The influence of unintended consequences on affordance actualization of EHR systems Thijmen Kroeze, S2905663
t.b.kroeze@student.rug.nl
Msc Business Administration – Change Management Faculty of Economics and Business
University of Groningen
25 - 06 - 2018 Supervisor: Dr. B. Mueller Co-assessor: Prof. Dr. J. Surroca
Word count (excluding references, appendices): 8811
Abstract
To bring about organizational change, organizations implement IT. However, present theories of IT-associated organizational change lack attention to change goals, the role of IT in organizational change and the multilevel nature of change processes. Researchers started using the theory of affordances to research IT use in organizations and the related organizational changes. This paper focuses on the concept of affordances to extend knowledge on how users interact with Electronic Health Record systems (EHR) in hospitals. To answer the research question, and in-depth single case study was conducted at a Dutch health care group. Findings indicate that unintended consequences are, similarly to EHR-affordances, interrelated in a web of affordances and unintended consequences. Some unintended consequences rise from the combination of previous affordances and unintended consequences and would not emerge without them. These unintended consequences influence the multi-level processes that aggregate into the organizational change process.
Key words: affordances, unintended consequences, Electronic Health Record, affordance actualization
Table of content
1. Introduction………... 4
2. Theory………... 5
2.1 Affordances………. 6
2.2 Affordance actualization………. 7
2.3 Organization-EHR affordances………... 8
2.4 Unintended consequences………... 9
3. Methodology………. 10
3.1 Case selection……….. 11
3.2 Data collection……….... 12
3.4 Data analysis………... 13
4. Results………... 13
4.1 Unintended consequences………... 13
4.1.1 Familiarization effort………. 13
4.1.2 Effects of standardization………... 15
4.1.3 Data entry volume.………. 17
4.1.4 Organizational cooperation……….... 19
4.1.5 Reflective activities………... 20
4.2 Relation unintended consequences and EHR affordances……….. 21
5. Discussion………. 23
5.1 Theoretical and managerial implications……… 25
5.2 Limitations and future research………... 25
5.3 Conclusion……….. 26
References………. 27
Appendix A - Codebook………... 30
1. Introduction
In current times, workers rely heavily on the use of advanced software-based information technologies (IT) (Leonardi, 2011). IT provides new affordances for and constraints on the use of information in organizations (Leonardi, 2007). To bring about organizational change, organizations implement IT.
However, present theories of IT-associated organizational change lack attention to change goals, the role of IT in organizational change and the multilevel nature of change processes (Strong et al., 2014). Strong et al. (2014) argue that IT researchers should develop theories to explain IT-associated organizational change processes in a manner that enables actionable recommendations. Researchers started using the theory of affordances (Gibson, 1986) to research IT use in organizations and the related organizational changes (Leonardi, 2013; Majchrzak & Markus, 2013; Volkoff & Strong, 2013; Zammuto, Griffith, Majchrzak, Dougherty, & Faraj, 2007) . The affordances theory is suitable for IS research as it integrates the social and technical components of IT (Wieneke, Lehrer, & Jung, 2016). Identifying specific affordances provides an explanation what matters about a certain IT to certain particular organization, rather than inferring that specific IT features alone drive change (Strong et al., 2014).
This paper thus focuses on the concept ofaffordances
, originating from ecological psychology. It
will further develop the affordance-actualization lens (AA lens) as theorized by Strong et al. (2014).
Leonardi (2011) argues that ITs have material properties. However, these properties provide different possibilities for behaviors based on the context in which they are used. Despite the material properties of an IT are the same for each person, the affordances and constraints of that IT are not shared; they are uniquely perceived by an actor. Affordances therefore emerge from the relation between the technology and an actor (Bygstad, Munkvold, & Volkoff, 2016).
Affordances can lead to actualization
, which Strong et al. (2014) describe as the actions to take
advantage of one or multiple affordances by using the IT. Actualization takes place to achieve immediate concrete outcomes to support an actor’s own goals. These immediate concrete outcomes are specific expected outcomes that are viewed as valuable for achieving encompassing organizational goals.
However, in a recent study conducted by Tim, Pan, Bahri, & Fauzi (2018), it is found that affordance actualization not only leads to positive, intended consequences. Rather they propose to incorporate unintended consequences that result from actualization of affordances. By also identifying unintended consequences, scholars get a comprehensive understanding on the use and implications of technology (Tim et al., 2018). An action potential could be (1) unpredictable, perceptible, or hidden and (2) its
realization could diverge from what was initially expected and intended (Tim et al., 2018). Therefore, it is imperative to understand both affordances as well as the actualization process to fully understand the concept of affordances (Strong et al., 2014).
This study will build on the suggestion of Tim et al. (2018) to further develop the concept of unintended consequences in affordance actualization. The research gap filled will be one rising from the unexplored phenomenon of unintended consequences in affordance actualization. Based on the previous, the research question of this study therefore will be: “ How do unintended consequences influence the actualization of organizational affordances of EHR systems?”
This research will contribute to both theory
and practice, as it will provide deeper understanding of how affordances are shaped and how the actualization process is shaped and influenced. The main academic value is the answer to scholars’ call to develop understanding of how affordances are actualized (Bygstad et al., 2016; Strong et al., 2014; Tim, Pan, Bahri, & Fauzi, 2018) . The study will add managerial value to the extent that managers gain deeper insights on unintended consequences of IT-supported organizational change and may be able to anticipate these unforeseen actions or to benefit from the learning process.
To answer the research question, this paper will be structured in the following manner: a literature review will be given to discuss existing literature on information system affordances. Following, the method section describes how data is collected, coded and analyzed in order to generate theory on the unintended consequences of EHR affordances. Findings are presented in the result section. To complete the thesis, the discussion section will present a summary of results and related theory, both contributions to theory and practice, the limitations of this study, and a conclusion.
2. Theory
In this section relevant literature will be discussed on affordances, affordance actualization, previous research on EHR-affordances and unintended consequences.
