This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Human Factors, is properly cited. The complete bibliographic information, a link to the original publication on http://humanfactors.jmir.org, as well as this copyright and license information must be included.
Our health care system fails to deliver necessary results, and incremental system improvements will not deliver needed change. Learning health systems (LHSs) are seen as a means to accelerate outcomes, improve care delivery, and further clinical research; yet, few such systems exist. We describe the process of codesigning, with all relevant stakeholders, an approach for creating a collaborative chronic care network (C3N), a peer-produced networked LHS.
The objective of this study was to report the methods used, with a diverse group of stakeholders, to translate the idea of a C3N to a set of actionable next steps.
The setting was ImproveCareNow, an improvement network for pediatric inflammatory bowel disease. In collaboration with patients and families, clinicians, researchers, social scientists, technologists, and designers, C3N leaders used a modified idealized design process to develop a design for a C3N.
Over 100 people participated in the design process that resulted in (1) an overall concept design for the ImproveCareNow C3N, (2) a logic model for bringing about this system, and (3) 13 potential innovations likely to increase awareness and agency, make it easier to collect and share information, and to enhance collaboration that could be tested collectively to bring about the C3N.
We demonstrate methods that resulted in a design that has the potential to transform the chronic care system into an LHS.
It has long been known that within the health care system, patients across care settings are prescribed half of indicated care [
What if we could create a vastly better chronic illness care system by harnessing the inherent motivations and collective intelligence of patients and families, clinicians, and researchers, so that all could collaborate, at scale, to improve health?
This provocation was the seed for the collaborative chronic care network (C3N) project [
Design processes are widely used to create and modify products, services, and systems [
The objective of this study was to report the methods used, with a diverse group of stakeholders, to translate the idea of a C3N to a set of actionable next steps. Although we have previously described elements of the C3N model [
This research was reviewed and approved by the institutional review board at the corresponding author’s home institution.
ImproveCareNow was launched in 2007 to improve health, care, and costs for children with Crohn’s disease and ulcerative colitis; together, IBD [
The C3N project, funded by an National Institutes of Health (NIH) Transformative Research Awards, aimed to design, prototype, and pilot a C3N—a potentially transformative system for chronic care. The C3N project partnered with ImproveCareNow to help it transform itself from a QI network into a C3N
The C3N team that led this work was composed of a pediatrician and epidemiologist, a behavioral and social scientist, improvement experts, designers, an expert in collective intelligence, and project management staff. Subject matter experts were integrated into the leadership team: pediatric IBD patients, parents of patients with IBD, pediatric gastroenterologists, and other pediatric IBD clinicians. All members of the leadership team also participated in the design process.
The initial design meeting included youth with IBD, parents of children and youth with IBD, pediatric gastroenterologists, nurses and other clinicians, and a variety of other experts including designers, technologists, artists, QI specialists, social scientists, intellectual property experts, and community organizers.
Participants in this initial meeting were recruited from ImproveCareNow by identification of individuals through literature and Internet search, as well as by snowball sampling, in which existing participants are asked to nominate other potential participants.
We combined several approaches to design the C3N. These are illustrated in
Design process used. KDD: key driver diagram.
ImproveCareNow collaborative chronic care network (C3N) key driver diagram. QI: quality improvement.
We used theories from leadership of social movements [
We built cross stakeholder collaboration by emphasizing the enormity of the challenge and the need for everyone’s expertise and effort using specific messaging around urgency, hope, and self-efficacy to encourage participation (eg,
We set out to create a small team of intrinsically motivated innovators working voluntarily together to realize innovative ways to tackle the thorny problem of systemic improvement of chronic illness care [
There are inherent power gradients in health care, and we managed these by explicitly acknowledging this dynamic, by privileging patient and family voices (eg, framing the context for all design meetings by having a patient or parent share their public narrative), and by active facilitation to ensure that patient and family voices were included.
We used the idealized design process [
Observation is the primary method for understanding patient needs and for generating ideas to meet those needs [
We conducted an environmental scan to identify ideas and concepts that could fulfill user needs. We used a broad set of tools including key informant interviews, literature review, Internet searches, and group discussions.
