We needed to scale our educational solution to potentially hundreds of geographically dispersed organizations and thousands of learners, which required the use of advanced technologies to allow for automated, accessible and evaluable learning solutions. Importantly, the interface had to be simple for diverse educators to implement. We prototyped and co-developed, with game design experts and a vendor, software that served as a scalable, self-authoring and self-serve platform that could be accessed online through a personal computer and/or mobile device. These software capabilities enabled us to register learners and to create, manage, monitor and edit the learning experience autonomously and just-in-time.

We assembled a large interdisciplinary team that included simulationists, educators, instructional designers, researchers and clinical subject matter experts to support content development and evaluation. Developers, technologists and graphic designers addressed software design, security and maintenance, as well as the user interface and experience; game designers developed the game world.

The alignment of content and this approach to learning was guided by Dr. L. Dee Fink’s (2013) constructivist Significant Learning Taxonomy [25]. Fink emphasizes the need for higher-level, learning-centred education to prepare learners for the real world. We integrated Fink’s six significant learning categories into the MSGW, including Foundational Knowledge, Application, Integration, the Human Dimension, Caring and Learning How to Learn (see Table 4 for a description of each learning category). Fink suggests these learning categories allow learners to carry meaningful, lasting change with them long after the learning experience has ended, which deeply resonates with our teaching and care philosophies and our aims for this project.

As the area of clinical specialization was LTC, we developed competencies for the following areas of instruction: gerontological and geriatric care, interprofessional care, relationship-centred care, and inclusive care planning and delivery. We created specific measurable objectives for each competency. For example, key competencies related to interprofessional care that we targeted included information sharing. A measurable objective associated with this competency was for the learner to demonstrate an appropriate method of communicating essential health information to others on the care team.

Our fictional game world universe was loosely based on the civilization of Carthage in 800 BC. In keeping with historical legend, the founder was portrayed as a racialized and gendered subject: a young Phoenician woman, known as Queen Dido. We purposefully chose this setting because of its historical and cultural significance and rich mythology, and because it is popularly considered to have been shaped by strong female leadership. Thus, we hoped it would resonate with and inspire our largely female workforce. Even though ancient Carthage was a pre-industrial society, setting the game here suited the performative range of the learning because our LTC homes rely on critical human skills for care interventions. Learners have the opportunity to perform the full range of behaviours they would be expected to demonstrate in LTC practice (e.g. skilled dementia care, communication with family and team members).

We selected the classic ‘hero’s journey’ for the storyline. That is, learners were called to go on a heroic mission in an unknown world, in service of humanity. In this fictional world, each learner met allies, experienced challenges, overcame an ordeal, experienced a revelation and returned home transformed. Learners took on the role of Alyssa, an apprenticing/novice healer in a guild of healers, on a mission in service of the care of older adults. Through this journey, learners were immersed in residents’ lives to promote their understanding of, and reflection about, each resident’s psychosocial care needs (e.g. their personhood, familial relationships, customs, history, environment). The storyline consisted of a prologue, 18 chapters and an epilogue. Within the chapters, learners could choose to follow different branches of the story on their journey where they interacted with other characters (e.g. care team members, the High Healer, allies, villains and mystics). At times, these characters role modelled values or acted as mentors. Through this journey, we hoped to elicit emotions in learners such as anticipation, satisfaction, curiosity, engagement, empathy and compassion. As learning activities in the game became increasingly complex, events in the storyline became increasingly challenging. Thus, the storyline reinforced the notion that becoming a healer requires effort, tenacity and mentoring to overcome setbacks and develop expertise. We harnessed the storyline to support learners in taking pride in their ability to overcome challenges as they advanced through the MSGW, and so, to become a ‘hero’ meant successfully acquiring the competencies needed to provide evidence-informed care to residents.

For the gaming assets, we curated a consistent style of images and videos (both real and AI-generated) of people, avatars, places, nature and artefacts that were meaningfully tied to the game world storyline. We adapted an ‘ancient-themed’ game user interface and game assets (e.g. hand-drawn icons representing ancient scrolls, pre-industrial tools). We selected photographic images of older adults to represent the simulated residents. Our intention was to honour the age and lives of residents whose stories were part of the MSGW. Photographs of landscapes, nature and artefacts were chosen to simulate the reported beauty of the region and to honour its rich history. Where possible, we sought to frame the view of the photographs and videos from the direct perspective of the learner (i.e. in a first-person point of view, as though the learner was in the scene, interacting with the simulated resident or team members, or within the setting). For audio, we used string and percussion sounds that were suited to a world set in 800 BC.

With respect to progressively complex simulations, we designed two levels of simulation. In the first level, we presented learners with text-based mini-simulations involving interactions with various older adults living in an LTC home who were represented by photographs. These brief simulations enabled rapid skill building (drawn from an interprofessional clinical framework) [30], whereby learners assessed part of a clinical situation and applied one discrete skill. For each skill, the learner worked through 10–15 short scenarios and completed multiple-choice answers. Learning was further supported by just-in-time resources such as a one-page fact sheet related to the skill that could be bookmarked within the game, downloaded or printed.

