Abstract
This study describes the concept of affordance-based procedure and its implementation in a triage station in a hospital emergency department. Rather than seeking to increase operators’ adherence to procedures, an affordance-based procedure (1) aims to induce task steps using affordances that also (2) support degrees of freedom for action. The design of this procedure was guided by the application of an extended version of cognitive work analysis, named “heuristic cognitive work analysis.” This design process produced a new procedural document: a reception card. Ten months after its implementation, a qualitative evaluation with 10 triage nurses shows that the reception card is viewed as supporting coordination between the different nurses’ tasks and providing an external memory to cope with frequent interruptions during high patient inflow, even though the document is used for convenience and with unexpected and partial uses of its items. The document assessed also afforded emerging benefits, that is, acceleration of ambulance release, higher level of confidentiality, assistance for staff hand-overs. Finally, novice triage nurses are particularly sensitive to the benefits brought by this affordance-based procedure.
Introduction
The role of task procedures in a work system remains an important issue in Human Factors studies. Even though researchers have warned of the limitations brought by a normative approach to procedures, the latter approach is still largely dominant in organizations to manage operators’ performance (Rebbitt & Erickson, 2016). Technical documents (e.g., guidelines, flow charts, checklists, protocols, or schedules) are commonly used as concrete means to request operators to comply with task procedures (Hale & Borys, 2013; Seferdjeli & Terraneo, 2015). In line with Frederik Taylor’s studies, detailed procedures are commonly viewed as the best way to cope with both routine actions and unusual work situations (Locke, 1982). Yet, numerous studies have shown that “work-as-done” differs from “work-as-imagined” and as written in the procedures (Braithwaite et al., 2015; Carim et al., 2016; Hollnagel, 2014; Leplat & Hoc, 1983).
Consequently, some researchers have highlighted the need for an alternative approach to task procedures (Blackstad et al., 2010; Dekker, 2005; Grote et al., 2009; Wachs & Saurin, 2018). In particular, the CWA framework aims at listing all the possible functions involved in the work domain in order to offer an opened window on the potential affordances relevant to control a situation (Vicente (1999). Operators are then assumed to be able to fully understand the state of the work domain and to select a strategy efficiently within a range of possible responses, notably in degraded work situations. Rather than prescribing a set of specific task requirements, the core idea is to make visible the degrees of freedom available to operators to control the work domain (see the notion of Ecological Interface Design, Vicente & Rasmussen, 1992).
The CWA framework has shown its effectiveness in designing assisting devices for tightly coupled work systems, in which causal relationships between task operations and events in the work domain are rather unambiguous and can be anticipated. In such work systems (e.g., nuclear plant, aviation, anesthesiology), individual supervision tasks to monitor the proper functioning of these relationships are essential.
The purpose of this study was to use an extended version of cognitive work analysis (CWA), named “heuristic cognitive work analysis” (hCWA), to design assistance devices, specifically a new kind of task procedures based on affordances (Bernard et al., 2020; Morineau & Flach, 2019). This process is illustrated through an application of hCWA to the work of triage nurses at the entrance of an emergency department.
The hCWA has recently been proposed to apply the CWA approach to loosely coupled work systems. These systems are characterized by the multiple degrees of freedom available to determine the course of operators’ activity. Work situations frequently vary, and emerging events may require ad hoc adaptation and self-organizing processes. In loosely coupled systems, operations mainly involve concrete transformations of the situation in the context of teamwork (Albrechtsen et al., 2001). Thus, the situation evolves in keeping with the emerging specificities and the dynamics of situated interactions between operators, tools, and the work domain (e.g., medical emergency, rescue services; see also the recent notion of self-organizing holarchic open systems in Miller et al., 2019).
Whereas CWA describes activity as a mental navigation through the means-ends relationships within the work domain, in hCWA, activity is depicted as a trajectory in a workspace composed of many degrees of freedom and bounded by basic work constraints (Rasmussen, 2003, Rasmussen, 1997). The current issue is to know whether the hCWA framework can be used to design assisting devices for loosely coupled work systems. In this paper, we assess the feasibility of using hCWA to design procedures based on affordances (Gibson, 1979; Woods & Hollnagel, 2006), that is, on concrete resources for action (Suchman, 1987, 2007; Wright et al., 1996) rather than on a strict compliance to task sequences.
