also has implications for both health professions how to perform their jobs and eventually the quality of healthcare. It is important to decide as organisation what you stand for and how this is reflected in your current enterprise system. Additionally, I would like to mention to system designers, that an information system is not only an information system but influences the role of different professions, and its collaboration. Therefore, close attention should be given towards these professions when designing a system.
5.5 Future research
39
practitioners and eventually the care delivery process. This study revealed that EHR changed the role of nurses and collaboration, and I would like to inspire other academics to create awareness towards the role of the nurse and other health professions. I argue that the changing role of health professions will eventually lead to a less patient-directive distinctive health care. Second, this study only focused on one type of EHR, however there are many designs of EHR and the same design can be used differently within another department. To strengthen the transferability of this study other academics can conduct future research within other social-historical contexts and with different methods. Lastly, this study found that with the use of CHAT and theory of affordances, academics can analyse the influence of information systems on collaboration. However, this study only focused on one case site and specifically on EHR. Therefore, I would like to invite other researchers to use this framework with other case sites and information systems to strengthen the use of this framework.
6. CONCLUSION
40 REFERENCES
Ahmadi, H., Sadoughi, F., Azadi, T., & Shahmoradi, L. (2018, October). Hospital Information System (HIS) Performance in the Post-implementation Stage. In 2018 59th International Scientific Conference on Information Technology and Management Science of Riga Technical University (ITMS) (pp. 1-5). IEEE.
Bardach, S. H., Real, K., & Bardach, D. R. (2017). Perspectives of healthcare practitioners: An exploration of interprofessional communication using electronic medical records. Journal of interprofessional care, 31(3), 300-306. Bardram, J. E., & Houben, S. (2018). Collaborative affordances of medical records. Computer Supported Cooperative
Work (CSCW), 27(1), 1-36.
Bærentsen, K. B., & Trettvik, J. (2002, October). An activity theory approach to affordance. In Proceedings of the second Nordic conference on Human-computer interaction (pp. 51-60). ACM.
Benbasat, I., & Zmud, R. W. (2003). The identity crisis within the IS discipline: Defining and communicating the discipline's core properties. MIS quarterly, 183-194.
Beyea, S. C. (2004). Improving verbal communication in clinical care. AORN journal, 79(5), 1053-1056.
Boonstra, A., Versluis, A., & Vos, J. F. (2014). Implementing electronic health records in hospitals: a systematic literature review. BMC health services research, 14(1), 370.
Bryant, A., & Charmaz, K. (Eds.). (2007). The Sage handbook of grounded theory. Sage.
Burton-Jones, A., & Straub Jr, D. W. (2006). Reconceptualizing system usage: An approach and empirical test. Information systems research, 17(3), 228-246.
Calhoun, A. W., Boone, M. C., Porter, M. B., & Miller, K. H. (2014). Using simulation to address hierarchy-related errors in medical practice. The Permanente Journal, 18(2), 14.
Campanella, P., Lovato, E., Marone, C., Fallacara, L., Mancuso, A., Ricciardi, W., & Specchia, M. L. (2015). The impact of electronic health records on healthcare quality: a systematic review and meta-analysis. The European Journal of Public Health, 26(1), 60-64. Catwell, L., & Sheikh, A. (2009). Evaluating eHealth interventions: the need for continuous systemic evaluation. PLoS medicine, 6(8), e1000126.
Cawsey, T. F., Deszca, G., & Ingols, C. (2015). Organizational change: An action-oriented toolkit. Sage Publications. Chao, C. A. (2016). The impact of electronic health records on collaborative work routines: A narrative network analysis. International journal of medical informatics, 94, 100-111.Charmaz, K. (2006). Constructing grounded theory: A practical guide through qualitative analysis. sage.
Charmaz, K. (2000). Grounded theory: Objectivist and constructivist methods. Handbook of qualitative research, 2, 509-535.
Chen, L., Baird, A., & Straub, D. W. (2019). An Analysis of the Evolving Intellectual Structure of Health Information Systems Research in the Information Systems Discipline. Association for Information Systems.
Collette, A. E., Wann, K., Nevin, M. L., Rique, K., Tarrant, G., Hickey, L. A., ... & Thomason, T. (2017). An exploration of nurse-physician perceptions of collaborative behaviour. Journal of interprofessional care, 31(4), 470-478.
Crawford, K., & Hasan, H. M. (2006). Demonstrations of the Activity Theory framework for Research in IS. DiPalma, C. (2004). Power at work: navigating hierarchies, teamwork and webs. Journal of Medical Humanities,
25(4), 291-308.
Edmondson, A. (1999). Psychological Safety and Learning Behavior in Work Teams Amy Edmondson. Administrative Science Quarterly, 44(2), 350-383. Edmondson, A. C., & McManus, S. E. (2007). Methodological fit in management field research. Academy of
management review, 32(4), 1246-1264.
Eggenberger, T. (2012). Exploring the charge nurse role: holding the frontline. JONA: The Journal of Nursing
Administration, 42(11), 502-506.
Engestrom, Y., & Expanding, L. B. (1987). An activity-theoretical approach to developmental research. Learning by Expanding.
Foronda, C., MacWilliams, B., & McArthur, E. (2016). Interprofessional communication in healthcare: An integrative review. Nurse education in practice, 19, 36-40. Gibson, J. J. (1977). The theory of affordances. Perceiving, Acting and Knowing. Perceiving. Hillsdale, NJ: Lawrence
Erlbaum Associates.
