Abstract
While it is widely acknowledged that co-design will enhance project outcomes, including for digital health innovation, the definition and application of co-design remain heterogenous. Efforts to systematise co-design as a meaningful and rigorous methodology, have been made over the past decades using co-design frameworks. However, we find that co-design frameworks present challenges when operational contexts are complex, and research problems nebulous, as is often the case in digital health research and development initiatives. Here, we present our experience of co-designing a self-guided supportive care digital health platform for Australians affected by brain tumours, called ‘Brain Tumours Online’. The Brain Tumours Online platform seeks to mitigate unmet supportive care needs of patients and their carers through a three-pronged approach: a repository of evidence-based information, an online peer support community, and a directory of validated digital therapeutic tools. Our co-design approach for this platform was influenced by three key contextual considerations: a) disparate operating models and competing priorities of partner organisations; b) patient population-specific needs; and c) the need to practice epistemic flexibility and adaptation to evolving project needs. Instead of following a standardised co-design framework, we adopted methodological pluralism, with a bespoke multi-modal co-design approach. This approach allowed us to combine strengths of various stakeholders and mitigate organisational barriers of working across sectors, characteristic of digital health initiatives in healthcare.
Layperson summary
In the realm of digital health innovation, co-design is recognized as crucial for achieving successful outcomes. Co-design refers to a participatory method whereby stakeholders, such as patients and carers, collaborate with design professionals to design solutions for problems. In the past, applying co-design methods has proved to be challenging, as there exist many different interpretations and contexts. In this paper, we reflect on the complexities we experienced when co-designing a supportive care digital health platform for Australians affected by brain tumours. Challenges encountered during this study included differing organizational priorities among partners, specific needs of patients, and the necessity for adaptable methods that respond to evolving project demands. To address these challenges, we used a variety of methods instead of adhering to a single systematized co-design framework. Our experience highlights how incorporating adaptability and engaging with different stakeholders in flexible ways can lead to better outcomes for digital health projects. By sharing what we have learned, we hope to encourage more flexible and collaborative approaches in digital health innovation, which can make treatments and tools more effective and useful for everyone involved.
Keywords
Co-design: history and heterogeneity
Co-design has its roots in Scandinavian participatory design practices. Democratic decision-making and distribution of power were key tenets of this paradigm 1 which was originally developed to give workers a voice in the development of computerised workplace systems, with the goal of improving overall system quality. 2 While a variety of approaches and methods are now adopted under the umbrella term of ‘co-design’, 2 they typically entail “designers and people not trained in design working together in the design development process”. 3 Co-design is increasingly recognised by health services, policymakers, funding bodies, and researchers, for its potential to improve the usability, usefulness, and legitimacy of products and services. 4
‘Co-design’ has often been used inter-changeably with terms such as ‘co-creation’ and ‘co-production’, all of which may fall under the umbrella of the Participatory Action Research (PAR) paradigm. 5 Each of these terms has similarities and broadly represent aspirations of researchers and professionals to meaningfully involve users across a project continuum. Nevertheless, various schools of thought exist in relation to the distinct properties of each of these concepts, depending on the disciplinary area of application. For example, within a public health context, Vargas et al. 5 posit that co-design and co-production are subsets of the overarching guiding principle of co-creation. Co-design is often considered as the active participatory collaboration between stakeholders to design solutions for a predetermined problem. Co-production in public health may be more focused on the implementation and collaborative delivery of the agreed solution. Co-creation represents sum of co-design and co-production, and is concerned with active involvement of stakeholders from initial problem definition to design, development, and implementation of solutions. 5 For the purposes of this paper, we focus on our application of the notion of co-design.
Sixty commonly cited definitions for co-design were identified by Masterson et al. in a recent scoping review exploring participatory concept definitions in health and social care. 6 Even as existing definitions have continued to evolve, the rate of new proposed definitions has increased. 6 Responding to this lack of consensus, Masterson et al. speculated that returning focus to the successful translation into practice of the values underpinning co-design (such as power distribution, mutuality) might be more useful than continuing to quest for universal or standard definitions. Indeed, one may also argue that pursuit of co-design guided primarily by varying definitions would have implications for rigour in the practice of co-design. Thus, as Masterson et al. note, there is greater translational value in operationalising the underlying principles of co-design.