After introducing the hCWA method, we present its application to a hospital triage station and discuss the findings.
Heuristic Cognitive Work Analysis to Design Affordance-Based Procedures
The hCWA framework is based on the CWA stages of analysis that examine the different sets and subsets of work constraints that shape the operators’ behaviors. The process starts with the analysis of the work domain constraints, followed by the elicitation of the constraints arising respectively from control tasks, operators’ strategies, work organization, and operators’ competencies. In the heuristic version of CWA, the work domain analysis is replaced with a representation of the workspace according to the dynamic safety model (Rasmussen, 2003). The task analysis stage involves a representation of tasks through the State Space Sketch, and the subsequent analysis stages are viewed as outputs observed during the activity analysis (Figure 1).
Stages of the heuristic cognitive work analysis (hCWA) framework. In the workspace, three examples of drift: tunnel effect with a focus on the patient (trajectory A), difficulties with equipment (
In hCWA, “heuristic” means that the work analysis does not first search for an exhaustive modeling of the work domain functions as proposed by the CWA. The purpose of hCWA is rather to highlight the basic constraints that significantly shape the work system and that require adaptive behaviors in order to avoid critical drifts.
Stage 1: Functional Workspace Analysis
Stage 1 involves considering the basic constraints emanating from the work domain and the work organization that delineate the set of opportunities or options available to operators, that is, their degrees of freedom during decision-making and effective actions. These degrees of freedom are constrained by dynamic organizational/environmental forces that can increase or decrease opportunities. In this space, the operators’ activity is represented as an operating point with a trajectory. A migration, or drift, of the operating point toward a risk margin may lead to adverse events (Rasmussen, 2003). This process can be illustrated through scenarios occurring in acute care in an emergency room (Bernard et al., 2020). The workspace of acute care can be delineated by the three following constraining boundaries: patient care, task management (a term used in studies on nontechnical skills to describe operations involved in the handling of equipment and in task distribution—Morineau et al., 2016), and information processing (inner and outer processing through mental activity, inter-individual communication, and the use of cognitive artifacts—Hutchins, 1995). A first scenario could consist in a “tunnel effect” during which the medical team is focused only on diagnosis and technical gestures without preparing medical equipment and neglecting team communication. In this case, the team would migrate toward the “patient care” boundary (Figure 1, trajectory A in the workspace). A second scenario could involve difficulties in the manipulation of medical equipment that lead caregivers to neglect the patient’s state and communication between them (trajectory B). A third scenario could relate to a disagreement between two physicians that leads to a stormy discussion, while the patient’s state is deteriorating, and a plan to prepare equipment is not triggered (trajectory C).
Stage 2: Task Analysis
In hCWA, task analysis involves describing the essential and critical task steps that the operating point’s trajectory must go through in order to remain inside the safe workspace envelope (Bernard et al., 2020). This description yields a State Space Sketch in which states are mandatory task steps to achieve. In complement, logical dependency relationships possibly existing between states are highlighted. A dependency link describes a given state as a precondition that is mandatory to move from one state to another. Mandatory states and dependency relationships reduce the number of degrees of freedom available in the workspace. For instance, being in a room is a state that requires as a precondition that an opening is accessible (a door or in exceptional conditions, a window). States that are not represented in the State Space Sketch may be either accessory or can emerge from the situation.
The Next hCWA Stages as Outputs: Strategies, Work Organization, Competencies
In hCWA, strategies, work organization, and human competencies are viewed as adaptive processes. They contribute to keeping the work trajectory in the safe envelope, hence satisfying the combination of the workspace constraints and those involved in the task state space (critical states and dependency relationships). Artifacts, as objects or procedures, may also be considered as adaptive responses implemented in material resources.
