Ever since communication technologies were adopted for health care purposes and defined under the umbrella term electronic health (eHealth), the concept of empowerment and the use of technological solutions have become intertwined [1]. As technological devices became more personal and connected, this relationship took on a new relevance. In particular, mobile health (mHealth) solutions, commonly defined as the use of mobile and wireless technologies to support the achievement of health objectives [2], have been used to enhance the self-management of various chronic conditions [3], such as diabetes [4] and asthma [5]. Evidence indicates that these mHealth solutions can foster self-management by addressing multiple risk factors [6] and sustaining long-term adherence to prevention measures, which remains a major issue [7].

Studies have examined not only the effectiveness of mHealth [8], but also its design and development process. Even though there is great potential for using mobile technologies for health purposes, findings show that many of the existing mobile apps are not theoretically grounded [9] and their contents are not based on evidence [3,7,9,10]. This is problematic because studies mentioned that theory-driven health interventions are more effective than those without theoretical grounding [11]. It is only recently that mHealth has started to adopt strategies informed by behavior change theories, but this adoption has not been systematic [12]. Some apps are only partially applying the principles of behavior change theories [13], whereas others have defined the app’s features or its mechanism (ie, goal setting) based on a set of different theories or models, but without clear reference to them [14-16]. Additionally, some authors have attempted to create a framework to develop digital behavior change interventions that integrate, for instance, behavior theories, design thinking, and user-centered design [17,18]. Despite these efforts, to date, many apps are still not based on theory, as attested by recent systematic reviews [19,20].

Another flaw of many existing apps lies in the quality of their content, which does not reflect the latest scientific evidence. Indeed, some ex-post examinations of mHealth apps underlined that their content rarely adheres to evidence-based knowledge [21-24]. Some authors have based the content of their interventions on the results of systematic reviews or additional participatory efforts (ie, involving different stakeholders) [25,26]. However, their procedures are not detailed and cannot be replicated. So far, the efforts to develop a framework for integrating evidence-based content into mHealth interventions have been limited [27]. Hence, it remains unclear how disease-specific recommendations and professional best practices should be selected to inform mHealth interventions. This is problematic as evidence-based information can enhance health literacy [28], which is a precondition for patient participation and informed decision-making [29]. Consequently, apps that are based on outdated or inaccurate content might negatively affect the users’ health and safety [30-32]. Considering the huge amount of incorrect and misleading information available on the internet, as well as in leaflets and other lay publications [33,34], it is of utmost importance that new mHealth interventions tackle the issue of content quality.

Participatory design is a democratic process involving different stakeholders from the early phases of the design process [35-37]. At least the following two premises provide the basis for different participatory design approaches: all stakeholders should be involved in the design phase to inform the approach and this will increase the likelihood of technology acceptance because it will help set clear expectations [38]. It is for a very good reason that many authors underscored the potential of a participatory design approach throughout various steps, such as requirement analysis, definition of features, and user interface design [39-41], but without providing much clarity on the most appropriate involvement of experts and other stakeholders for the selection of content. Participatory design could be a viable approach for achieving the evidence basis of an app. Several guidelines exist, but they are mostly designed for health care professionals rather than for community-dwelling individuals or patients. Selecting the content of an app through a participatory design approach involves understanding which of the recommendations are not only impactful in terms of prevention, but also feasible and applicable for people living in the community.

The objective of this study was to fill this gap by describing a structured procedure for the participatory identification of evidence-based content to ground the development of a self-management app. To illustrate the procedure, we used a project based in Switzerland aiming to develop an app for the prevention and self-management of pressure injuries (PIs) in individuals with spinal cord injury (SCI).

SCI is a complex chronic condition affecting human functioning in all aspects [42], and it is associated with a number of complications [43,44]. People with SCI have a high risk of developing PIs [42]. The incidence of PIs in the SCI population is 25% to 66% [45], and approximately 85% of individuals with SCI will experience PIs at some point in their lifetimes [46]. PIs impact the quality of life of the affected individuals, as their treatment necessitates prolonged inactivity, which often results in a loss of income and a feeling of social isolation [47]. Moreover, evidence shows that PIs can account for approximately one-fourth of the cost of care for individuals with SCI [48].

There is general agreement on the fact that PIs might be often preventable in individuals with SCI [49] and that prevention is more cost-effective than treatment [50]. Prevention is possible through active self-management. However, this self-management remains challenging owing to the many different factors that need to be taken into account. Indeed, individuals with SCI have to play an active role in the prevention of PIs by, for instance, adapting their behavior, which includes repositioning, performing pressure-relieving movements, and keeping the skin clean [51]. The prevention and management of PIs in individuals with SCI could benefit from the development of an evidence-based mobile app that supports individuals in performing the many preventive measures, as well as monitoring and treating early stage PIs.