Uncertainty that standardisation should be a principal goal of co-design methods research is also apparent in recent reviews of frameworks intended to systematise involvement approaches. Green et al. found numerous variations in the application of experience-based co-design (EBCD), one of the most commonly cited co-design frameworks in healthcare, with studies often eliminating or altering certain EBCD process phases. 7 More explicitly, in a systematic review of PPI frameworks, Greenhalgh et al. identified 65 different frameworks with a variety of intended objectives, including, but not limited to: a) surfacing and addressing imbalanced power dynamics; b) setting priorities and scoping research questions; c) involving lay people in conducting research activities; d) developing demonstratable and auditable governance structures for academic-public partnerships. 4 Greenhalgh et al. observe that, partly due to the combination and operationalisation of diverse objectives, these frameworks often have limited transferability to contexts other than for which they were designed. They ultimately conclude that, “A single, off-the-shelf framework may be less useful than a menu of evidence-based resources which stakeholders can use to co-design their own [project-specific] frameworks.” Similarly, Singh et al. discuss the importance of ‘context in co-design’. 8 They posit that real-world problem-solving requires consideration of the complex interlinking socio-political, cultural, behavioural and other contextual factors that influence health problems. Thus, pursuit of standardised approaches to co-design may impede successful context-informed co-design of sustainable and impactful solutions.
Recent calls to reconsider approaches to co-design, whether through greater values orientation
6
or ‘build your own framework’ methods,
4
introduce the idea that
Here, we present insights gained during development of a bespoke, multi-modal co-design approach to drive development of a digital health supportive care platform for patients with brain tumours and their carers. We identify salient contextual considerations and describe how these influenced the co-design process that followed. We consider how the issues we identify might operate as inputs into emerging ‘build your own framework’ approaches, juxtaposing these considerations with those suggested by others.
Contextual considerations in digital health
Digital health solutions hold great promise for improving patient care and patient activation. Self-managed or self-guided digital health solutions can bridge gaps in care, reaching patients and carers who face access barriers for supportive care due to geography, disability, stigma, or social deprivation. A scoping review of the role and impact of digital health solutions in oncology supportive care identified myriad potential benefits, such as improved symptom management, reduced distress, decreased unplanned hospitalisations and associated care-related costs, and improved survival and overall quality of life. 9 People with brain tumours, in particular, experience debilitating symptoms, shortened life expectancy and high unmet supportive care needs. Previous studies have shown that alongside medical treatment, digital technology can be used to support symptom self-management for patients with cancer. 10 Our consortium of stakeholders including consumer advocates, individual health professionals (from Royal Melbourne Hospital, Peter MacCallum Cancer Centre, St Vincent's Hospital Melbourne, and Barwon Health), universities (The University of Melbourne, University of Technology Sydney), and IT development company DEPT (formerly ‘Two Bulls’), came together to co-design and develop a self-guided supportive care digital health platform for Australians affected by brain tumours.
The value of co-design and PPI approaches for digital health innovation projects is widely recognised. 11 However, digital health projects present unique challenges and opportunities for meaningful co-design. Given the diversity of patient circumstances, selected technologies need to be flexible, and where possible, personalised to address individual needs. 12 Without inclusion of consumers in their design, there is risk that digital solutions do not reach those with greatest need or those most disadvantaged. 13 Successful, evidence driven co-design is characterised by synthesis of many perspectives across the industry-academia-health service-consumer continuum 14 and is essential if translation of new interventions into clinical practice is to be accelerated, equitable and impactful. 15 Systematised co-design frameworks offer one way of achieving these goals. However, as outlined below, our project's operational context and associated constraints made application of a systematised co-design framework challenging, and warranted original solutions to devise a meaningful, but adaptive co-design approach.
