**6. Discussion**

This report presented a proof of concept of a pervasive gamified platform framework that is a unique combination of integrated sensors (wearables), a virtual coach and serious games in healthcare. This description of this platform contributes to research on pervasive technology approaches in healthcare [49,61]. Behaviour change theories and game design principles underpinning the framework were presented with particular attention to how these theories and principles were combined and applied to develop content across the technology components to promote positive health behaviours among young people with T1D. This work contributes to efforts to link theories more closely to interventions in the research literature [62]. This report also presented the results of pilot tests of

early prototypes of the framework. Consistent with previous literature documenting issues that often arise in early evaluations of prototypes, [18,61,63,64], we found that similar issues of usability and technical quality assurance presented challenges for users. We received several comments that the games were difficult to play and were "overly complex" due to design issues and should be more simple and straightforward to play. At the same time, participants commented that the "look and feel" of the games seem to be targeted to a much younger audience and should be designed for a more mature audience. These issues made it difficult to answer some of the original questions of the research (e.g., Do players like to step into the "magic circle", and Does striving for progress in the game motivate healthy behaviours). In addition, we were unable to evaluate fundamental questions about use of the platform and engagement in health behaviours because our findings did not confirm a widespread assumption in the literature that users carry their mobile platforms with them "everywhere" and that the platforms are constantly connected to data networks [61,63,65]. Specifically, we found that because of the central server architecture, progress in the game and coaching critically depended on the user being continuously connected. What became clear from the evaluation "in the wild" is that we incorrectly assumed that users would be continuously online, they would wear their physical activity wrist band all day and they would consistently send their BG data to the sensor network immediately after measuring. Some carried their own mobile phone (not the phone provided) and used the phone provided for gaming only at home. This led to a less than optimal user experience. Thus, the prototype we evaluated was not truly "pervasive" in practice given technology constraints and individual differences in user preferences for using the technology. Finally, the variation in use and feedback of the platform also supports claims that simply applying gamification principles to an intervention does not guarantee increased engagement with a technology, much less, increased motivation to engage in health behaviours targeted by the intervention [49]. The games, and the platform overall, could be improved by simplifying the interface so that it is more intuitive and easy to use, reducing the difficulty of the games for increased playability, modifying the look of the game so that it is more appealing to the target users, and addressing technical issues reported so that the platform is free of major technical bugs that inhibit enjoyment and prevent use of the technology. The implementations of these improvements should be further evaluated because recent research suggests that participatory design processes do not always lead to positive effects on health behaviours [66].

The results from the pilot study contribute to and further the literature on challenges faced when deploying pervasive technologies in healthcare [61]. The negative feedback we report in particular is significant because it presents a more balanced report of technologies (reporting negative as well as positive findings) that contribute to theoretical and practical progress in the field. This addresses a pervasive problem that negative reports of findings including negative user feedback of new technologies are not reported [23,61,67]. The findings from the pilot study also contribute to research on the state of information technologies that suggest that more reliable technology systems should be developed and deployed in healthcare [68]. In sum, the balanced reviews provided in this report not only contribute to scientific efforts to understand the application of technology to healthcare but may also be of practical value for informing the strategic planning of technology development efforts to improve healthcare. Specifically for the latter, in the future, evaluations of similar prototypes such as ours should consider the current findings that suggest specific areas of challenge (e.g., usability, acceptability of look and feel, technical performance) should be addressed before evaluating the impact of the prototype on health outcomes.

The pilot studies show the importance of finding a good balance between flawless and non-obtrusive technology, between challenging but not too difficult game play, and the relevance of designing and fine-tuning to the age group as well as to the level of knowledge and experience of the individual user. Furthermore, the use of mobile technology in combination with web-based elements has been shown to be feasible with potential to change healthcare by supporting a vulnerable group of patients in self-managing their chronic disease. We emphasise the value of mixed methods long-term user studies "in the wild" with patients from the target group in their daily environment, as well as the involvement of the medical team in their role as care providers in these types of user studies.