We used goal-directed design—described in-depth elsewhere [
Understanding user needs and the current state informs the development of high level outcome measures. During design meeting 1, participants predicted the needs and goals of patients or families, clinicians, and researchers. These were synthesized to create measurement concepts and then outcome measures to assess the ability of a C3N to achieve its aims, centered on human needs, during the testing phase.
Design synthesis is the abductive process of organizing and manipulating observations, data, and ideas into a coherent whole, both synthesizing observations into interventions and synthesizing interventions into a concept design. Observations were synthesized into a conceptual framework (a high-level description of what a C3N is and ought to do), a set of
During design meeting 2, participants used the secondary drivers and personas to generate scenarios and a set of innovations (prototypes) that could be tested for their ability to change outcomes, whether individually or in combination. Participants also screened and elaborated on the KDD, metrics and targets, and possible innovations.
Ideas were screened through criteria such as “Is it...desirable?” “...different?” “...feasible?” and “Will the idea move us beyond current best practice?” In design meeting 3, participants rated each intervention concept based on potential impact and degree of understanding or knowledge for implementation. Using this 2 x 2 matrix, intervention concepts could be rated high impact and high knowledge (implemented relatively easily to good effect), high impact and low knowledge (could have a positive effect but require further development), or low effect with or without high knowledge (screened out because of little or no expected impact). Intervention concepts classified as high impact but low knowledge became prototyping candidates owing to their potential for teaching us the most about particular interventions employed as part of a peer production knowledge network such as C3N.
Design meeting 1 objectives:
Collaborative chronic care network (C3N) design meeting participants:
Meet and develop a level of comfort and familiarity with each other
Develop an appreciation for the broad range of expertise, experience, and approaches that each brings to the design process
Develop a shared understanding of the common purpose of the C3N
Understand the phase 0 design process
Develop a shared initial vision of the final C3N
Understand how their work fits into the C3N
Develop a clear articulation of the problem(s) that the human-centered design should address
Understand the health care ecology model
Give feedback to a proposed set of specific characteristics (for patients, families, clinicians who are part of networks, and researchers) that will be sampled for during the human-centered design process
Develop predictions of what the end users will say about their needs
Design meeting 2 objectives:
Introduce and reintroduce C3N participants to each other
Obtain input on a refined vision, purpose, values and principles, and metrics
Screen and elaborate a proposed initial system driver diagram
Screen and elaborate information, technical and experience architectures for the patient-facing portions of the design
Identify possible studies and prototypes
Design meeting 3 objectives:
Align participants around the C3N system driver diagram
Rate the secondary drivers as to their contribution or importance to desired system outcomes
Ensure that potential prototypes or work products for the next phase of the project sufficiently cover the highly rated components of the system driver diagram
Prioritize the prototypes or work products for the next project phase based on those that will advance the improvement and research and development efforts
Begin to scope the required effort, team composition, and other resources for a number of highly rated prototypes or work products
The design process began with a relatively small team of 25 members. By actively reaching out to more participants and inviting them to “make a difference” and “be part of the solution,” we increased the number of people involved over the design phase to 150 people (9 C3N team, 28 clinicians, 9 designers, 6 informatics experts, 11 patients, 5 parents, 54 collaborators, and 27 project and research staff), exchanging over 1700 posts and messages on the private social media site. There were 8 clinicians, 3 patients, 3 parents, 2 staff, and 18 C3N team members and collaborators at design meeting 1; 6 clinicians, 3 patients, 2 parents, 6 staff, and 18 C3N team members and collaborators at design meeting 2; and 7 clinicians, 2 patients, 0 parents, 3 staff, and 12 C3N team members and collaborators at design meeting 3. All work teams that formed had representation from patient or family, clinician, and researchers stakeholder groups.
Interviews with thought leaders provided the following C3N design imperatives:
Design for all stakeholders at once—not separately for patients, clinicians, and researchers.
Technology is only a means to an end. The focus of design must be on enabling people to gracefully achieve their goals, with appropriate technology deployed in service of those goals.
Notwithstanding the above, an upgradeable set of modular technologies is likely to prove to support system evolution.
Design with acknowledgement of health care as a service that is coproduced, not a product to be delivered. This shifts the paradigm from health care as a transaction to health care as shared work [
Design to enable large, diverse groups of people to identify and test many solutions to many problems. No one person and no one solution will transform care, outcomes, and cost.
