What is the Learning Unit about?
“There is nothing so practical as a good theory” is a famous quotation attributed to the psychologist Kurt Lewin. Following this idea, we initiated the D-CREDO series of learning units with a gentle introduction to Distributed Cognition Theory. Its tenets help explain how digital health tools fit into the clinical reasoning process, and in particular, how they interact with clinicians working in teams, as well as how they are influenced by the context of use (Fig. 1).
The goal of this learning unit was also generalistic: to present participants with the different classes of tools used in the D-CREDO learning units before they begin exploring specific topics related to individual tool categories. As this is a starting learning unit and a prerequisite for many other modules, we assume that participants may be early-year medical or nursing students who know the basics of clinical reasoning and are familiar with common cognitive biases affecting clinicians.
The practical exercises and discussions planned in the unit are intended to help students understand the potential benefits of digital technologies in clinical reasoning, articulate their strengths and limitations, and describe the basic ethical and legal aspects of using digital technologies in the clinical reasoning process.
Two virtual patients accompany students in this module: Karol Krawiec, an elderly patient from a remote mountain village on the Polish-Ukrainian border who is diagnosed with atrial fibrillation, and Franz Huber, a former car factory worker from Rotterdam who presents with chronic fatigue and is eventually found to have type 2 diabetes mellitus. Through these cases, students recognise both the opportunities and the risks associated with hand-held six-lead ECG devices (Fig. 2), medical calculators on-line, videoconferencing, patient portals, electronic health records with integrated warnings and reminders, and patient-facing diabetes-management applications integrated with digital glucometers.
Maintaining a bird’s-eye perspective, students are encouraged to discuss with their peers their intuitive perceptions and lived experiences of using these tools to date and their positive and negative impact on clinical reasoning.
How We Developed It!
The first step is always a well-designed plan. In D-CREDO, we formalised this process by requiring all working teams to complete a structured learning unit outline. This spreadsheet prompted us to deconstruct the high-level description of the learning unit from the blueprint into three phases: asynchronous, synchronous, and asynchronous again. Each phase was then divided into individual learning activities with estimated time allocations.
All activities were mapped to the ICAP (Interactive, Constructive, Active, Passive) framework, which served as a continuous reminder to design student-centred, interactive, and constructive tasks, while avoiding passive learning.
Although we always look for opportunities to reuse existing learning resources, an extensive internet search confirmed that no available material provided a sufficiently concise, medically oriented introduction to distributed cognition theory at a level appropriate for novice health professions students. Likewise, we identified the need for a dedicated video explaining the typology of digital tools in clinical reasoning that we apply throughout the learning unit. For both purposes, we first developed screenplays, discussed them collaboratively, and then converted them into voice-over presentations. Generative AI was particularly helpful for producing cartoon-like and photo-realistic images that effectively illustrated the content.
A key pedagogical intention was to integrate digital tools directly into the virtual patient cases so that students could interact with them in authentic ways. For the case of Karol Krawczyk, this entailed using a medical calculator to determine the CHA₂DS₂-VASc score. For Franz Huber, students used a diabetes-management smartphone application to document glucose measurements and food intake across the day (Fig. 3). Because a physical health device cannot be easily shared in an online course, we recorded a video demonstrating the operation of a handheld ECG device on a patient with atrial fibrillation. Other tools – such as smartwatch activity monitors and digital glucometers – were represented through photographs.
Recognising that assessment drives learning, we embedded both formative and summative assessment activities throughout the unit. All assessments include detailed model answers, either immediately available to students or accessible to teachers as needed (Fig. 4). Students completed MCQ questions assessing their understanding of the videos on distributed cognition theory and the categories of digital tools. They also collaboratively completed tables in which they mapped components of clinical reasoning, opportunities, and risks to specific technologies in the context of the virtual patient cases.
With these elements in place, the learning unit was ready for internal review by expert project collaborators who had not participated in its development.
What did we learn?
During the development of the learning unit, we frequently encountered challenges related to the subtle yet important differences in how responsibilities are distributed among health professionals – general practitioners, medical specialists, and nurses – in different countries. Our authors’ real-world experiences also had natural limitations, as several desirable uses of digital tools are still uncommon in routine practice or are avoided in favour of more established, traditional non-digital procedures. These contextual variations required careful deliberation to construct realistic yet forward-looking learning activities.
What surprised us positively was the extent of support generative AI could provide in the creation of digital learning resources. For the first time, we photorealistically generated a nurse persona (Mrs Berman, Fig. 2) who explains to the virtual patient’s family how to use a handheld ECG device. For that purpose the HeyGen platform proved to be a valuable companion of our teachers. Likewise, it was striking how easily DALL-E integrated with ChatGPT could adjust with a short prompt the digital device displays in real-world images made by us, modifying numerical values to maintain internal consistency across the virtual patient cases.
The development process also benefited substantially from experience gained in the earlier DID-ACT project, as many strategies for designing learning units could be adapted and transferred to D-CREDO. Additionally, the two virtual patients used in the learning unit- Krzysztof Krawczykowski (later renamed to Karol Krawiec due to pronunciation challenges among partners which was also an interesting lesson learned) and Franz Huber – were originally created within the former iCoViP project. Although very little of the initial material remained in the final cases, the original versions served as valuable starting points that helped focus our discussions and inspired the overarching storylines.
Stay connected with D-CREDO and follow our journey on LinkedIn for more updates, insights, and stories.