2.1 Affordances
The concept of affordances originates in ecological psychology research. Gibson (1986) argued that a person or animal looking at an object does not see its qualities, but rather perceives what the object will permit them to do. An affordance can therefore be described as “what is offered, provided, or furnished to someone or something by an object”
(Strong et al., 2014) . Wieneke et al. (2016, p. 5) provide a clarifying
example:
“For example, a museum guard, standing all day long, perceives the affordance to sit down when he sees a chair (with the object’s properties being horizontal, flat, stable, rigid and knee-high). An affordance is related to the actor’s characteristics and objectives, so that people with different goals can interpret the likely uses of an object differently. To clarify this, we use the previous example again: a craftsman needs to replace a ceiling lamp in the museum. When he sees the same chair as the museum guard, he perceives the affordance to stand up on the chair. Accordingly, each object can afford multiple action possibilities to different actors; it can also afford one actor with different objectives.”
The identification of affordances is a method to understand the action possibilities as perceived by the user. Affordances in IT research emerge from the relationship between the actor and the object and refer to potential uses of an IT and not solely to actual uses or realized action (Markus & Silver, 2008;
Wieneke et al., 2016) . Affordances materialize when a person interprets a technology through his goals for action (Leonardi, 2011). In this paper, affordances are defined as “the potential for behaviors associated with achieving an immediate concrete outcome and arising from the relation between an artifact and a goal-oriented actor or actors
” (Strong et al., 2014, p. 69).
Researching the affordance theory is not without its problems. Strong et al. (2014) argue three key issues in using the affordance theory. First, most IS scholars focus solely on the possibilities for action rather than taking the actual actions taken and their outcomes into account. Therefore, it is important to clearly distinguish the affordance itself (the possibilities for goal-directed action), its actualization (the actions actually taken), and the outcomes of those actions. Second, while affordances relate to a goal-directed actor, it is still to be determined which or whose goal this is: individual, group or organizational goals. To solve this, it is recommended to focus on the immediate concrete outcomes for the individual. Third, actualizing an affordance does not happen in a vacuum as there are multiple affordances to an object. These bundles of affordances need to be considered and the ways these affordances interact. Following these key issues, (Volkoff & Strong, 2017) defined six key principles to follow when researching affordances in IT. These principles are presented in table 1.
Principle 1 Affordances arise from the user/artifact relation, not just from the artifact.
Principle 2 Maintain the distinction between an affordance and its actualization.
Principle 3 Focus on the action, not the state or condition reached after taking the action.
Principle 4 Select an appropriate level(s) of granularity for the affordances.
Principle 5 Identify all salient affordances and how they interact.
Principle 6 Recognize social forces that affect affordance actualization.
Table 1 Key principles of studying affordances
2.2 Affordance actualization
Where the affordance is the potential for achieving a goal, the actualization relates to an individual actor and details regarding the specific actions that actor has taken(Volkoff & Strong, 2017). As mentioned in the introduction, the affordance actualization process is described as the actions to take advantage of one or multiple affordances by using the IT, to achieve immediate concrete outcomes that support organizational goals. Focusing on the actualization process and the immediate concrete outcomes enables researchers to capture the real moments where affordances are realized (Tim et al., 2018). The concept of affordance actualization is underexplored and needs more development, which Strong et al. (2014) laid the theoretical basis for.
Identifying affordances is a first step for understanding how organizational change is implicated by a technology (Strong et al., 2014). Focusing on actualization provides the foundation for increasing the understanding of resulting changes in organizations after affordances are actualized (Volkoff & Strong, 2017). As users actualize an affordance, they engage in a learning process in which they may encounter many difficulties. Actualization is not a linear process and therefore affordances may not be fully actualized. According to Strong et al. (2014), actualization involves 1) a set of individual journeys occurring in a dynamic organizational context as users learn and managers intervene and 2) a multi-level process in which the journeys of many individuals interact and aggregate to form an organizational journey, which is the organizational change process. A recent study by Tim et al. (2018) provided an early attempt that explores the affordance actualization of an emerging technology.
2.3 Organization-EHR affordances
This paper will build on previous work of Strong et al. (2014), who identified eight affordances arising from the relation between the EHR and the healthcare organization. Those eight affordances are essential to analyzing of the use of an EHR can result in organizational change. An organizational affordance is an affordance where the potential actions are associated with achieving immediate concrete outcomes on organizational level, in order to achieve organizational level goals. Table 2 provides an overview of the affordances and the organizational context needed to actualize this affordance.
Affordance Goals-directed actions needed to actualize an affordance 1) Capturing and archiving data
about patients
Individuals record all appropriate data about patients and interactions with them in the EHR, including provider notes.
2) Accessing and using patient information anytime from anywhere
Individuals retrieve all relevant and available data pertaining to each patient encounter.
3) Coordinating patient care across sites, facilities and providers
Individuals involved in coordinating care use the EHR’s
coordination features, rather than phones, for communication and coordination about patients and their care.
4) Standardizing data, processes and roles
Individuals create and use templates in the EHR.
Individuals follow standard procedures for data, tasks, and roles, and participate in setting standards as needed.
5) Monitoring organizational operations
Individuals regularly check the EHR and take actions to ensure smooth operations and completion of their own or others’
workloads.
6) Substituting healthcare professionals for each other
Individuals communicate work requests to message pools.
Individuals use standardized data formats in patient records.
7) Incorporating rich information into clinical decision making
Individuals consult the EHR’s alerts and recommendations as part of their decision making process.
8) Shifting work across roles Individuals use templates with embedded rules for performing a task, and do not deviate from the standardized rules without first checking with someone in the appropriate role.
Table 2. Organization-EHR affordances based on Strong et al. (2014).
Strong and colleagues also argue that many individuals encounter multiple affordances at once while engaging in actualization of the EHR affordances. These multiple affordances are interrelated in various ways and actualizing them may contribute to achieving multiple goals. Relationships between affordances are temporal, as affordances are not all actualized at once; instead actualization occurs at different times. The actualization of affordances in some cases can depend on the immediate concrete outcomes from preceding actualization of other affordances. Furthermore, While the affordances exist as potentials, some actors, EHR aspects or some organizational context aspects may not be ready yet for actualizing them. For instance actors can actualize basic affordances and therefore building their knowledge and skills in manners that enable them to perceive and actualize more advanced affordances they could not before. The web of dependencies is presented in the figure below.