The Design of Affordance-Based Procedures
Classically, a technical document is viewed as the concrete implementation of an abstract procedure requiring compliance. However, this top-down approach may be replaced with a bottom-up approach, whereby a technical document is viewed as an affordance in situated action. In the ambient environment, affordances are environmental properties that furnish opportunities for action, supporting the individuals’ behavior according to their capacity and intentions (Gibson, 1979). Affordances exist in the environment as properties that can be potentially used, but will not necessarily be.
An affordance-based procedure is defined as a required set of steps in a task procedure the completion of which is induced by easily accessible resources for action, while many degrees of freedom are still available.
Table 1 synthetizes a comparison between procedures involving compliance and those driven by affordances. The table shows that an affordance-based procedure fits well with the hCWA idea that degrees of freedom must be preserved in a loosely coupled work system.
The next section presents the application of hCWA to the triage station in a hospital emergency department. This analysis enables the design of a reception card to assist nurse triage, a technical document implementing the notion of affordance-based procedure.
Comparison Between a Procedure Involving Compliance and a Procedure Based on Affordances.
A Heuristic Cognitive Work Analysis of a Triage Station in a Hospital Emergency Department
Triage at the entrance of an emergency department (ED) sorts and prioritizes patients for medical care. As the frontline of the patient pathway, triage represents a critical stage. An incorrect triage decision or symptoms that are missed can delay care and increase the risk of worsening or deterioration of the patient’s condition. Typically, triage is performed by a nurse trained in triage protocols, who is sometimes combined with a physician’s assessment. They conduct a preliminary screening to determine how long the patient has the capacity to wait before receiving a more thorough clinical examination and eventual treatment (Yang et al., 2016).
As in any workflow, triage involves a basic task sequence including different phases (e.g., patient arrival, registration, triage). The order of the elements of this task procedure is not fixed and may vary among EDs as a function of different factors: the presence of a specific registration station or of a common area shared by the nurse and registration teams, the rapid expedition of patients to treatment areas while registration is performed at the bed side, and so on (Huddy, 2017). In this study, we examined the issues posed by a given triage/registration procedure in the context of an overcrowded ED.
Method
Context
The Atlantic Brittany Hospital Center is a French “county” hospital in a town of 53,000 inhabitants. It is a popular vacation region that also attracts retirement-age people with gradually increasing healthcare needs. This situation leads to a high number of visits in the ED, roughly 70,000 per year, with peaks of about 200 visits per day during vacation periods.
Setting
Figure 2 shows the architectural configuration of the ED entrance. During high inflow, two nurses and a physician are in charge of triage, a monitoring nurse is in charge of reassessing the condition of waiting patients, a nursing aide and two secretaries also occupy this area. In periods of low inflow, there may be only one triage nurse, with one nursing aide and one secretary, as for example during the night.
Simplified map of the emergency department entrance.
Procedure
Table 2 presents the different steps of the work analysis and main data sources. Empirical observations and in situ interviews with the clinicians were iteratively performed. Prior investigations of hCWA in the context of a medical emergency were used to specify the foundations of the workspace model. Observations supplemented by quantitative data on task interruptions and general statistics on ED visits were used to specify the constraining pressures exerted on the triage nurses’ workspace. Interviews with clinicians helped identify the critical main task states during registration/triage. Then, a logical analysis based on the aggregated data brought to light the dependency relationships between task states. The following data collection documented the strategies, work organization, and required competencies deployed by clinicians. These data will serve to design and evaluate a solution.
Application of hCWA and Main Data Sources Used for Each Step.
Note. The hCWA stages are given in boldface. hCWA = heuristic cognitive work analysis.
Data collection
Observations were performed over 2 weeks, that is to say about 50 hr, in the first part of August 2019, a period known for overcrowding. The same observer, specialist in Human Factors studies, was located in the triage zone over the period. The research observer was oriented to the triage station before observations began. The data were collected through pen and paper notes, photos, a stopwatch, and in situ informal interviews with the team members. Task interruptions were counted during a specific moment of nurses’ activity: recording the patients’ medical file. Statistical data on the number of visits automatically recorded by the ED software were also used. Overall data collection ended when understanding the work situation reached saturation level, after debriefing with the ED working group on the collected data.