C3N leaders must empower others to achieve common aims under uncertain conditions. This entails fostering a rapid learning ethos and an embrace of failure in service of learning.
Our environmental scan uncovered 64 people, organizations, products, or services that provided inspiration for parts of the C3N. These included 5 blogs, 10 information clearinghouses, 3 design firms, 4 potential funders, 8 experts or innovators or innovations, 15 networking or community platforms, 3 stakeholder representatives, 4 information technology innovations, and 12 thought leaders. Examples include commons-based peer production models (eg, Linux, Wikipedia, TripAdvisor, Slashdot, and Science Commons); patient communities (eg, PatientsLikeMe, Crohnology, CureTogether.com, and e-patients.net); crowd-sourcing platforms such as Innocentive and Many Eyes; and QI and research collaboratives such as the Northern New England Cardiovascular Disease Study Group and the Children’s Oncology Group.
We also recognized challenges. Unlike Wikipedia or TripAdvisor, medicine has inherent power and knowledge differentials and regulatory and oversight constraints. Developing the right mix of incentives to engender collaborative behavior is challenging, as is attracting individuals to contribute and foster contributions. A free-for-all where everyone’s opinion is equal risks introducing and propagating harmful ideas and suggestions.
Creating a C3N
Purpose:
To enable patients and families, clinicians, and researchers to work together to create a
Vision:
To be healthier together
Patients, families, clinicians, and researchers all have the same goal when it comes to chronic disease—for those affected to be healthy and live gracefully with a condition that they didn’t ask to have
The C3N will enable ImproveCareNow to become a collaborative innovation network—a community with shared purpose, values, tools, and technologies (both human and digital) to enable patients, families, clinicians, and researchers to share responsibility for achieving dramatically better health for children and adolescents with IBD
Values and principles (How we behave in the community and act toward one another):
Hope and compassion (to cause the enthusiasm, curiosity, and the will to solve problems)
Privacy (must be data literate to participate, patients own their data: have rights of possession, use, and disposal)
Trust (individuals demonstrate credibility, information is credible; scholarly norms for attribution; openness)
Shared responsibility for outcomes—we all have the responsibility to improve the health of the entire community (the entire population of patients with IBD)—you can make a difference, and you are expected to
Urgency and daring to create and try new ideas
Creativity and innovation (to generate and test new solutions)
Self-determination (agency)
Descriptions of keywords:
Collaborative—patients or families, clinicians, and researchers engaged as partners in a shared task.
Community—a distributed, voluntary organization that is interdependent, has shared responsibility, and is greater than the sum of its parts.
Shared purpose—improve the health of the entire IBD community
Common values—compassionate, safe, trust (privacy and credibility), open, self-determination, or agency
Flexible set of tools—human (quality improvement, leadership training, motivational interviewing, social networks, and incentives) and information (asynchronous communication, social media, network analysis, data mining, and multimedia) technologies
The goal-directed design method is presented elsewhere [
Including all relevant stakeholders as codesigners enlarged the discussion of relevant outcomes beyond traditional clinical measures. By focusing on people and their goals or needs and by insisting that the system must meet the needs of all people, we were able to arrive at a set of measures that reflected the multistakeholder perspective. The following system performance measures were proposed and approved:
Participation, engagement, and interaction among all types of users as measured by attendance at webinars, monthly calls, and community conferences, as well as contributions of data and ideas.
Health outcomes (eg, steroid-free remission and improved quality of life) as measured by physician global assessment.
Reliability and effectiveness of chronic illness care (eg, more appropriate medication use and disease activity monitoring) as measured by the degree to which a bundle of clinical interventions were delivered as part of clinical care.
Self-management, as measured by self-report of adherence.
Production of new knowledge and discoveries as measured by research products, including grants, abstracts, presentations, and publications.
Representatives from all stakeholder groups cowrote the purpose, vision, values, and principles for the C3N (
Design meeting 2 also resulted in ~140 potential innovation ideas. We deduplicated and combined the ideas into 33 unique potential innovations.