Figure 1. Affordance dependency diagram (Strong et al., 2014)
2.4 Unintended consequences
Furthermore, Tim et al. (2018) discovered a feedback effect as the unintended consequences generated by actions in actualizing affordances will provide feedback for adjusting the actions of actors, and therefore adjusting the use of technology. Predominantly, previous studies subconsciously avoid acknowledging IT-related harms (Majchrzak, Markus, & Wareham, 2016). Ultimately, these unintended consequences influence the multi-level process that aggregates into the organizational change process. Unintended consequences can seem harmful at first sight, nevertheless the outcomes of these consequences stimulate goal-oriented actors to take further actions, i.e. working around or addressing the undesirable implications (Tim et al., 2018) . Concluding, Tim et al. (2018) build on the conceptual framework first drafted by Strong et al. (2014) to illustrate the affordance actualization process and included the unintended consequences to deepen existing theorization (Figure 1, next page). This research will explore between the unintended consequences and adjusted actions of goal-oriented actors.
Figure 2 Conceptual framework (Tim et al., 2018)
3. Methodology
In the following section the methodological approach for this study will be discussed. The section contains three parts: selection of cases and participants, data collection and data analysis. As previously argued, this research contributes to the theoretical development of the use of the affordance theory in IS research. It tries to explain a currently underexplored phenomenon: the unintended consequences of affordance actualization and the feedback effect as found by Tim et al. (2018). As by answering the research question new and accurate insights in a specific topic are generated and the phenomenon is still exploratory in nature, the most suitable approach is one of theory development (Brown & Eisenhardt, 1997; van Aken, Berends, & van der Bij, 2012).
This study will follow a grounded theory method as it serves as a specific method for theory generation and in this research to develop new theories of information systems phenomena (Corbin &
Strauss, 1990; Urquhart, Lehmann, & Myers, 2009) . The grounded theory approach advocates that there should be an ongoing interplay between data collection and analysis (Urquhart et al., 2009). This enables the researcher to capture all potentially relevant aspects of the topic as soon as they are perceived (Corbin
& Strauss, 1990). Urquhart et al. (2009, p. 359) list four distinctive characteristics of the grounded theory method:
1) The main purpose of the grounded theory method is theory building.
2) As a general rule, the researcher should make sure that their prior – often expert – knowledge of the field does not lead them to preformulated hypotheses that their research then seeks to verify – or otherwise. Such preconceived theoretical ideas could hinder the emergence of ideas that should be firmly rooted in the data in the first instance.
3) Analysis and conceptualization are engendered through the core process of joint data collection and constant comparison, where every slice of data is compared with all existing concepts and constructs to see if it enriches an existing category (i.e. by adding/enhancing its properties), forms a new one or points to a new relation.
4) ‘Slices of data’ of all kinds are selected by a process of theoretical sampling, where the researcher decides on analytical grounds where to sample from next.
3.1 Case selection
The selection of cases is an essential aspect of building theory from a case study (Eisenhardt, 1989). The goal of theoretical sampling is to choose cases that are likely to replicate or broaden the emergent theory (Eisenhardt, 1989). Crucial in this research is to find a case that matches the requirements in table 3.
Requirement Explanation
1. The organization uses information technology to improve working methods
As this study focuses on the affordances theory in IS research, an appropriate case should make use of IT.
2. The technology fundamentally changed the way organizational members work
It is likely for unintended consequences to be most visible when the technology implemented has a significant impact on the worker’s methods, as workers have to change their routines.
3. Organizational members are willing to cooperate in the research
Data will be collected through semi-structured interviews with the workers, managers. Therefore it is essential that organizational members are willing to participate.
Table 3 Case selection requirements
The case selected is a Dutch healthcare group, comprising three hospitals and seventeen nursery centers.
The organization employs 300 medical specialists and 6,100 staff members. In total, the hospitals have 800 beds and provide general care for a catchment area of 300,000 citizens. As many hospitals in the Netherlands in current times, the healthcare group migrated from a paper-based working method to an
Electronic Health Record (hereafter: EHR) to ensure a more cost and time efficient method of working and therefore meets requirement 1 as stated in Table 2. The EHR drastically changes workers’ routines and working methods as argued in requirement 2. The organization agreed to serve as a case study and logically meets requirement 3.
Expectations regarding EHRs are high, and the benefits from the use of technology in healthcare environments can be substantial. Yet, introducing EHR is a complex and often unpredictable undertaking.
An EHR has proven more difficult to adapt to clinical practices than initially assumed (Jensen &
Aanestad, 2006). The cradle of this EHR implementation lies in a merger between three regional hospitals into one hospital group. Either hospital had its own IT architecture and their own EHR, both not compatible across the sites. As pressure from governmental inspection organizations to deliver better registration quality grew, the need for a central architecture and one EHR software suite increased. With the pressure rising, top management decided for an accelerated implementation of a cross-site EHR system, with the goal of increasing patient safety and registration quality, and cancel the enhanced surveillance of the governmental inspection agency.
3.2 Data collection
In this study primary data will be used to answer the research question. The qualitative instrument used is the semi-structured interview. Semi-structured interviews allow the researcher to collect consistent primary data across multiple individuals, as well as also providing the opportunity to ask more in-depth questions (Yin, 2009). Further, archival documents will be analyzed to gather secondary data to ensure triangulation (Yin, 2009). Multiple data collection methods are combined to provide stronger substantiation of constructs (Eisenhardt, 1989). To improve reliability and ensure consistent data, interview protocols will be drafted. One interview will last approximately sixty minutes. With informed consent of respondents the interviews will be recorded and transcribed. Reliability of the study will also be increased by writing memo’s of all procedures during the research to enable replication (van Aken et al., 2012). Van Aken et al. (2012) propose four potential biases that need to be controlled, which can be found in table 4.
Bias Control
Researcher bias Inter-rater reliability will be ensured through letting the data also be coded by a second researcher. An interview protocol is applied as well.
Instrument bias By using triangulation data collection methods can correct or complement each other.
Respondents bias Both crucial roles (managers, workers) are included in the research.
Circumstances bias Respondents will be interviewed at different times in a private room.