Ethical considerations
A research convention signed between the Hospital and the University in charge of the study ensured confidentiality about the information collected from both the hospital personnel and the patients. The personnel were told the study was an ergonomic study on reception and triage. The collected data were anonymized. The data were not shared by the research group with management or anyone else outside the group.
Findings From the Heuristic Cognitive Work Analysis
This section shows the key findings related to the work analysis.
Stage 1. Functional Workspace of the Triage Station
Generic constraints selection
Figure 3 shows the triage station as keeping with three basic constraints previously observed in studies on other medical emergency situations (Morineau & Flach, 2019), that is, Patient Care (screening, triage), Task Management (task distribution, ambulance management, equipment handling), and Information Processing (mental workload, management of interruptions).
Functional workspace of triage with its basic constraints exerting pressure on the triage nurse’s activity and leading to migration into the risk margin.
Specification of triage workspace constraints
Data collected on these three basic constraints allowed to assess their pressures on the work system.
Patient care
The statistics collected in the ED under investigation about patients’ visits during the observation period, in particular during eight overcrowded days from August 7 to 14, showed an average of 227.1 visits per 24 hr (SD = 18.9) with a minimum of 206 visits and a maximum of 252 visits per day. In such situations of overcrowding, patient care is subjected to contradictory and competing pressures of the need for rapid new patient intake versus the need to have sufficient time for a rigorous clinical examination for sense making.
Information processing
Task interruptions during the completion of a medical record constitute a cue revealing the pressures originating from information processing. Interruptions produce mental workload and interferences in short-term memory. For two mornings and three afternoons from August 7 to 14, interruptions during this task were recorded. A total of 34 triage medical file completion tasks performed by five triage nurses were observed, which corresponds to a cumulated time of 190 min. During these periods, 249 interruptions were noted, that is to say one interruption every .9 min (SD = .5), that is, an average of 1.4 interruptions per min (SD = .7, n = 34 observations). In comparison, Johnson et al. (2014) observed a mean of seven interruptions per hr at the triage station in another hospital.
Figure 4 shows the sources of the interruptions observed during triage medical file completion. The main source of interruptions is coworkers in the triage/registration station (55%).
Distribution of interruptions during triage medical file recording by the triage nurse.
Task management
Observations show that the frequent use of equipment constrains nurses’ activity; examples include printing the triage medical file, handling the mobile station on which a computer laptop is standing, or installing the medical equipment during a clinical examination with the help of the nursing aide. The digital documents are shared between the secretary and the triage nurse through communal software, and their paper format equivalent is also shared through a communal patient’s folder. This task distribution imposes document processing synchronization between the two operators, which represents a supplementary constraint. The triage nurse must also implicitly consider the need to release the ambulances arriving in the ED as soon as possible.
The combination of all these pressures increases the risk of a drift of the operating trajectory toward the risk margin, as shown in Figure 3. For instance, whilst the triage nurse is focused on the registration of a medical record after a clinical examination (patient care), she is interrupted by a paramedic who is waiting to deliver a patient (task management). This interruption interferes with the short-term storing of information of the patient examination that she has performed before and means that she has to collect this information again (information processing).
Stage 2. Triage Task Analysis With State Space Sketch
Figure 5 shows the registration/triage workflow. The first state is the reception that normally is carried out by the triage nurse in order to detect a deteriorating patient as soon as possible. The output of this reception stage (arrows) shows that the reception is a precondition to several subsequent states: administrative registration by the secretary, clinical examination by the triage nurse, and the generation of a triage medical record. At this level, the main issue is that in overcrowded conditions, there may be no triage nurses available to receive a new patient.
State space sketch of triage (white circles) and administration registration tasks (blue circles). Arrows between states represent dependency relationships.
Another specific state is the medical electronic recording that requires several preconditioned states to be triggered. In addition to the preliminary reception, the triage nurse needs to have the patient’s digital file created by the secretary after registration and its equivalent paper version available to read complementary documents emanating, for instance, from the patient’s general practitioner (GP). Overall, this task analysis demonstrates that the triage nurse must be both at the front line of the ED entrance and capable of taking over processing the patient’s file after the administrative registration.
