Physical activity (PA) is important for children with a chronic disease [1,2]. Asthma is the most frequently diagnosed medical condition in children [3]. For children with asthma, PA can positively affect their asthma control by improving their physical fitness. This can reduce the threshold of triggers causing asthma symptoms, which in turn may lead to decreased use of medication and increased quality of life [1,4-6]. Besides the positive effects of PA regarding physical fitness, stimulation of PA in children with asthma is also important for their psychosocial functioning, because PA is often intertwined with social life. Children with physical disabilities are less involved in social activities compared to their healthy peers, and children with asthma often feel left out of the group [7-9]. Despite all benefits of PA, children with asthma—especially girls—seem to be less physically active than their healthy peers [10-13].

In the Netherlands, children are treated for asthma according to the Dutch Pediatric Society guidelines for asthma in children [14]. This treatment includes taking medication, receiving information, and having regular check-ups. Children with asthma who are inactive are referred to health care providers, such as pulmonary nurse practitioners and pediatric physical therapists. They provide health education, encourage children to increase their PA levels, and pay attention to social participation and children’s self-confidence regarding PA. Moreover, they educate them on how to cope with exercise barriers and provide exercise training. Such training interventions, which include, for instance, swimming and aerobic exercises, are shown to be effective in increasing cardiovascular fitness and in improving pulmonary function of children with asthma [6,15-19]. However, health care providers indicate that they lack interventions that result in maintenance of improved PA levels. This is because existing training programs do not aim to achieve behavior change in children with asthma [15,17,20].

This raises the question as to how PA levels can be both increased and maintained in children with asthma who are inactive. Various studies and reviews show that effectiveness of PA stimulating programs can be increased by technology-supported interventions [21-29]. Although most studies focus on adults, several studies show that a technology-supported approach can have positive effects on PA and physical functioning in children and adolescents as well [30,31].

Features of technology-supported interventions, especially behavior change interventions, are often inadequately described [23,24]. Therefore, it is difficult to compare interventions and their effects. To ease and improve this comparison, it is important to describe the design process and the developed components of an intervention [32,33]. Several frameworks are already used to describe interventions that stimulate PA, such as the MRC (Medical Research Council) framework to stimulate PA in older adults, the IDEAS (Integrate, Design, Assess, and Share) framework to promote PA in adults, the CeHReS (Centre for eHealth Research) framework to promote PA in people with depression, and the intervention mapping (IM) model to stimulate PA in patients with heart failure [33-40]. For our intervention, we chose the IM model, because it explicitly differentiates between personal and environmental factors affecting PA, and it takes into account creating an implementation and evaluation plan. Moreover, IM was already successfully applied for developing health interventions for children [41-43]. The framework enables systematic development of evidence-based interventions and enhances adoption, implementation, and maintenance of the developed intervention [39].

As part of a project that aims to promote PA in children (aged 8 to 12 years) with asthma, we attempted to develop an intervention that supports health care providers in stimulating PA in children with asthma using the IM model. In a preliminary study, we explored promoting factors for PA in children with asthma according to children with asthma themselves, their parents, and their health care providers [44]. In this study, we applied the IM model to develop a technology-supported intervention that incorporates both these promoting factors and behavior change principles that were discovered. We aimed to transparently report the development process and the final components of the intervention using IM to meet the demand of sufficiently describing technology-supported interventions [32,33]. We answered the following research question: “How can scientific evidence on stimulation of PA, the needs and wishes of future users, and behavior change principles be translated into features of an intervention that facilitates health care providers in promoting PA in children with asthma?”

By systematically describing the translation from requirements to functional components and our specific design and implementation choices, we give insight into the background of our intervention and ease comparisons with other interventions. In addition, we report the contribution of future users, because the intervention was developed in cocreation with children with asthma, their parents, and health care providers.

The IM protocol includes six steps: (1) conduct a needs assessment in which a distinction is made between personal and environmental determinants; (2) define the intervention outcome, performance objectives, and change objectives; (3) select theory-based intervention methods and strategies; (4) create components of the intervention and conduct pilot tests; (5) create an implementation plan; and (6) create an evaluation plan.