Table 4 Potential biases (Van Aken et al., 2012)
3.4 Data analysis
At the heart of building theory from case studies is analyzing data, which is regarded as the most difficult and least codified part of the process (Eisenhardt, 1989). In order to convert transcripts into a useful depiction of the gathered data, the transcripts will undergo analysis through the process of grounded theory research where three basic types of coding are used: open, axial and selective which is described in the next section following the definitions of Corbin & Strauss (1990). All transcripts are coded by one researcher, starting at the first moment of data collection.
Open coding is the interpretive process by which data are broken down analytically in purpose to provide the researcher with new insights through standardized thinking methods of interpreting phenomena from the data. Open coding help the researcher to guide questions when returning to the field.
In axial coding, codes found in open coding are converted into more abstract categories and related to sub-categories. In later stages of research selective coding is used. This is defined as the process by which all categories are unified around a core category, which represents the central phenomenon of the study.
4. Results
The findings of this study led to a theoretical understanding of unintended consequences of EHR affordances. This section will provide an overview of all unintended consequences and their relationship to the eight affordances defined by Strong et al. (2014). As discussed in the literature review, an unintended consequence has a feedback effect generated by actions in actualizing affordances. Through this feedback effect causes actors to adjust their actions and therefore adjusting the use of technology.
Conclusively, these unintended consequences influence the multi-level process that accumulates into the organizational change process.
4.1 Unintended consequences
4.1.1 Familiarization effort.The first unintended consequence results from familiarization effort before go-live of the EHR. This effort is two-sided: management and the project team (hereafter: project
team); and users (medical assistants, nurses, physicians). The project team tried to familiarize the users through training and information provision such as newsletters and tutorial videos. As respondent 2 describes: “We provided them with a lot of information on forehand. But the question is how much the receiver is going to read it in the workload. [...] So we covered people in all kinds of information, but they didn’t read it. So we made short movies, they didn’t watch them. [...] So on forehand we sent a lot, but it wasn’t received.”
The project team tried to create a sense of urgency with users pre-go live, but noticed
that little of what they sent was received. The overall tendency was that using a patient focussed, cross-site EHR would not change their working methods to a large extent. Users figured that the introduction of the cross-site EHR would not impact their work significantly, since they already used digital EHR-prequels. A small e-learning was introduced and made compulsory for system access on go-live. Respondent 5: “Yes, I would have liked to have a little more support, but that’s not only the organization’s fault. It’s also our own fault. Then I had to close my clinic for one or two weeks to go deeper into the system to learn all ins and outs. At a certain moment we had a tiny basis course for which we had to do a little test. I passed that one, but at some point you don’t know anything anymore. And if it really happens you know a few fundamental actions in no time, in order to work. But then you miss the details. For that, you need a lot more training. But that comes at the expense of patient care, that’s the major issue.”
The e-learning was
considered too basic and too early to remember upon go-live. Users felt they were not able to make use of the EHR’s full potential as they did not have the time to familiarize themselves with it. For this however, they do not solely blame the project team, but also looked at themselves. Mainly physicians struggled to make the system their own as they argued that they had to close down their clinics for one or two weeks in order to learn ins and outs of the EHR. This responsibility is echoed by respondent 1: “It was of course also up to the physician himself if he was going to await it, and to shout a day before the EHR was introduced that he doesn’t want THAT [the EHR in the way it worked then]. I mean, there is some responsibility for the management, but also for the physicians, also for the secretaries.”
The little interest in training pre-go-live was also noticed by the project team, as the attendance at training sessions is low, especially by physicians. Medical secretaries and nurses spent more time on training and receiving information which resulted in easier use on go-live. As respondent 9 explains: “We implemented the EHR, also for the nurses. And when you see that all the nursery meetings we held were attended well. It was an enthusiastic group of nurses. I don’t know what it is with physicians, but you invite them twenty times, they won’t show up nineteen and will show once. And at the time of implementation they say: ‘But I haven’t had any training.’”
Top management recommended clinics to
halve their amount of patient visits in the first weeks after go-live, but the decision of doing this was still with the physicians. Some departments decided to follow the advice, while others continued their work.
UC1: Users’ lacking system knowledge.
The following unintended consequence is that users
lacked system knowledge upon go-live. Users were able to use the very basic features of the EHR but many were not able to use it to support their work processes in full. Due to this lack of knowledge, workers struggled to integrate the EHR in their daily routines, as can be read in the next paragraphs.
Respondent 6 exemplifies: “ I think it has been implemented too quick, as many people don’t know how to work with it, they still don’t. I notice that as I work on all departments. I notice how departments use it differently. Just something simple as a patient note. You can just print it in the EHR. Almost no department does that, they make their own patient notes as they consider it easier. It doesn’t make any sense.” In this example, nurses fail to use standardized, quicker features due to their lack of understanding the possibilities of the EHR.
4.1.2 Effects of standardization. The second unintended consequence results from the effects of standardization through EHR. Using an EHR enables organizations to exchange staff, patients and to improve service and care quality. However, this requires the organization to standardize its work processes and data entry methods related to EHR. As respondent 7 illustrates: “Why do we want that standard?
Because we can have predictable service, quality is just predictability of outcome. Better service to our customers, but also the support for the physician.” When registration emerges uniformly, reliable quality parameters can be established. Yet, as the hospital sites used different EHR systems pre-merger, uniform registration caused deteriorating registration first at sites after implementation. Respondent 7 explained:
“It means that for one it is an improvement, but for the other its a step back. You first have to equate your clocks. [...] So first you’re going to equalize and from there you have a new foundation to build on. The illusion with many was that we went to a new EHR, so it will be better than what we had [before].”