Our first level also presented learners with fundamental quests (e.g. mini-simulations with short quizzes related to introducing and reinforcing basic knowledge about age-related changes). A quest is an in-game mission that is part of the Hero’s Journey and is designed to enrich performance in the simulations and support socialization to the role and setting (e.g. recognition of one’s responsibility as an advocate for residents). Learners were prompted to complete brief learning activities, supported by quizzes, in order to move forward in their Hero’s Journey.

In our second level of simulation, learners applied first-level simulation learning to engage with more complex case-based scenarios and more in-depth, specialty quests. These quests might include access to a journal article to deepen learners’ understanding of the resident’s health conditions, activities of daily living and advanced care needs. This second level involved text-based scenarios accompanied by photographs of the residents that involved the ongoing care of eight residents to illustrate the diversity that exists in an LTC resident population. Because care practitioners needed to understand the experiences of daily life in care, the simulations needed to highlight residents’ clinical needs, as well as the social experience of living in care. Our intention was to have players witness, engage with and be drawn into the lives of these residents who experience changes in health conditions and care needs over time. Each resident had three progressively complex sequential scenarios [31] that followed their life path in LTC.

To balance engagement loops and to manage cognitive load, we worked with game designers to titrate the content and the learning activities. This meant that The Carthage Codex was designed with 80 levels of game-play for an estimated 35 hours of content. This strategy aimed to ensure adequate frequency of levelling up (i.e. every 26 minutes). A new story chapter unlocked every fourth level. Within each gated section of the game, learners explored and made choices about how to move to the next level (e.g. choosing the order in which to assess residents in the mini-simulations, which then unlocked a challenge with the High Healer). Levelling up was celebrated in the form of a pop-up window with a scenic Mediterranean image and a harp sound. With respect to other engagement loops, learners could self-monitor progress bars for levels, library resources and badges. Because learners played solo and asynchronously in this version of the game, social interaction engagement loops consisted of an automated leaderboard (for experience and mastery points) and activity feed.

A digital badging system recognized and rewarded the achievement of key care competencies (e.g. advocacy, family engagement, assessment, acute deterioration, activities of daily living, falls prevention, palliative care), care of virtual simulated residents (i.e. tokens of appreciation at end of care scenarios), and mastery and game milestones (e.g. related to repetition of simulation activities and progression in the library and storyline).

The game economy consisted of a system of points and currencies that could be used to purchase items. For example, we called our system of points ‘experience points (XP)’, which were a cumulative resource earned upon successful completion of each learning activity. Upon successful completion of quest and simulation activities, gold was earned as a currency; in turn, gold could be spent to purchase special powers and guidance to advance through the game. For example, interprofessional team consultations could be purchased with gold to provide team perspectives, fostering collaboration for each simulated case scenario. We used game resources to weight learning activities to ensure that learners were drawn to and rewarded for higher-order learning (e.g. 50 XP earned for a library quiz vs 250 XP earned for a full resident simulation).

We created a set of in-game prebriefing materials for learners, which included short exercises that oriented them to the game world universe, user interface and game rules. We also provided them with a game map to support their understanding of game flow, self-monitoring of progress, exploration and self-directed learning by topic, tracking of achievement, and navigation of their journey.

In response to early player feedback requesting further prebriefing and orientation, we created three short interactive quests related to the fictional game world environment, their role as a healer within this environment, and the rationale for having a pre- and post-assessment. Within this last quest, we made an explicit statement that we valued and appreciated their willingness to engage in the pre-assessment, while reassuring them that no preparation was necessary. We presented the pre-assessment as an opportunity for players to discover independently their areas of strength and their learning needs. We reassured learners that they would subsequently receive support and resources to succeed in the game. Finally, to foster enthusiasm for the game process, we stated our confidence in their ability to progress and succeed toward their goal of graduating as a full healer. We also provided contact information for them should they experience technological difficulties with the website (see Table 5).

In lieu of a facilitator and in-person debriefings, we used a variety of methods for learners to receive feedback and to self-assess their progress. Written with an appreciative tone, feedback for answers was provided through instant, automated tips and reflective prompts, which (1) reinforced best practice and/or rationale behind correct responses; (2) prompted reflection to improve application of learning within the simulations; (3) encouraged and rewarded the learner for meeting targeted outcomes; (4) redirected the learner to key resources or learnings to promote success after an incorrect response; and (5) acknowledged the complexity of clinical practice where many factors must be considered correctly for an optimal response.

We tested The Carthage Codex students, direct care providers, serious game experts, technologists, instructors and administrators. Testers provided feedback related to game design, usability, game flow, content and the storyline. Iterative revisions were then made to the MSGW to address confusing, missing or underdeveloped elements prior to implementation. We piloted the MSGW with two discrete groups of learners (students and LTC practitioners) and implemented their feedback before scaling.

We conducted a comprehensive process evaluation [40]. Although it is beyond the scope of this paper to report in detail on the evaluation process, data collection methods included implementation notes, app usage metrics, platform analytics, in-app and follow-up surveys, and interviews with end users.