To assess the health problem, background information and scientific evidence regarding children with asthma, the impact of low levels of PA, and available interventions had to be gathered according to the IM model. In our preliminary study, 3 stakeholder groups were questioned to explore promoting factors of PA in children with asthma. To do so, concept-mapping sessions were held with 25 children with asthma (aged 8 to 12 years), 17 parents of children with asthma, and 21 health care providers of children with asthma. The health care providers were lung nurse practitioners and pediatric physiotherapists, experienced in behavior change and supporting PA in children with asthma. The resulting factors were labeled as either personal or environmental factors according to the Physical Activity for People with a Disability model [45]. More details regarding the needs assessment can be found in our preliminary study [44]. On the basis of this needs assessment, we defined the aim of the intervention. The reported factors will be used in subsequent steps of the IM framework, such as designing the program’s requirements and functional components.

To examine the target behavior, we created a logic model of change according to the IM framework. First, we defined the program outcomes based on the stakeholders’ reported clusters in the needs assessment. They describe the desired changes in the behavior and the environmental conditions that are required to reach the goal of the intervention [39]. Second, we translated the program outcomes in performance objects based on the stakeholders’ reported supporting ideas in the needs assessment. Those indicate the required behavior to reach the goal of the intervention. On the basis of these performance objectives, change objectives were defined which indicate the necessary behavior to accomplish the required behavior change and environmental conditions.

On the basis of the needs assessment and the logic model of change thus created, we selected several methods and strategies that were translated into design considerations for the intervention. This was performed in a stakeholder session with a pediatric pulmonologist, a pediatric physiotherapist, and researchers specialized in exercise behavior in children, behavioral change, health technology, and game development. The final selection was a product of mutual agreement.

On the basis of the needs assessment, logic model of change, and the design considerations, we formulated several requirements for the intervention to be developed. With a multidisciplinary team involving health care providers, data scientists, game designers, and software developers, we translated the requirements into functional components. This translation was performed in 2 sessions and the final list of requirements was a product of mutual agreement. To develop those components, we performed an agile method consisting of several design and develop iterations. Each iteration lasted for 2 weeks, and there were 10 iterations in total.

In the first 3 iterations, we created story lines, mock-ups, and user journeys resulting in a prototype. In a stakeholder session, this product was tested by children with asthma and their health care providers. Children who participated in the needs assessment were asked whether they wanted to participate in this stakeholder session as well. Children who were diagnosed with asthma and aged between 8 and 12 years could be included. The 3 participating children were aged, on average, 9.3 (SD 1.5) years, and 2 of them were female. One of the children was diagnosed with severe asthma, one of them had mild asthma symptoms, and one of them had moderate asthma symptoms. The 4 participating health care providers were all pediatric physiotherapists with >5 years of work experience with children with asthma. In the stakeholder session, the participants individually tested the prototype and shared their experiences with the researchers. The children were asked to focus on the app’s functionalities, the story line, and the usability of the PA monitor, whereas the health care providers were asked to focus on the dashboard’s functionalities and the clarity of the graphs showing the children’s PA. Afterward, the researchers compared the experiences of the participants and translated them into requirements for improvement of the prototype. In the remaining 7 iterations, these improvements were made, and the prototype was further developed into a usable product.

In a stakeholder session with a pediatric physiotherapist, a pediatric pulmonologist, and a lung nurse practitioner, we examined the health care provider’s daily practice and the current health care situation for children with asthma who are inactive receiving care regarding PA. We explored the available time and skills of the health care providers, to support children with asthma who might benefit from PA with an intervention.

For a blended intervention, it is important to evaluate the usability and feasibility in children’s everyday life and in their health care provider’s daily practice. Therefore, we selected methods to evaluate user experiences, adherence, and usability of both children and health care providers. In addition, to evaluate the possible impact of the intervention on children, we selected methods to look into the self-perception of PA among children, their social participation, and their enjoyment of PA.

The Amsterdam University Medical Center Medical Ethical Committee approved the research protocol with trial number METC 2017 191. The study was registered in the International Clinical Trial Registry Platform with trial number NTR6658 on August 21, 2017.