Following implementation and standardization, work routines had to be adapted to the new situation. Respondent 3 elaborates: “[previously] You could shape it as you wanted in your clinic. We wanted to see that again in the EHR, but that didn’t work out completely. It makes sense, as we all had to work uniformly.” Medical staff and secretaries had to adapt their daily schedules and work methods to ensure time availability for data entry and preparation of information. Tasks that were completed routinely over the years required more effort and thought in order to fulfill them. Automatisms had to be renewed and the secretary-physician relationship was affected as task management shifted from direct contact to
digital task lists. Due to the use of the EHR and the standardization tasks shifted, resulting in changing cooperation between physician and support staff. Respondent 4: “But I do the work, not my secretary. I used to have a tape and that was brought over there, and she wrote it up. Now I am typing myself and with hotkeys I finish that letter. So that costs time, but then expensive workers do cheap labor. And because of that I see fewer patients per timespan, and it means that my support staff do nothing in that timespan, so to speak.” Shifting documentation responsibility from secretaries to physicians caused a decline in productivity for the latter, as can be seen on more occasions in this case.
UC2: Affected professional identity.
The subsequent unintended consequence of standardization
therefore is the effect on professional identity of medical professionals. Professional identity is defined as one’s professional self-concept based on attributes, beliefs, values, motives and experiences (Slay &
Smith, 2010). Nurses view themselves as craftsmen who can put their skills and experience to work in order to provide best possible care for their patients, depending on the situation. This diminishes through standardization, as the EHR rather than the professional dictates which care patients need while situationality, experience and craftsmanship is not considered. As respondent 6 argues: “If you look at the younger generation, I think the clinical eye diminishes. It isn’t taught anymore. You have to learn it in school, but in practice you don’t have to use it, so to speak. So I believe that the quality of the profession decreases on that point.” Nurses are, since the coming of the EHR, bound to planned patient check-ups three times a day, even if they do not deem is necessary. This will elaborated in the next paragraph on data entry volume.
While historically seen physicians were exempt from administrational tasks as they were delegated to secretary staff, EHR affects their professional identity as they are responsible for filing their patients’ medical documentation themselves. Physicians’ self-image as organizational VIPs is diminishing as support staff decreases and administrative tasks are performed by themselves. Respondent 2 explains:
“The work of a medical secretary, compared to 15 years ago while we were still working on paper. At the time the secretary was the right hand, she took care of everything for the physician. Really everything.
Then it was a very varied job. That job has been gradually emptied out and is being differentiated. The secretary wants to keep her job and holds on to what she can get. She wouldn’t say to the doctor that he has to do it by himself. [...] That physician doesn’t want to lose his ‘girls’, as its called, he doesn’t want to lose his secretaries.” The previous quote also indicates a changing professional identity of the secretary.
While previously seen as a jack-of-all-trades fully supporting their physician, most of their work portfolio
has been reduced to specific supportive tasks as making appointments and guiding patients in clinic visits.
Considering the shift to upcoming online platforms where patients will make their own appointments and have the ability for direct contact with the physician, the secretary function seems to be disappearing.
4.1.3 Data entry volume. Following the standardization of data entry and work processes, the volume of data entry for nurses and physicians grew as illustrated by respondent 2: “When the EHR came, it was very noticeable that the registration burden became huge. Irrespective of the digital registration, third parties requested all sorts of quality indicators, flags, quality marks. It was crazy how much they grew at the time. It was like: ‘Hey, you register, so you can deliver. Then we’re also going to ask more.’”
Dutch hospitals are required to deliver quality data to regulators which they can extract from the EHR.
Besides this mandatory data, hospitals can choose to strive for quality marks by third parties in order to achieve special recommendations for certain treatments, to improve market position. The third parties require other, specific data that is not necessarily available from what is entered into the system by regular standards. Management chose to add this functionality to the EHR and thus increase the volume of data, enabled by the use of the EHR.
Another goal of data entry is to enhance transparency of treatment and ensure quality care. As Respondent 5 points out: “The reason that it’s done this way, is to get it transparent, to make it clear what we’re doing and to adjust compensation. [...] It became a very unclear whole unfortunately.”
Insurance
companies required more specific information on treatments in order to improve billing and compensation.
However, this increased complexity and volume of data entry of treatment plans, which caused physicians to spend more time on registration rather than delegating work to secretaries. As respondent 4 exhibits:
“But the discussion was: what is the secretary allowed to do? They’d prefer that the secretary wasn’t allowed to fill in anything. [...] They may want it by law and from the [professional] roles you have, but it is not going to work. If you want that, we’ll see even less patients.” Management implemented key performance indicators at nursery departments to ensure data entry. This caused nurses’ workload to increase, as respondent 6 exemplifies: “I look at my KPIs, you have so many of those scores you have to do everyday, how does that make sense? If someone does not have a wound on his bottom in the morning, usually he won’t have it in the evening. And then I think: ‘I can enter that, but why do I have to do that three times a day, mandatory?’ It’s also not one checkmark, you have to there and tick three-four things. I think it got very intensive and a lot more.” The KPIs led to an increase in mandatory checks which not always considered necessary by the nurses, as will be explained in the effects of standardization in the
next paragraph. Another managerial issue was a conflict of interests between the business side and the care side of the organization. The business side requires precise registration to invoice health insurers without loss. However, physicians want to record as little as possible, as record keeping costs time that can not be spend on care.
UC3: Less time spent on care.
Following the above, staff experienced an increase in workload.
This increase led to an unintended consequence of less time to spend on providing medical care. As the volume of data entry gained, staff had less time to spend on doctor-patient and nurse-patient contact. Some doctors decided to see fewer patients over the same timespan as before the EHR, due to the time needed for administration. Another reason for this, is that many nurses and physicians struggle with their typing ability. Typing courses were provided by the organization, but potential candidates decided not to follow them. Also, informal contact moments with colleagues reduced, lowering regular meetings to discuss ways of working. As respondent 4 captures it: “Since then it influences my daily routine that I have to skip my coffee break, that my office hours run out, so I decided to see fewer patients over the same timespan.”
The
disappearing of these moments caused the department to rarely discuss the newly implemented EHR on best practices and issues with physicians and support staff. Also consultations on departmental issues and improvements lowered since the coffee break was the daily moment to recap across functions how the department was performing.
More innovative users started using software to enhance their ability to quickly produce data and enter it in the system to reduce data entry volume and time. For example, when requesting a printed medication prescription directly from the EHR, the system will first download a file to the computer, asks for opening in Microsoft Word from where the file can be printed, rather than providing a standard template that is printed. Innovative users coded a path that automates these steps. This software requires coding skills and is therefore not suitable for the majority.
