This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.
Knowledge of how to design digital systems that are ergonomically sound, high in usability, and optimized for the user, context, and task has existed for some time. Despite this, there are still too many examples of new digital health care systems that are poorly designed and that could negatively affect both the work environment of health care staff and patient safety. This could be because of a gap between the theoretical knowledge of design and ergonomics and the practical implementation of this knowledge in procuring and developing digital health care systems. Furthermore, discussions of digitalization are often at a general level and risk neglecting the nature of direct interaction with the digital system. This is problematic since it is at this detailed level that work environment and patient safety issues materialize in practice. In this paper, we illustrate such issues with two scenarios concerned with contemporary electronic health care records, based on field studies in two health care settings. We argue that current methods and tools for designing and evaluating digital systems in health care must cater both to the holistic level and to the details of interaction and ergonomics. It must also be acknowledged that health care professionals are neither designers nor engineers, so expectations of them during the development of digital systems must be realistic. We suggest three paths toward a more sustainable digital work environment in health care: (1) better
There is no doubt that the digital transformation of health care is changing how health care is delivered [
When technologies are used, they always help to shape the context in which they fulfill their function. They help to shape human actions and perceptions, and create new practices and ways of living.
Researchers have studied the impact of digital systems on the work environment since the early eighties [
In parallel with the introduction of digital systems in the workplace, the field of human-computer interaction (HCI) has developed, and the usability (ie, the efficient, effective, and satisfactory use of a system with respect to the specified user, goal, and use context [
Despite this, basic usability-related problems continue to emerge in health care practice [
In this paper, we identify new ways to overcome obstacles related to poorly designed digital health care systems. Specifically, we highlight the potential effects on the work environment of violations of HCI design principles and of deficient knowledge of the user, use context, and task. This is illustrated through the analysis of two scenarios, providing insights into how these shortcomings materialize in health care practice. Based on insights from these scenarios, we argue for three paths forward that we believe could lead to a more ergonomically sound digitalization of health care.
In the following section, we first present two scenarios concerning the use of contemporary electronic health care records (EHRs). EHRs are used as a means to demonstrate our viewpoints, as they are used by basically everyone in the health care sector, whether you are in home care, in primary care, or in specialized care, and whether you are a doctor, a nurse, a lab assistant, or an administrator. They are, furthermore, frequently used throughout the day. We then outline four insights from the scenarios built on basic HCI theory, which illustrate how digital systems affect health care professionals’ work environment.
The two scenarios are synthesized from real situations witnessed in data from in vivo observations of Swedish home health care and primary care performed in 2014 and 2018 as part of two research projects. One project aimed at studying digital systems used by practitioners in home health care. The other one studied the implementation of a new EHR for primary care and included both an expert evaluation of the usability of the system as well as in vivo observations of the system at primary care centers.
On Monday morning, Anna, a 47-year-old registered home health care nurse, logs in to her computer and opens the EHR to get an update on what has happened with her patients since she last saw them on Thursday the week before. She starts by looking through the messages sent by the staff working during the weekend. There are 51 new messages in the message box. With the introduction of the new EHR system, the number of messages has multiplied due to a new way of sharing information in which messages are sent to work groups instead of individuals. Many of the messages are obsolete because they were already addressed during the weekend, while others are not relevant to Anna’s work. She understands that she must read 20%-25% of the messages more carefully and that the only way to find them is to go through the list one message at a time and delete the irrelevant ones.
Anna needs to check what has happened to one of her patients during the weekend, so she opens the health record for that patient. The information is sorted chronologically with the oldest post first, and since the patient has had home health care for 3 years, the health record is long. To find the most recent post from the past weekend, Anna must scroll through several pages of records. To avoid this time-consuming activity, which entails a lot of clicking, Anna has learned that she can access the information in reverse chronological order by creating a PDF file for printing. By using this workaround, she can access the information she wants about the patient more quickly.