During design meeting 3, a total of 20 intervention concepts were rated as having high impact and high understanding or knowledge. These were interventions that ImproveCareNow was either doing currently or else were sufficiently specified so that no further design or testing was necessary. These should simply be done. There were 13 intervention concepts rated as having high potential impact and low understanding or knowledge about how to implement. These intervention concepts, listed in
Mentoring in inflammatory bowel disease (IBD) clinic
Leadership training
Privacy education
Facebook connector app and community building
Branding ImproveCareNow as a collaborative chronic care network (C3N)
Model care or quality improvement metric explorer
Android device—gateway to C3N
Virtual camp oasis
Self-management support curriculum
Open-source practice Wiki
Patient driven n=1 trial
Restructured IBD education day
Patient interface—virtual C3N
Our codesign process resulted in a community of over 100 people willing and able to self-organize to pursue a shared overall concept design for the ImproveCareNow C3N, a logic model for bringing about this system, and 13 potential innovations likely to increase awareness and agency, make it easier to collect and share information, and to enhance collaboration. Developing and testing these potential innovations to determine the degree to which they could collectively bring about the C3N were the actionable next steps for ImproveCareNow.
It is not intuitive that thousands of people could self-organize and collaborate to achieve shared aims. But we see they do across many industries. In addition to Wikipedia and other examples uncovered in our environmental scan, more recent examples such as AirBnB, Uber, Lyft, and crowd funding sites such as KickStarter and GoFundMe affect the lives of more and more people. The C3N design is a way to translate peer production to health care.
Distributed networks are especially relevant to children with chronic diseases that the NIH identifies as rare diseases [
There is growing awareness that health care is a coproduced service—that professionals and patients create value through collaborative interactions [
We conceived of the C3N as a health care system in which patients (and their families), providers (physicians and other clinicians), and researchers could collaborate, at large scale, to achieve shared aims. By codesigning with representatives of all stakeholder groups, we were able to translate this idea into a design concept, including a set of measures, a logic model, and a set of innovations that could be tested together to achieve the goal of improving care, spawning innovation, and accelerating research. The C3N model challenges the dominant chronic illness care paradigm that views patients as objects on which to intervene, structures care around episodic one-to-one patient-physician interactions, and assumes an inherent power differential based on knowledge. The C3N is designed to engage patients as coequals in care delivery, designing innovations, and research; make learning continuous; and level the knowledge gradient.
We encountered several challenges during this design phase. Because a C3N had never been created before, we did not know what the end product ought to be, and this was frustrating to some stakeholders who wanted to know what the answer was. Over the course of the design process, most stakeholders came to realize that there was no predesigned product and that the point of the codesign process was to come up with this answer. Another challenge was managing expectations of how transformative the changes would be. Some stakeholders were nervous that the design would be too much of a change, whereas others feared the opposite. We regularly introduced the topic of change and attempted to calibrate expectations, in part, by reiterating that the codesign process itself would ensure that the final product was acceptable to the community. A third challenge was the need to translate across stakeholders so that a common perspective and even a common language emerged. Words like
Because a C3N had never been made before, there was no way to know in advance what steps to follow to bring it about. We developed, rather, a collaborative team of more than 100 diverse stakeholders aligned around a common goal and with a common plan for testing our way into this new system. This team was able to identify user needs and generate a sufficient set of novel ideas that could be potentially transformative for ImproveCareNow [
The clinicians and patients with whom we worked are likely to be systematically unrepresentative of the population, having relatively high levels of skills, insights, or resources. This set of conditions risks creation of a design that would work only for these users. Our rationale for this strategy is based on von Hippel’s theory of lead-users [
The generalizability of the C3N design is unknown. Although this design was built for IBD, the noncategorical approach to chronic illness care [
Finally, the absence of formal feedback from our codesign participants limits our ability to understand how acceptable different users found the process.
Our current health care system cannot achieve the results we need. Incrementally improving the current system is not enough, but designing a new system is a daunting task. Our experience suggests that codesigning with representatives from all relevant stakeholders, using the idealized design process, can result in a potentially transformative design for the chronic care delivery system.
collaborative chronic care network
inflammatory bowel disease
key driver diagram
learning health system
National Institutes of Health
quality improvement
This work was funded by the NIH through the NIH Roadmap for Medical Research (R01 DK085713-01). The authors would like to acknowledge the ImproveCareNow Network for their contribution to this work.
None declared.