As previously mentioned, nurse-patient contact decreased after EHR introduction. As respondent 8 capture:“Yes, I think we spend less time with the patient. I can remember during my internship in the clinic downstairs, that I would sit at the patient's bed and wrote on the table in front of the patient. The files were still behind the bed then. That’s gone. You do ask [the patient], you save that information and [later] you will type it.” Nurses experience that writing on paper and talking to the patient simultaneously is easier than typing and talking. Also, writing provides a more personal approach for the patient. Both physicians and nurses experience that looking at a screen while typing and talking creates a distance
between patients and care providers where in some cases patients made remarks on this fact and asked for another approach. Nurses also spend more time on data entry as they struggle to bring their computers-on-wheels (COWs) in nursing rooms due to the available room, the heat and noise of the computers and the danger of infection for specifically vulnerable patients. Therefore, they write information down on paper or remind it and enter it into the EHR afterwards.
4.1.4 Organizational cooperation. Due to the vast amount of data entry in the EHR, the availability of information increased significantly. Introduction of the EHR reduced the number of hard copy files to a bare minimum, improving availability of patient information and therefore enhancing the quality of care. As respondent 6 describes the situation pre-EHR: “You had files of the doctor, of nurses, some behind the bed. There were files everywhere. And then you had to wait on one as it was somewhere else. That was the biggest expectation I think, that you could get everything. And everything was available.”
The EHR also enabled the organization to remove archives and discard many of its
paper-based document management activities. Where pre-EHR notes of physicians and nurses tended to turn out missing and potentially losing vital information, EHR provides extensive history, notes, lab results and other forms of patient information. Medical professionals are able to provide better quality of care due to comprehensive knowledge of patients’ medical history, notes from colleagues.
The largest improvement however, was that through EHR departments were facilitated in multidisciplinary cooperation, leading to improved decision making, diagnosing and mood. The latter being explained by respondent 1: “At that time it was implemented by phase and we we’re already up and running for quite a while before the next clinic was done. It happened regularly that someone came: ‘Can I see how you do it?’ and that was quite mood improving.” Enablement of multidisciplinary cooperation and rich information availability also improves quality mark ratings, which was explained in the first paragraph.
Pre-EHR, departments existed as separate entities with little reciprocal interaction in two dimensions: 1) between disciplines on a single site and 2) between identical disciplines on multiple sites.
Interaction between disciplines on a single site grew more naturally as staff started to enrich their knowledge and decision making by using the available data from multiple disciplines. In a few cases multidisciplinary meetings became the norm rather than the exception. One spokesperson was appointed per patient to serve as the connecting element between the departments. This development improved quality of care significantly. Interaction in the second dimension grew as management decided that one
discipline over three locations had to use the same design of their EHR and adapt its work processes accordingly. Growth in the second dimension was therefore forced by management.
UC4: Use of same language.
The unintended consequence of growth of multidisciplinarity is the
use of the same language. Organizational cooperation strengthens standardization as multidisciplinary communication requires the use of the same language such as similar terminology and standardized reporting structures. Pre-EHR every department had the liberty of using own templates and physicians had their own style of reporting. The EHR provides this structure and supports the use of similar terminology, as respondent 8 explains: Now you’ve got an overview, like every report divided in several themes:
excretion, nutrition, liquids, so you can enter at every theme what deviates. [...] Now that is really checked.” The quality of information increased through the use of the same language, enabling multidisciplinary cooperation.
4.1.5 Reflective activities. The last unintended consequence identified from the data follows from an aggregation of the previous discusses consequences.
UC5: Reflection and improvement. Organizational members noticed work that by the introduction of the EHR their work routines were heavily influenced and their professional identity changed. To cope with these changes, some workers started to look for ways to improve their work methods without waiting for initiatives from ranks higher up the organization. As respondent 1 describes:
“What we did here was to brainstorm an entire day and put all our activities on paper. How are we going to do it, is it necessary, can we do it better. In that sense is the implementation of an EHR or what system it may be, a positive thing.”
The idea of a brainstorm session was formed by the workers themselves,
rather than pushed by their practice manager. Workers tried to use the introduction of the EHR as a handle to reflect on the way they worked and why they did it that way. They came up with an own framework to reflect their activities and adapt them to the EHR, to benefit as much possible from the digitized way of working. The provided training was solely focused on features, rather than promoting work methods reflection.
The project team also supported workers reflecting on their processes following missing feature requests after implementation. Respondent 2 provides an example: “They came to me for a missing form.
That form had to be in the system. [...] If they had to prepare a patient for surgery, they had to fill it out.
Why? Yes, it went to the admissions office, but that paper trail was gone. Then it had to go to the nursery,
as they have to put it in a folder. But the folder was also gone. And this way we looked at many things and in the end it turned out the the last person tossed it away, so it has always been nonsense.” Similarly many unnecessary forms were disposed through the EHR introduction, in order to reduce waste and information pollution.
# Unintended consequence Explanation 1 Users’ lacking system
knowledge
As users did not recognize the urgency of the EHR implementation, training and information consumption was low. Subsequently at go-live, users lacked sufficient knowledge of the EHR-system to use its full potential.
2 Affecting professional identity
Use of EHR requires standardization of data entry and processes. This standardization causes 1) physicians to spend significantly more time on administration, 2) lower nurses’ craftsmanship and 3) medical secretaries lose their reputation as jack-of-all-trades. These three factors combined results in a different view of their professional identity.
3 Less time spent on care As data entry and processes became standardized, the volume of data entry for nurses and physicians grew. Combined with insufficient system knowledge, medical professionals’ workload grew so physicians decided to see less patients over the same timespan as pre-EHR. Nurses have less contact with patients due to time spent on data entry.
4 Use of same language Cooperation between departments on-site and cross-site required the use of standardized templates and terminology, as was provided through the EHR. Using the same language enables cooperation.
5 Reflection and improvement
As by the introduction of the EHR staff’s work routines were heavily influenced and their professional identity changed, groups of users engaged in bottom-up improvement activities to integrate the EHR in their routines.