In the afternoon, Anna returns to the office from a patient visit and opens the EHR to document the visit. This procedure is by far the one most often performed in the system. The process is cumbersome. Besides the actual writing, it takes Anna 19 clicks to find the patient, open the patient’s health record, locate the correct place to insert information, and then verify and save the information. To make things even more annoying, the system takes more than a second to respond to each of Anna’s clicks, with the consequence that she makes an extra click or two just to be sure. As part of the documentation process, Anna must also classify the intervention performed at the patient’s home. Each intervention has a name, for example, “Ostomy control,” and a code, for example, “EK007,” selected from a drop-down menu with a long list of all available interventions (
An illustration of a drop-down menu with interventions sorted by code and not by name. The user must scroll through the list to find the desired intervention.
General practitioner Helen is finishing her shift at the primary care center. Before she can go home, she has a long list of EHR posts to verify and sign. She has postponed this activity for too long since it is so cumbersome to perform. There are several ways this signing can be done, but she has not found a way that suits the workflow and that she can easily remember, so she is unsure whether she is doing it correctly. She needs to stay focused when jumping between the view in which she keeps track of which posts are still unsigned and the view in which the actual information to be signed off is presented. Furthermore, the feedback on each post is difficult to grasp. A post with a certain hashtag symbol has been signed and needs no further attention. Two smaller, overlapping hashtag symbols indicate a post that was signed but has been changed and thus must be viewed and signed again. A post that is unsigned has no symbol at all and is thus not highlighted in any way (
Helen is interrupted in her work by a medical laboratory assistant. He wants Helen to change the lab test order she placed for a patient. The lab test that Helen selected from the list in the EHR system cannot be performed at this primary care center. It is not that she did not know this, but that the list from which lab tests are chosen is long and the names are often similar, so she sometimes clicks on the wrong one. Helen is not the only physician making this mistake, and to prevent this from happening, the medical laboratory has printed a separate list on paper showing only the available lab tests. Helen is annoyed that she has to interrupt her work to change the lab test order. She opens the list of lab tests on the computer, verifies that she has now chosen the correct one by double-checking against the printed list, and then finishes the signing before going home 25 minutes later than planned.
An illustration of how one view of the health record for one patient can look in the system. There is one input line per date, and all inputs here are either notes or information about the patient contact made. The “Initials” column shows who made the input, and the “Sign.” column shows the signing status: an empty space indicates that all signing has been completed; a single hashtag indicates that signing needs to be done; the overlapping hashtags (one or two) indicate that a post was signed but then changed and needs to be looked at and signed again.
A different electronic health care record layout. This view presents information organized differently from the view shown in Figure 2; for example, the signing icons now appear in the center of the table and not to the right.
The scenarios have been synthesized from situations witnessed in health care practice to illustrate what working in a state-of-the-art EHR can be like. Several insights (
Insight 1: The digital system forces the user into inefficient workflows
Insight 2: HCI design principles—independent of user, task, and context—are not adopted
Insight 3: Workarounds become permanent solutions
Insight 4: Different domains of ergonomics interact in creating the digital work environment
How different functions in the digital system are designed forces the user to adapt how tasks are performed. When the new EHR was introduced in the home health care organization (scenario 1), the communication routine changed. The design of the new system increased the number of messages that the user had to process each day. Similarly, the general practitioner at the primary care center (scenario 2) postponed the signing task since the system hindered rather than supported her in this process by not providing an intuitive workflow or a consistent design. This resulted in an inefficient procedure, causing both frustration and physical strain due to excessive clicking. Demands for more extensive signing in the primary care EHR (scenario 2) were deliberately introduced as a new work routine when implementing the new EHR in order to increase patient safety. Hence, the extra effort required to perform this task may be justified. However, how it was implemented in the system, with a confusing workflow, poor feedback, and inconsistency, could have been avoided by applying better knowledge of how the user works. In contrast, the increased message handling in home health care (scenario 1) was not deliberately introduced, and its effect on the time and effort needed to process the messages was not acknowledged or understood until the system came into practical use.
