The design patterns developed in this study show how practical, clinical, organizational, and technical knowledge about support of children with a chronic disease can be shared between patients, relatives, and caregivers using a Web 2.0 system compliant with the WHA eHealth resolution. The pattern language describes solutions to issues related to the design of Web 2.0 systems and visualizes some of the generic qualities of those solutions. Modern management of chronic disease, not the least Type 1 diabetes in children and adolescents, is today focused on the empowerment of individuals and groups in order to enable them to address the prevalent issues in daily life and their own local environment. A reliable and efficient information exchange between patients, parents, and clinicians is particularly important in this setting. The design patterns identified in this study mediate values that have been found to be important in community-based management of chronic disease, e.g. the value of allowing individuals to modify information services according to personal preferences and the value of experiencing 'belonging' to a larger community that shares one's own needs and interests .
Almost a decade ago, a shift in health care delivery from biomedicine to infomedicine was described as the Country of PeoplePower, an utopian scenario where communities, patients, and health workers joined in informed, shared governance over healthcare and clinical decision-making . Drawing on ubiquitous computer communication technologies, patients and clinicians were foreseen to contribute actively to patient records, transcripts of clinical encounters were shared, and patient education occurred primarily in the home, in school and in community-based organizations. Patients and clinicians jointly developed 'quality contracts', serving as foundations for quality improvement systems that aggregate data, while also reflecting attributes of individual patients and clinicians. Exploiting Web 2.0 technologies and the present architecture, management of childhood chronic disease according to the country of PeoplePower scenario is no longer utopian. However, to realize such an infomedical shift in chronic disease care in local communities requires a focused effort involving a broad range of professional groups together with patients and their relatives. Even though central tenants of the Web 2.0 concept are self-evolution and networking, the core technical components must still respond to high software engineering standards. This implies that the implementation of the Web 2.0 architecture in local communities will require systematic approaches that allow capturing of broad perspectives on health service delivery and supporting multi-disciplinary decision-making, e.g. participatory design methods .
Systematic accreditation of learning materials is essential for a Web 2.0 system in the pediatric chronic disease setting. With the evolution of the Internet, traditional authorities have partly been replaced by apomediaries, which are tools and peers standing by to guide information seekers to trustworthy information, or adding credibility to information . For apomediation to be a successful model for children with a chronic disease and their families, it has to be adjusted to the degree of maturity and autonomy of each child and parent. In Web 2.0 environments, the implementation of apomediary credibility at the application level may therefore become equal to or even more important than guaranteeing source credibility, e.g. as expressed by the HoN code . This implies that there are several practical challenges for developers of pediatric chronic disease websites which aspire to come across as credible. Children and their parents need and want to be able to be co-creators of content, not merely an audience that is broadcasted to. Web 2.0 technology enables such sites. At the same time, to reduce late complications, modern clinical management of chronic disease is based on appropriateness and calibration principles [8, 9], where the guidelines provided to patients are individualized and precise. Hence, developing credibility in the Web 2.0 context is also about developing the actual community in which the system is implemented. Chronic disease communities are built upon personal and social trust, and organized means to build and maintain that trust. It is thus necessary to find the appropriate balance in each community between patient co-determination and medical authority in chronic disease management, and to express this balance in the design of the apomediation components of the Web 2.0 system.
In the Web 2.0 context, accreditation issues are, however, not only limited to learning processes and content. Use of participatory design methods for implementation of Web 2.0 systems based on the architecture raises concerns about the management of responsibility in the total health service delivery perspective. Unlike system developers, practitioners and patients that participate in Web 2.0 system development and maintenance cannot be made professionally accountable for shortcomings in the system performance. In this context, it has to noted that the pattern language does not indicate which features require continuous organizational service offerings, and which can be managed by user communities, e.g. whether workspace assignments, addition of new software modules, and protection of existing software against hostile code can be administered by the user community. Already the original Internet's design relied on few mechanisms of central control. This lack of control has the added generative benefit of allowing new services to be introduced without up-front blocking by either individual users or public authorities . To avoid downstream problems when implementing the pattern language in healthcare settings, ambiguities regarding the responsibility for software quality and maintenance need to be straightened out at an early stage. To mitigate the risk for late design failures, participatory design groups implementing Web 2.0 systems based on the pattern language have to include software engineering competences to cover areas such as system extensions and maintenance, and the design groups have to accept full responsibility for decisions made in these areas.
In design areas such as urban planning, pattern languages have been extensively used to transfer value-bearing features between different contexts . A core set of design patterns constitutes the defining part of a pattern language developed for a particular area. Using the present language as a basis, a module-based Web 2.0 system design to support community-based childhood chronic disease management can be developed in different social and cultural contexts. Availability of high-level web programming languages, such as Joomla! http://www.joomla.org, facilitates the software implementation process, providing access to modules corresponding to most components of the general architecture. The core set of design patterns delineate a smallest common denominator for the design specifications that are developed in order to implement particular systems. System developers in different community settings can then add the lower-level modules they need around this core in order to represent their actual program, and also add new sets of design patterns. In this way, each system design is tailor-made for its particular patient population and environment, while it still shares basic value-bearing features. Nevertheless, present high-level web programming languages are not adequate for development of all applications included in the pattern language. For instance, confidential patient-practitioner communication will in most settings require encryption to be added to the email messaging module. It is possible that over time, generic software modules can be developed for all application-level instances in the pattern language.
From a health informatics point of view, the most important feature of the pattern language is that it allows professional information system developers, clinical practitioners, and patients to share a common expression of the most important organizational and technical design issues during the development of Web 2.0 systems for use at the community level. What most previous methods, such as the Enterprise Architecture Framework  have in common is that they have been based on concepts developed for use by IS specialists. However, there is a substantial difference between an IS aimed to be used by professionals trained in formal analyses, and a Web 2.0 system that is to be shared by clinicians and laypersons of different ages at the community level. In the latter setting, identification of the needs of each different user group, lay and professional alike, is essential and requires that intelligible input can be delivered from all these users into the system development process. This, in turn, requires that the materials describing the system design are legible also for persons without a technical background.
The next step in the development of the Web 2.0 system design is the formative evaluation of the implemented prototype. This step is to address human-computer interaction and information content issues. A number of methods for evaluating health-related Web 2.0 systems are available, however, most of these use differing conceptual definitions and scoring approaches [24, 25]. Additionally, the majority of methods proposed are also limited in their interpretability and ease of use, because they are usually defined to assess either health information content or usability from a human-computer interaction perspective. However, preliminary versions of comprehensive multi-disciplinary methods that cover both aspects have recently been published. For instance, the WebMedQual guideline for website assessment contains constructs for the assessment of information content, the authority of source, design, accessibility, links, user support, and privacy . When such guidelines have been further developed into assessment standards, they will provide essential complements to pattern languages in the development of health and safety promotion websites.