Table 5. Summary of unintended consequences
4.2 Relation unintended consequences and EHR affordances
In this paragraph the previously named unintended consequences will be linked to Strong et al.’s (2014) eight organization-EHR affordances. For readability, unintended consequences will be abbreviated to UC(s)(for plural). As previously discussed, individuals can encounter multiple affordances at once while engaging in actualization since affordances are interrelated. Therefore, UCs are also part of this web of affordances. Some UCs are only possible through the actualization of an affordance and influence the
outcome of multiple affordances. Furthermore, unintended consequences also are mutually interrelated, meaning that throughout this case UCs rising from early affordances caused new UCs to emerge, as is shown in the figure below.
Figure 3. EHR-unintended consequences and affordances dependency diagram
UC1: Users’ lacking system knowledge
The first UC precedes the first affordance, namely capturing & archiving digital data about patients
. By
lacking system knowledge, users had difficulties with recording all appropriate data about patients in the right places and to fit it in their routines. It influenced the extent to which users were able to actualize this affordance, as they did not perceive the possibilities for action the IT artifact provided them. Combined with the affordance of standardizing data, processes and roles, users were able to spend less time on care and more time on administrative tasks, which is UC3. UC1 also has its influence on UC2, which will be explained hereafter.
UC3: Less time spent on care
Following the order of the figure above, UC3 will be discussed first. Cause for the reduced amount of time that could be spent on care is the first affordance combined with the fourth affordance. As digital data had to be captured and archived, the data entry volume and time spent on entering increased. At the same time, processes and data entry was standardized, requiring workers to enter more data on specific moment in their shift.
UC2: Affecting professional identity
Adding to the previous, the UC2 emerges from the standardization of data, processes and roles, the fourth affordance; UC3 on the reduced amount of time spent on care and UC1 on lacking system knowledge. The standardization affected medical staff’s self-image as independent craftsmen who had the knowledge and
skills to provide the best care for patients, without following strict procedures and subsequent administrative tasks. The decrease in time spent on care affected the professional identity as medical staff had to increasingly spend time on administrative tasks, which they experience as ballast and is not why they chose the work they did. Concluding, the lacking system knowledge affected their professional identity resulting from feelings of incapability of doing the job which they did for years.
UC4: Use of same language
The fourth UC is more independent than the others. As a result of multidisciplinary cooperation, people started to use the same language in terminology and reporting structures. Through this the EHR enabled staff to discuss work with other departments and sites, enabling the fifth UC.
UC5: Reflection and improvement
The last UC emerged from a combination of UC2, UC4 and impact affordance 5 on monitoring organizational operations. As physicians, nurses and medical secretaries’ professional identity changed, and throughout the organization people started to use the same language, activities emerged within departments and sites where groups of co-workers joined forces to reflect on their routines and improved these to integrate the EHR in their work. This influences the fifth affordance as people strive to improve the quality of operations and recognize opportunities for organizational improvements.
5. Discussion
This research focused on the unintended consequences resulting from the actualization of EHR affordances. It attempts to answer the following research question: “ How do unintended consequences influence the actualization of organizational affordances of EHR systems?” Findings indicate that unintended consequences are, similarly to EHR-affordances, interrelated in a web of affordances and unintended consequences. Some unintended consequences rise from the combination of previous affordances and unintended consequences and would not emerge without them. These unintended consequences influence the multi-level processes that aggregate into the organizational change process and therefore confirms the feedback effect in affordance actualization as theorized by Tim et al. (2018).
The result of this research are five unintended consequences existing in the web of dependencies as argued by Strong et al. (2014) in their research on organization-EHR affordances, which was fundamental to this research to built upon.
In the following section these answers will be explained in depth by analysing our findings and enfolding the literature to analyse conflicting as well as supporting literature. Furthermore, contributions for practice as well as literature, limitations of this study and possibilities for future research will be discussed. The section ends with a conclusion.
The first unintended consequence is the lack of user’s system knowledge due to a disinterest from users in the EHR, pre-implementation. Attendance at provided training was low, especially from physicians. Also information provided through other communication channels such as newsletters, physical letters, tutorial videos, were hardly received. The lack of knowledge influenced affordance actualization, especially shortly after go-live, as actors were not ready for actualization. This is in line with Strong et al. (2014), who argue that actors, EHR aspects and organization context aspects may not be ready to be actualized. Actors first have to actualize more basic affordances so they can build their knowledge and skills, to perceive more advanced affordances. Good training and information can help actors to perceive advanced affordances, and potentially actualize these earlier on in the adaptation phase.
This can increase innovative use and may able users to use technology in novel ways to support their daily work.
The second unintended consequence is the effect on professional identity. After implementation, the EHR caused nurses and physicians to increasingly spend more time on administration and logically less time on actual contact with patient, or providing care. Following, medical professionals felt that they were becoming secretaries rather than doing what they were educated to do and experienced in. Also, where secretaries traditionally were considered jack-of-all-trades and right hands of physicians, their tasks increasingly focussed on administrative work, meaning spending entire days on a computer.
The third unintended consequence results from a aggregation of standardization and the increased data entry volume. As the volume of data entry gained, workers had less time for on doctor-patient and nurse-patient contact. Actors adjusted their actions by reducing their amount of patients in their clinics and canceling informal breaks with co-workers. This both contradicts and confirms earlier research on time spent by nurses and physicians. Previous research suggest that nurses might be more time efficient than physicians, as autonomy and accountability of nurses is different (Poissant, Pereira, Tamblyn, &
Kawasumi, 2005). Therefore, physicians may spend more time on documentation than nurses to ensure quality of information. Poissant et al. (2005) however do report a decrease in time spent by nurses on documentation, while nurses in the case organization experienced this different. An explanation for this
may be that while in the paper-based way of working most reporting was done at the end of the shift and all at once, registering throughout the day may feel as spending more time on it.
The fourth unintended consequence was the use of the same terminology and reporting structures as a result of multidisciplinary cooperation. The fifth consequence comes about from a combination of the second and fourth unintended consequence. As physicians, nurses and medical secretaries’ professional identity changed, and throughout the organization people started to use the same language, activities emerged within departments and sites where groups of co-workers joined forces to reflect on their routines and improved these to integrate the EHR in their work.