A user-centered process is always desired, but even without such a process there are
In scenario 1, the nurse must fake a printing procedure to create a PDF that presents the patient’s health record in the desired chronological order. This solution works around the current default, which presents the information in reverse chronological order, not in line with how the nurse reads the information. It is not technically difficult to implement this correction, so one wonders why it was implemented incorrectly from the beginning. This could have happened because of a lack of understanding of the user’s tasks, revealing that a user-centered processes did not guide the system development. Since redesigning the software entails additional costs, a so-called workaround solution—creating a PDF—is offered. Likewise, the medical laboratory assistant in scenario 2 had created a workaround solution for handling the long list of lab tests that caused many incorrect orders. This solution involved a printed paper with a shorter version of the list, including only the available lab tests. These examples of workaround solutions are indicative of faulty system implementation and should lead to changes in the digital system. Instead, it is the users who must adapt, and the risk is that such workarounds can become a standard part of the interaction [
Although HCI is traditionally associated mainly with the domain of cognitive ergonomics, we would like to emphasize that the development, implementation, and usage of a digital system entails interaction between several ergonomic domains: physical, cognitive, and organizational ergonomics [
In the scenarios, we see examples of the users having to traverse long lists of information, interpret vague feedback, wait for the system to load, and be interrupted in tasks, all of which are related to cognitive ergonomics. Time-consuming or cumbersome interactions are frustrating and tiring to deal with and increase the cognitive load [
Physical ergonomics is also part of the interaction. A design that involves small buttons, many clicks, and an inefficient layout, as in the two scenarios, can cause strain, mainly in the neck and shoulders [
How tasks are performed in the EHR system is closely connected to, for example, how work is performed and how staff interact and communicate. We saw examples of this in scenario 1, in how communication was performed through messaging in home health care, and in scenario 2, when the physician at the primary care center had to allot extra time for administration in order to handle the new way of signing posts in the system. Hence, organizational aspects of ergonomics are also part of the digital work environment [
What the four above insights share is that none of them is really novel: they all concern issues that have long been known, and in a sense solved, by HCI and ergonomics researchers. The insights were extracted from observations of real use situations involving modern EHRs and typical everyday work by health care professionals, meaning that the related issues are still very real for the typical users of such systems. Furthermore, several recent scientific publications support the insights noted in the scenarios, and we conclude that usability issues are still present, workarounds are still common, and there are no indications that these issues are decreasing in prevalence [
Concerning the first insight about introducing
The second insight about violating
The third insight about
The fourth and last insight concerns the interaction between
As the ongoing digitalization introduces new ways of working, including new hardware (eg, mobile documentation on tablets or wearable sensors) as part of the work equipment, interaction between different ergonomic domains can be expected to become more complex. A few relevant studies have already emerged. Johnson et al [
Based on the above discussion, we identify three paths toward a more sustainable digital work environment. The first path focuses on providing HCI theory–based
Studies exploring system usability are typically based on the users’ subjective grading of various usability statements related to the system (eg, [
When procuring a new digital system, such as an EHR, requirements related to HCI and ergonomics are often either completely lacking or expressed in unverifiable terms [
A lack of adequate
By creating guidelines for procurement that include user involvement, together with
As mentioned above, ergonomics has many aspects, and frameworks for combining these aspects have been proposed to allow for the simultaneous achievement of individual employee and organizational goals [
To meet this need, we propose the introduction of
This definition is based on the International Ergonomics Association’s [
Discussion of the digitalization of work is often general and insufficiently detailed to capture the full complexity of digitalization, which includes everything from effects on efficiency, safety, ethics, power relations, the work environment, and new ways of organizing work to designing the technology itself. Much research into digitalization emphasizes how health care is being transformed as a practice. As demonstrated in the above scenarios, many problems associated with digitalization only appear in actual use situations, where they also become issues for individual workers. Furthermore, many of these problems are completely unnecessary, since they are not novel and knowledge of how to avoid them has long existed.
In order to bridge this gap, initiatives are needed to more actively integrate ergonomics and HCI theory into practice. We suggest three paths forward so as not to overlook fundamental aspects of implementing and using digital systems in order to create a sustainable digital work environment: (1) development of
electronic health care record
human-computer interaction
None declared.