Several studies show that children with asthma benefit from PA. Increased levels of PA are related to both improved cardiovascular fitness and quality of life [4-6,16,19]. However, children with asthma experience limitations in PA and social activities because of their disease [7-10,46]. Reported explanations for these experienced limitations are, for instance, inaccurate symptom perception of airflow limitation, misinterpreting healthy exercise-induced shortness of breath as an asthma attack, and low self-efficacy in relation to PA [47,48]. Parents endorse the limitations that children experience and report parental fear for exercise-induced asthma, challenges with asthma management, and lack of trust in teachers or sport coaches as parental barriers [49,50]. Increasing self-esteem and improving knowledge about PA in relation to asthma and medication are reported as strategies to reduce children’s perceived limitations in PA [51,52].

Figure 1 shows differences and similarities among all reported factors of the 3 stakeholder groups. Factors that were rated as very important were getting positive feelings of PA, playing together, being stimulated and motivated by relatives, good asthma control, making PA enjoyable, rewarding PA, tailored PA, playing outside more often, setting realistic goals, relatives of child have sufficient knowledge about asthma in relation to PA, child itself has sufficient knowledge about asthma in relation to PA, digital interventions for PA, increasing extrinsic motivation, positive attitude of child, scheduling PA together with child, and variation in PA.

The aim of the intervention was defined as follows: the intervention needs to support health care providers in improving PA levels in children with asthma who might benefit from increased levels of PA. The intervention will be developed for children who are inactive, with mild to moderate asthma symptoms, who are referred to a health care provider, either a lung nurse practitioner or a pediatric physiotherapist, to improve their PA levels.

All program outcomes, along with their corresponding performance and change objectives, are shown in Table 1.

The following three design considerations were formulated: (1) support behavior change, (2) blended technology, and (3) integration in everyday life.

To start with, the intervention should support behavior change regarding PA. Gamification is frequently used in the improvement of health and well-being and has shown to be particularly effective in behavior change regarding PA [53]. Gamification is defined as “the use of game design elements in non-game contexts” [54]. Conceptually, gamification combines design elements for behavior change techniques from persuasive technologies, intrinsically motivating qualities from serious games, and methods for tracking individual behavior from personal informatics [53,55-59]. However, in practice, most gamified eHealth applications merely implement short-term engagement through extrinsic rewards [60]. For children with a chronic disease specifically, the nonsignificant results of several serious games aiming to increase PA might be explained by the lack of incorporating behavior change theories in the game [61]. To reach the full potential of gamification, it is necessary to design tools based on theories that support the overall goal, behavior change, and the effects of game mechanics [60].

To maintain improved levels of PA, children must develop new habits. An important factor in habit formation is rewarding the required behavior [62]. Rewarding increases positive feelings in relation to the behavior, in our case performing PA [63]. To develop intrinsic motivation, such positive feelings are important [64]. Increasing self-efficacy is also shown to be effective, for instance, through action planning and shaping knowledge [65-67]. Moreover, goal setting, monitoring and feedback are shown to be effective persuasive methods for improving PA behavior [67-69]. Personal coaching from health care providers, for instance, with the widely used motivational interviewing technique, might enhance the behavior change process [70,71].

Second, the intervention should use blended technology, meaning that individual use of digital technology is combined with face-to-face interaction with a health care provider. Digital technology was shown to be valuable for improving health in children and supporting behavior change [30,72]. However, the results of such digital technologies can be optimized by offering them in combination with personal coaching [30]. Personal coaching is important to provide feedback and support goal setting, which are both identified as success factors when behavior change is aimed [67-69]. Moreover, personal coaching by health care providers reinforces a tailored approach, which was also identified as a success factor of eHealth interventions that induce health behavior change [73-76]. Both a tailored approach and support of a health care provider are important supporting factors of shaping knowledge, which is a frequently used technique in supporting healthy behaviors [67]. Children with asthma specifically need to learn more about the positive relation between PA and asthma control, how to properly take asthma medication, and the difference between healthy shortness of breath and an asthma attack [47]. Moreover, feedback and knowledge regarding accurate symptom perception is important for perceived exercise limitations. Some children report substantial discomfort when there is limited bronchoconstriction, whereas others report no symptoms even if severe obstruction is present [48].