We used a consensus method for the participatory identification of evidence-based content to ground the development of a self-management app for PIs in individuals with SCI. Indeed, to ensure that individuals with SCI have access to sources of information that are credible, of good quality, and up to date, the information provided in the app should be consistent with the latest available clinical recommendations, including those that are indeed the best available evidence for pressure ulcer prevention and remain the foundation of a prevention program [52].

The recommendations were identified through a three-stage research procedure developed following the main steps of the consensus development method [53]. First, a review of existing recommendations for the prevention and management of PIs in individuals with SCI was conducted. Second, a consensus meeting [54] to select the most important recommendations that individuals with SCI should apply in their daily lives was performed. Finally, the results were consolidated by the expert team. Both the review of the recommendations and the consensus meeting were conducted at the end of 2017, while the third phase was performed during the first quarter of 2018.

Published recommendations on the prevention and management of PIs were identified through an electronic search and consultation with experts between March and July 2017. Keywords for the search were combined from three different domains. The first was related to PIs (ie, pressure ulcers, pressure injuries, decubitus, pressure sores, bedsores, and skin problems), the second was related to self-management (ie, prevention, detection, treatment, self-management, reduction, and risk factors), and the third was related to SCI (ie, spinal cord injury, tetraplegia, quadriplegia, and paraplegia). The online search applied these keywords in both English and German languages. The search was performed in Google as well as PubMed. The research team extracted all recommendations for the prevention and management of PIs that were directed toward or could be applied by community-dwelling individuals with SCI.

The review obtained a comprehensive collection of recommendations that were screened and synthesized (ie, similar recommendations from different sources were merged). The results of stage 1 were presented in a document that was sent to all participants of the consensus meeting for preparation.

A purposive sample of health professionals and community-dwelling individuals with SCI were invited to participate in a consensus meeting [53]. With the help of SCI-specialized medical doctors, we identified health professionals who may have experience with PIs in individuals with SCI. For the recruitment of those working in the inpatient setting, we contacted the different departments of the four SCI rehabilitation centers in Switzerland and requested for collaboration. For the recruitment of health professionals working in the outpatient setting and individuals with SCI, we relied on informal networks. Through this process we contacted a total of 35 individuals, and they were offered two possible dates for the consensus meeting.

The final sample of 15 participants [53] included SCI-specialized medical doctors, nurses, wound experts, psychologists, occupational therapists, physiotherapists, and nutritionists who were from different parts of Switzerland and working in SCI rehabilitation centers in Switzerland, as well as home care providers, home care counsellors, representatives from an accident insurance fund, and individuals with SCI. Table 1 presents the participants’ characteristics.

The consensus meeting was grounded in a systematic consensus planning process that helps to prioritize issues of a different kind during experts’ discussions [55]. The meeting lasted one day and was structured in two parts. A person experienced in consensus meetings moderated the plenary sessions. Three persons facilitated the working groups. They were trained for the technical tasks (eg, dealing with the voting system) and were knowledgeable about the project.

After the consensus meeting, the research team together with two experts from the project scientific advisory board consolidated the results by refining them and taking into consideration the input of the participants. For instance, special attention was devoted to the recommendations that remained ambiguous after stage 2. They were screened and sorted out by the research team based on eight logical rules for their inclusion or exclusion. The rules (Textbox 1) referred to the size of the discrepancy between the results of vote A and vote B, and between the two working groups. Additionally, new recommendations were developed for domains that, according to the participants, were insufficiently covered by the existing recommendations. The consolidation stage resulted in a newer and more complete set of recommendations. These recommendations were shared with all the participants of the consensus meeting via email. Feedback from the participants was collected and integrated.

The sources presented in Table 2 have been identified, and their documents have been systematically reviewed [58-65]. The recommendations extracted from the documents were categorized by applying a deductive-inductive approach. At first, the recommendations were ordered according to the four categories defined by Keast et al (ie, appropriate support surfaces, regular repositioning of the patient, optimizing nutrition, and skin care) [66]. This categorization, however, was not exhaustive. We therefore started an inductive process by grouping together those recommendations that were not covered by the categories defined by Keast et al. We created new categories until all recommendations belonged to one category. We then revised the categories with the aim of reducing their number. We compared among each other the recommendations included in every category and with those included in other categories, and when possible, we merged the categories. Following this procedure, we reached the number of 12 categories. This procedure is similar to the basic rule of the constant comparative method often used in qualitative research, namely the comparison of a new incident with the previous incidents coded in the same category [67].

The result of the review and recommendation categorization was a list of 130 recommendations for the prevention and management of PIs by individuals with SCI organized in relation to the following topics: (1) Support surface (code A); (2) Repositioning (code B); (3) Nutrition (code C); (4) Skin care (code D); (5) Skin assessment (code E); (6) Exercising (code F); (7) Collaboration with health professionals or caregivers (code G); (8) Transfers (code H); (9) Clothing (code I); (10) Body function and structure (code J); (11) Personal factors (code K); and (12) General (code L). The orders of the categories and recommendations within a category do not reflect a priority order. The recommendations were then collected in a preparatory document, which was sent to the participants prior to the consensus meeting.