Stage 3: Strategies, Work Organization, Competencies of the Triage Station
To cope with these constraints, the triage nurses deploy strategies, ad hoc work organization, and cognitive competencies.
At the organizational level, sometimes the nursing aide, or even the secretary, carries out the reception of walk-in patients, whilst this function is not included in their prescribed roles. When patient reception is performed by the current available team member, information is transmitted orally to the triage nurse, which increases the likelihood of interruptions
Health managers in the ED explained that several months before our study, this operating mode of task distribution among triage/registration team members led the staff to impose a specific task procedure, whereby there were two triage nurses at the station: one nurse would be in charge of reception whereas the other one would deal with creating the patient’s medical file. The triage nurses reported that this system was unsatisfactory, and it was consequently abandoned. At the time, the deeper reasons explaining this organizational failure were not explored.
Alternatively, if the same nurse performs both the reception and the medical registration stages, loss of information is likely to occur. To avoid these situations, the strategy developed by the nurses was to write some critical information informally on the paper folder of the patient’s file or to rely on their own short-term memory. As pieces of information are written down informally, more often than not, they can be understood only by the nurse who has written them. Writing is thus viewed as an aid for inner cognitive processes rather than as cognitive support for teamwork.
In conclusion of this work analysis, in addition to the pressures emanating from the work context of intense activity, the triage nurse is submitted to conflicting task goals between being present at both the patient’s reception and the triage medical file recording stages. Since allocating these two tasks to two triage nurses, when possible, did not work, a new solution was needed.
Design and Evaluation of an Affordance-Based Procedure
The design and evaluation of an affordance-based procedure are presented. Concretely, this procedure took the form of a reception card. This document was designed to enable the triage nurse to collect the patients’ clinical information during the first contact with patients and their family. A participatory design method was used with a working group that iteratively refined the first reception card design until reaching a satisfying version. This version was implemented in the hospital setting and evaluated.
Method
Setting
A working group already existing in the ED and having the purpose to address potential organizational issues participated in the design of the device. This group was composed of three physicians, two healthcare executives, one expert triage nurse, one nursing aide, and one ED secretary. Four meetings lasting about 3 hr each took place between September and November 2019 to implement an iterative design loop. During this period, the working group continuously informed the ED personnel on the design process. Two nursing science teachers practicing in the Nursing School of the hospital and who had previously worked as triage nurses gave advice on the different versions of the document by email. Before and after the design process, the project and its outcomes were validated by two general staff meetings that included the Head of the Hospital. In December 2019, the hospital’s senior administration gave the green light to the triage reception card implementation.
Procedure of evaluation
In September 2020, 10 months after its implementation, the document was evaluated via a questionnaire sent by email to the triage nurses of the ED. Four questions were formulated: 1/Do you use the reception card during your activity? Yes, systematically – Sometimes – Never. 2/For what reasons? 3/If tomorrow the Emergency Department staff decided to return to the previous organization (with no reception card) would you…? Strongly disagree, Disagree, Somewhat agree, Agree, Strongly agree. 4/For what reasons?
Data collection
Ten triage nurses out of a total of 45 nurses certified to work as triage nurses at the entrance of the ED completed the evaluation questionnaire. They had an average age of 33.5 years (SD = 9.3; min. = 24 years; max. = 48 years). Their experience as triage nurses varied from 8 months to 17 years. With such a small sample, only a qualitative thematic analysis of reasons justifying the responses was carried out. Since the two questions concerning the reasons provided similar types of findings, their responses were aggregated. The topics extracted from the thematic analysis integrated all the participants’ responses.
Ethical considerations
The evaluation phase was performed in the same ethical context than the design phase. The questionnaires were anonymous. Participants had been previously informed of the study.
Results
Features of the designed reception card
An affordance-based procedure (Figure 6) can be characterized as a set of functions, physical forms, information content, and procedural instructions.
Example of triage reception card from in situ work (Copyright Atlantic Brittany Hospital Center).
Functions
We formulated a set of design goals on the functions of the reception card. Our main hypothesis was that the reception card would induce the work organization that had previously failed and involved putting one nurse at reception whereas the other would focus on the electronic medical record and the clinical examination. Second, we expected that the reception card, acting as an external memory, would reduce the risk of information loss caused by interruptions. These functions would be particularly useful during high patient inflow.