Finally, it is important that use of the intervention and performing PA could be integrated in everyday life of the users. Although participation in everyday activities is a vital part of children’s development, children with physical disabilities or chronic diseases are less involved in such social activities than their healthy peers [8,58,77,78]. Children with asthma report that they often feel left out of the group [9]. Integration in everyday life is therefore important from 2 perspectives. On the one hand, we have to prevent that using the intervention creates a special situation as this might reinforce children’s feeling of being different and excluded. On the other hand, PA is often intertwined with social life. By focusing on PA in everyday life situations instead of sports specifically, children might experience better social participation.

The developed intervention Foxfit will be tested by approximately 15 children (aged 8 to 12 years) with asthma who are inactive. They will all be treated in a hospital in the Netherlands, specialized in pediatric asthma. During the evaluation study, the intervention will be provided by pediatric lung nurse practitioners and functional exercise therapists from those hospitals.

To evaluate the usability and feasibility in children’s everyday life and in health care providers’ daily practice, both children and health care providers will be interviewed halfway and at the end of their participation about their user experience, adherence, and suggestions for improvement. In addition to the qualitative interviews, they also fill in the System Usability Scale, which measures the usability of the developed intervention [84,85].

To obtain insights into the effect of the intervention, children will fill out several questionnaires at the start and end of the intervention. Those questionnaires regard asthma control, PA participation, enjoyment of PA, and self-perceptions with, respectively the Asthma Control Questionnaire, Activity Questionnaire for Adults and Adolescents, PA enjoyment scale, and the self-perception profile for children (competentie-belevingsschaal voor kinderen) [86-91]. Regarding children’s physical fitness and endurance, they perform a 6-minute walk test. In addition, activity data, collected with the PAM, of all participants will be analyzed. We will compare activity levels of the baseline week with activity levels of the last intervention week.

A detailed description of the evaluation together with its corresponding analyses and results will be published separately.

This study illustrated the way IM was used to translate theory and stakeholder experiences into design requirements and functional components and finally develop a complete blended and technology-supported intervention. IM was useful for describing the translation and development in a structured way. The needs and wishes of children with asthma, their parents, and health care providers were combined with scientific evidence regarding behavior change techniques and gamification in a technology-supported intervention. This resulted in the blended intervention Foxfit that consists of an activity tracker, a smartphone app for children (aged 8 to 12 years), and a web-based dashboard for health care providers. It focuses on PA in everyday life to improve social participation and contains behavior change elements such as goal setting, rewards, action planning, monitoring, shaping knowledge, a tailored approach, personal coaching and feedback, and a gamified story. Because of the cocreation with future users, the support of the resulting product has been strengthened among children with asthma, their parents, and health care providers. The structured description of the development process ensures that Foxfit can be compared easily with other interventions.

The participation of stakeholders in this project was very strong, because the intervention was not only created for the target group but together with them. Consistent with other studies that applied IM to develop health interventions for children, we experienced that the process of IM is very time consuming, but the systematic approach improves clarity and analysis of the intervention [41,42]. The theoretical foundation increases the potential of the intervention to realize the desired outcome of supporting PA, because it tends to provoke more significant effects on PA [30,65,92,93]. Moreover, clarity and analysis of the intervention is important, because many gamified interventions for children with chronic disease lack a detailed description of the final intervention [61].

Because of the structured approach, it is relatively easy to apply this approach to other situations. On the one hand, the current gamified story and integrated behavior change techniques can be applied for children with other chronic diseases. Only the content of the knowledge components and the specific questions about asthma symptoms should be adjusted to the new disease. On the other hand, the integrated behavior change techniques can be applied for children in older age groups in case the gamified story is adjusted to their interests and experiences. Because children should be able to handle a smartphone with the Foxfit app themselves and understand why they need to wear the activity tracker, we do not think that the intervention is suitable for children younger than the current minimum age of 8 years.

An important factor of Foxfit is the support of the health care provider. Because of the blended approach, children receive personal feedback, and the intervention can be tailored. These components are success factors when behavior change is aimed for and requires considerable time from both children and health care providers [67-69]. The intervention takes 7 weeks, which will not be long enough to reach lasting behavior change. However, 7 weeks of monitoring and coaching can lead to positive PA experiences and therefore, improved PA levels. Extended use of the intervention, possibly with decreasing involvement of the health care provider, can help maintain these improved PA levels and form new habits. In the evaluation study we will obtain an indication of the effect of the intervention on improved PA levels.