5.1 Theoretical and managerial implications
This research has made two contributions to the extant change management literature. First, as previously mentioned in the introduction, on contribution of this research is extending the knowledge on the concept of affordance actualization. This study indicates that from actualizing organizational affordances, unintended consequences can emerge that will influence the actualization of other affordances. Another contribution is the extension of literature on EHR systems by reviewing a detailed case and profit from the richness of the data.
In organizational change, managers usually prefer a more structured, planned approach. This research has managerial value in that managers gain deeper insights on unintended consequences of IT-supported organizational change and may be able to anticipate on unforeseen actions following from actualized affordances. Also, managers should strive to recognize the differences in affordance actualization between different professions within one department to improve earlier advanced use of EHR.
5.2 Limitations and future research
This research has several limitations. First, the list of EHR-affordances used was from an American health care group. As this case study is conducted in the Netherlands and both hospital systems do differ, some affordances might be overlooked. Another limitation is the number of respondents that can restrict fully grasping the entire story of with nurses, secretaries and physicians from all three locations of the group. A third limitation is that all respondents had to recall information on events that began three years ago.
Therefore, some information might be missing or overlooked.
Future research directions can be to research other systems’ affordances and emerging unintended consequences to further conceptualize affordance actualization and explore the phenomenon of unintended consequences.
5.3 Conclusion
This study explored how unintended consequences influences affordance actualization of EHR-affordances. Research showed that those unintended consequences emerge in a web of affordances and influence the actualization of both basic and more advanced affordances. The unintended consequences are summarized in table 5. The web of relations between affordances and unintended consequences is visualized in figure 3. Multiple theoretical and practical implications and contributions of this study are discussed. To conclude, opportunities for future research are provided.
References
Brown, S. L., & Eisenhardt, K. M. (1997). The Art of Continuous Change: Linking Complexity Theory and Time-Paced Evolution in Relentlessly Shifting Organizations. Administrative Science Quarterly
,
42(1), 1–34.
Bygstad, B., Munkvold, B. E., & Volkoff, O. (2016). Identifying generative mechanisms through
affordances: a framework for critical realist data analysis. Journal of Information Technology Impact
,
31(1), 83–96.
Corbin, J. M., & Strauss, A. (1990). Grounded theory research: Procedures, canons, and evaluative criteria. Qualitative Sociology
, 13 (1), 3–21.
Eisenhardt, K. M. (1989). Building Theories from Case Study Research. Academy of Management Review.
Academy of Management
, 14 (4), 532–550.
Gibson, J. J. (1986). The Ecological Approach to Visual Perception
. Psychology Press.
Jensen, T. B., & Aanestad, M. (2006). How Healthcare Professionals “Make Sense” of an Electronic Patient Record Adoption. Information Systems Management
, 24 (1), 29–42.
Leonardi, P. M. (2007). Activating the Informational Capabilities of Information Technology for Organizational Change. Organization Science
, 18 (5), 813–831.
Leonardi, P. M. (2011). When Flexible Routines Meet Flexible Technologies: Affordance, Constraint, and the Imbrication of Human and Material Agencies. Management of Information Systems Quarterly
,
35(1), 147–167.
Leonardi, P. M. (2013). When does technology use enable network change in organizations? A comparative study of feature use and shared affordances. MIS Quarterly
, 37 (3), 749–775.
Majchrzak, A., & Markus, L. (2013). Technology Affordances and Constraints Theory (of MIS). In Encyclopedia of Management Theory
(pp. 832–836). SAGE Publications, Inc.
Majchrzak, A., Markus, M. L., & Wareham, J. (2016). Designing for digital transformation: Lessons for information systems research from the study of ICT and societal challenges. The Mississippi Quarterly
, 40 (2), 267–277.
Markus, M. L., & Silver, M. S. (2008). A foundation for the study of IT effects: A new look at DeSanctis and Poole’s concepts of structural features and spirit. Journal of the Association for Information Systems
, 9 (10/11), 609.
Poissant, L., Pereira, J., Tamblyn, R., & Kawasumi, Y. (2005). The impact of electronic health records on time efficiency of physicians and nurses: a systematic review. Journal of the American Medical Informatics Association
, 12 (5), 505–516.
Slay, H. S., & Smith, D. A. (2010). Professional identity construction: Using narrative to understand the negotiation of professional and stigmatized cultural identities. Human Relations
, 64 (1), 85–107.
Strong, D. M., Johnson, S. A., Tulu, B., Trudel, J., Volkoff, O., Pelletier, L. R., … Garber, L. (2014). A theory of organization-EHR affordance actualization. Journal of the Association for Information Systems
, 15 (2), 53–85.
Tim, Y., Pan, S. L., Bahri, S., & Fauzi, A. (2018). Digitally enabled affordances for community-driven environmental movement in rural Malaysia. Information Systems Journal
, 28 (1), 48–75.
Urquhart, C., Lehmann, H., & Myers, M. D. (2009). Putting the “theory” back into grounded theory:
guidelines for grounded theory studies in information systems: Guidelines for grounded theory studies in information systems. Information Systems Journal
, 20 (4), 357–381.
van Aken, J., Berends, H., & van der Bij, H. (2012). Problem Solving in Organizations: A Methodological Handbook for Business and Management Students
. Cambridge University Press.
Volkoff, O., & Strong, D. (2013). Critical Realism and Affordances: Theorizing IT-associated Organizational Change Processes. MIS Quarterly
, 37 (3), 819–834.
Volkoff, O., & Strong, D. (2017). Affordance Theory and How to Use it in IS Research. In R. D. Galliers
& M. K. Stein (Eds.), The Routledge Companion to Management Information Systems
. Routledge.
Wieneke, A., Lehrer, C., & Jung, R. (2016). Exploring Affordances of Business Intelligence & Analytics with regard to Customer-Oriented Work Practices. In PACIS 2016 Proceedings
. Retrieved from
http://aisel.aisnet.org/pacis2016
Yin, R. K. (2009). Case study research: Design and methods
. Sage publications.
Zammuto, R. F., Griffith, T. L., Majchrzak, A., Dougherty, D. J., & Faraj, S. (2007). Information Technology and the Changing Fabric of Organization. Organization Science
, 18 (5), 749–762.