Physical format
The paper format facilitates rapid use without any of the constraints related to the use of a digital device (task management constraint). The paper format also enables annotations decided in the course of the action (information processing). For easy handling, the document is a paper notepad of A5 size that can be put it in the pocket of the nurse’s gown. Each sheet can be easily detached from the notepad.
Information content
The purpose of the procedure is to structure the main information items that must be requested of a new patient arriving at the ED (patient care constraint). The items composing this document are listed in Table 3.
Items Composing the Reception Card.
These items concern formal information about the patient, but also record the first clinical data collected on symptoms and primary care. Information about the patient’s living situation and autonomy are particularly important for older people.
Procedural instructions
The reception card was presented with the following instructions: (1) The document may be used at your convenience for patient reception, if necessary. (2) If it is used, only the items you consider important may be completed.
Figure 6 shows an example of a reception card used by the triage nurses. As an affordance-based procedure, a reception card is expected to be used as a resource for cognitive processing rather than a document requiring by-the-book application. The document shown in Figure 7 is partially used; not all the items are completed. An unexpected use can also be observed through the annotation of physiological data (e.g., temperature, cardiac frequency, blood pressure) and the evolution of these data through several measurements. The presence of a different handwriting means that this reception card has been shared between two clinicians and supported the coordination of their interventions.
Reasons arguing in favor (in blue) or in disfavor (in orange) of the reception card extracted from a thematic analysis, with their frequencies of occurrence (participants = 10).
Findings Concerning the Evaluation of the Affordance-Based Procedure (Reception Card)
The data indicate that eight nurses out of ten used the reception card sometimes only. One nurse, the less experienced one, used the card continuously, at each patient’s intake. The most experienced nurse in the sample never used the document.
Figure 8 shows that seven out of ten nurses indicated that they would like to retain the reception card, and three either did not respond or did not find the card useful. Qualitative feedback about the lack of usefulness indicated that the card was “time consuming to complete and duplicated information capture.” In this case, only the “reason for ED visit” was viewed as useful. Another drawback concerns “the lack of space on the mobile station near the computer laptop that constrains the use the paper sheet,” despite the small size of the reception card (Figure 7).
Responses to the question “If tomorrow the ED staff decided to return to the previous organization (with no reception card) would you….?” (n = 10).
As hypothesized, the reception card was seen as “helping to cope with patient inflow” (cited seven times). Five times the document was judged to be “decreasing the risk of information loss,” and associated benefits are “error avoidance” and “protection against interruptions.” Two nurses mentioned the benefit for “nurse coordination between the reception and the triage medical file recording stages” and of “facilitating the waiting for administrative registration.” The document “reduces oral redundancy” in team coordination. A set of unexpected contributions also appear: the possibility of “accelerating the release of ambulances” owing to the rapid notification of the patient information, “higher confidentiality through avoiding oral communication,” “time saving,” and “hand-over facilitation” as another aspect of clinician coordination.
A novice participant illustrates how the document is integrated into her practice:
It is practical. It enables me to note the information from the ambulance drivers, families, or escorts when I am not processing the file directly. It also enables triage nurses to process the file without having carried out the reception themselves, and during hand-over. I’ve never known the previous system, I’m not sure I would be able to remember everything. I find that the oral system carries a greater risk of errors in records, identity, etc. When some colleagues don’t use it and I retrieve a file from which they have taken the reception information, they are obliged to tell me again orally, it’s a waste of time, according to me.
Discussion
This preliminary study provides some support for a reception card that aids the coordination between the triage nurses’ intake reception task and their other competing tasks. It also provides an external memory to cope with frequent interruptions. The document appears to be used for convenience, and with unexpected and partial uses of its items. As an affordance, the document is perceived by users to provide emergent benefits such as the acceleration of ambulance release, a higher level of confidentiality, and assistance for staff hand-overs. Whereas the most expert nurse in the sample does not use the document consistently, the less expert nurse uses it continuously. The latter finding supports the sensitivity of novices to verbal mnemonics aids, whereas experts likely prefer physical cues embedded in the environment (Beach, 1993). In the design process, hCWA was used to identify the main workspace and task demands that constrain the work system. This analysis has highlighted the basic functional purposes that the affordance-based procedure needs to meet to deal with these demands.