Time is scarce and expensive for the health care providers. Therefore, we tried to diminish the coaching time by enabling web-based coaching meetings and by minimalizing health care providers’ preparation time. Despite these optimizations, it might be difficult to structurally implement the Foxfit intervention in the health care system. Because of the advantages of personal feedback from health care providers on children’s motivation, we would not recommend to completely omit this. However, the investment time of health care providers might be further diminished through the use of artificial intelligence (AI) for automatically personalizing the intervention. Such a technical system might, for instance, suggest health care providers weekly activity goals based on activity data of the past weeks. This approach takes the attitude of both patients and health care providers toward AI into account. Patients appreciate the human factor of health care providers and might have a negative attitude toward medical AI because of the absence of this humanistic care factor [94]. Moreover, health care providers appreciate AI systems to make the process more efficient or support them by giving suggestions, but ultimately, they want to make their own choices [95]. By giving AI-based suggestions for weekly PA goals, there still is clinical expertise and personal feedback, but the health care providers are supported in their decisions and the process is made more efficient.

Although coaching from the health care provider is an important component of the intervention, it would be interesting to study whether the intervention can be used as standalone app for children as well. Although the intervention was designed for children with asthma who are inactive, there are also several children who already have moderate PA levels but who may still be able to diminish the impact of their disease when their PA levels are improved. A stand-alone app might, for instance, be helpful for them as well as for children who completed the entire blended Foxfit intervention previously and now need a repetition. To examine the possibilities of a standalone app, we should study whether some of the health care provider’s tasks and dashboard functionalities can be implemented in the smartphone app. Moreover, in some cases, highly educated parents could possibly take over the coaching role of the health care provider when they receive proper instruction.

Regarding the hardware, the PAM that children must wear is extremely easy to use and safe in terms of personal data. No personal data are stored, and the activity data are not shared with a commercial organization. The PAM does not require charging during the intervention and it synchronizes automatically. Thus, the impact on children’s everyday life is minimal. However, this system requires Raspberry Pi with Bluetooth and 4G dongle at home. This takes up space, requires an electrical outlet, and comprises of small components, which might be inconvenient when having very young children at home.

Although a large group of children participated in the needs assessment, it was challenging to include children with low PA levels for our pilot test. In the end, more children with sufficient PA levels than children with low PA levels pilot-tested our prototype than children with low PA levels. This might have influenced the feedback of the prototype and therefore the choices made for the final intervention. For example, in the needs assessment, children suggested that competitive elements would help them become physically more active, whereas participants of the stakeholder sessions indicated that competition is not desired, because children’s motivation might be reduced when they compete against children having better PA possibilities. A solution might be to compete against each other based on children’s personalized goals. In the feasibility and usability tests, user experiences and suggestions for improvement of a larger group of children with asthma who are inactive will be explored.

According to the Fogg Behavior Model, someone will not automatically perform the desired behavior when the opportunity, capability, and motivation are sufficient, but there will be a moment of opportunity in which someone can be persuaded to perform the desired behavior in case he receives a trigger [96,97]. In technology-supported interventions, providing such an external trigger at the moment of opportunity is called a just-in-time notification. Although such notifications might have reinforced children’s motivation, we did not choose to implement them, because one of our requirements was that the intervention should be usable without the children having to carry the smartphone with them all day. The PAM did not technically support such notifications, but incorporating these just-in-time notifications might strengthen the motivational aspect of our intervention in the future.

IM was useful to structure the development of the blended and technology-supported intervention Foxfit, and to describe this development process. Foxfit facilitates health care providers in promoting PA in children with asthma. The intervention consists of an activity tracker, smartphone app for children, and web-based dashboard for health care providers. Important behavior change elements of the intervention are goal setting, rewards, action planning, monitoring, shaping knowledge, a gamified story, personal coaching and feedback, and a tailored approach. Not only is the intervention based on behavior change techniques, it also meets the needs and wishes of its users because it was cocreated with children with asthma, their parents, and their health care providers.