Classically, triage in an ED is viewed as a logistics matter of patient prioritizing designed to evaluate the performance of several triage scales (e.g., Hinson et al., 2019). However, this approach bypasses the characteristics of the actual work context in which these triage systems are implemented.
To solve work issues at the triage station, it is tempting to use procedures requiring compliance. In the case of the ED investigated, such a procedure did not work. Alternatively, an affordance-based procedure is at an intermediate level between procedures to be complied with and “free” affordances provided either by well-designed objects (Norman, 1999; Wears & Perry, 2002) or by self-designed resources for action, as for instance unofficial paper documents created and used by clinicians (Balka, 2010). The notes written on the patients’ paper folder by triage nurses observed at the triage station belong to this type of resources for action.
Previous healthcare design studies in the literature have already implicitly integrated this affordance-based procedure approach. For instance, Bakdash and Drews (2012) developed a prototype medical kit composed of pockets with icons, colors, and brief text descriptions. These pockets are adjusted according to the required sequence of equipment use (Drews, 2013). A second example is the Broselow tape that is dedicated to pediatric emergency. The tape is divided into color-coded segments that give an estimate of a child’s weight based on the child’s length, but also correspond to medical tasks in terms of doses of some emergency drugs, required equipment, and IV fluid volumes (DeBoer et al., 2005). However, before the current proposal and to our knowledge, there was no general methodological framework modeling the design of procedures based on affordances.
As a first application of hCWA in a design process, limitations can be pointed out due to the lack of quantitative data on triage performance that could confirm the positive effects of this intervention. Moreover, the sample of qualitative feedback is small. However, the reception card is currently fully integrated in the work organization at the entrance of the ED. Some anecdotes have revealed the relevance of this document. First, during a triage training session with nurses from another hospital, the trainees had the opportunity of using the reception card during simulated scenarios. At the end of the training session, they spontaneously asked if they could possibly obtain the document for their own triage activity. Second, the reception card has shown its relevance in a defective work environment context. Also, during the spring of 2020, a computer shut-down lasting 2 hr was fruitfully navigated by the use of the hCWA designed triage reception card as a full document for triage; another unexpected form of affordance thus emerged from this resource.
Conclusion
This study shows that hCWA can serve as a framework to design effective cognitive device interventions for loosely coupled work systems in which adaptive strategies are deployed to cope with multiple degrees of freedom and variety. Future studies are needed to confirm the relevance of hCWA to design task procedures that are based on affordances rather than on demanding compliance from operators or on “free” affordances offered by object available at the workstation.
Footnotes
Acknowledgment
The authors wish to thank all the members of the triage station who participated in this study. Special thanks to Yasmina Galateau, triage nurse; Régis Lorguilloux and Pierre-Yves Renahy, teachers at the nursing school of Vannes; and Sophie Oberlé and Agathe Texier-Pauton, healthcare managers at the Emergency Department for their helpful comments. The authors are also grateful to the editor and the experts who reviewed this manuscript for their useful comments and pertinent advices. This work was partly supported by the French National Research Agency (ANR) as part of the VR.MARS project.
Author Biographies
Thierry Morineau is a full professor in cognitive psychology and ergonomics at the University of Bretagne Sud, France. He is a member of the Human Factors & Ergonomics Society. He received his PhD in psychology at the University of Angers. Currently, his research focus is to adapt the cognitive work analysis framework for healthcare systems.
Mounia Djenidi-Delfour is a hospital practitioner and assistant manager in the emergency department of the Atlantic Brittany Hospital Centre. She received her Doctorate in Medicine at the Medical Faculty of Rennes University. She is currently involved in the improvement of safety in patient management procedures.
Fabrice Arnault is an emergency doctor and currently the Department Manager at the Emergency Department of the Atlantic Brittany Hospital Centre. In addition to his medical doctorate, he has received degrees in disaster medicine and medical management.