Disparate operating models and competing priorities of partner organisations
Digital health projects are frequently run by multi-disciplinary teams, which on the one hand can enhance project outcomes, but on the other pose administrative and organisational challenges. 16 Our project was funded under the Medical Research Future Fund (MRFF) program, a $20 billion investment fund that aims to transform healthcare research and innovation and improve health system sustainability. Our project budget was AU$2.6 million with another AU$1.2 million in in-kind support. Under the MRFF funding governance requirements, the project was subject to stringent delivery timeframes and reporting milestones decided at the grant planning stage. Furthermore, there was tension between lead times for seeking Human Research Ethics Committee and institutional Research Governance approval, and the software development partner's operational model which assumed that development could commence early in the project. As a result, software developers were allocated to our project and ready to build, while the academic partners were still waiting for clearance to conduct researcher-led co-design interviews and workshops.
Patient population-specific needs
Patients with brain tumours often face complex symptomatology with varying needs. Furthermore, poor prognoses and rapid changes in illness trajectory can translate into high attrition rates among research participants. Additionally, physical and cognitive ability to participate may vary, with fatigue and cognitive dysfunction prominent symptoms.17,18 For this patient population, a linear co-design framework, which assumes consistent participation from the same set of patients over an extended period of time, was not feasible. Instead, we needed an approach that permitted flexible patient involvement through a variety of engagement avenues, and that could elicit a diversity of perspectives, with options to participate in one-off activities.
Epistemic flexibility and adaptation to emerging project needs
As the project progressed, so the research questions evolved. Co-design activities required responsiveness to be commensurate with the (changing) question at hand and accommodate epistemic flexibility. For example, initial co-design activities necessitated exploration of systemic healthcare challenges and interpretation of the participant lived experience. As the project progressed, the methods needed to be more applied in nature, to inform pragmatic design decisions such as users’ preferences for platform aesthetics; colour schemes, fonts, pictures, and the platform name.
Our multi-modal co-design solution
We sought to create dynamic and locally adaptive organisational partnerships, which could support an evolving collaboration between a wide range of stakeholders contributing equitably to value-creating practices over the course of the project. Our solution was to create a gamut of activities with varying degrees of participant demand, fit-for-purpose relative to the project needs at any given time and impending design decisions. Furthermore, different project partners had varied needs for end user input, interpreted against their own tacit subject-matter knowledge. Thus, the various partner organisations had the opportunity to facilitate individual co-design activities, with inputs from other stakeholders, as evidenced in Table 1.
Overview of our bespoke multi-modal co-design approach.
It should, however, be noted that the overarching co-design program also required a dedicated point of coordination. A full-time project manager was employed under the project to help manage invisible administration tasks, such as coordinating co-design activities, liaising with consumer experts, and ensuring that the relevant stakeholders had the necessary information and support they needed to execute their own co-design activities.
Ethical approval to conduct this research was obtained from The Royal Melbourne Hospital Human Research Ethics Committee (HREC Reference Number: HREC/77238/MH-2021). Formal written consent was obtained from research participants.
The emergent platform
Outputs from the co-design processes guided the development of a web-based digital health solution, and defined the short-, medium-, and long-term outcomes for the project. Our platform, now known as “Brain Tumours Online” (a name chosen by our co-design workshop participants), seeks to mitigate unmet supportive care needs through a three-pronged approach:
LEARN: A curated database of evidence-based information pertaining to subjects ranging from (but not limited to) symptom management, emotional, financial, transport and accommodation needs, as well as existing and new treatments and research. CONNECT: An online peer support community that allows patients, carers, and health professionals to connect with each other, ask questions, share their stories on discussion forums and take part in online webinars. TOOLBOX: A directory of validated digital symptom self-management interventions for patients and carers.
Discussion
In this paper, we have presented the various constraints which made the application of a stringent ‘off-the-shelf’ co-design framework challenging. Our co-design approach was influenced by three key contextual considerations. First, we had to accommodate operational constraints that affected the ability of different stakeholders, such as the software development partner, to contribute to co-design (or consume its outputs). Second, the approach required flexibility and sensitivity to the constraint that not all potential co-design participants would be able to contribute across the lifespan of the project, or to the same extent. One patient consumer expert who had made significant contributions to the project passed away during the course of the project. This patient's involvement as part of the DRG was invaluable and certainly left a lasting impact on our project team. Similarly, other expert consumers who had been involved in varying degrees also experienced rapid changes in their illness prognoses which impacted their ability to be involved in the project over an extended period of time.
Finally, we observed that a commitment to meaningfully shape the design of the platform
While we did not adopt the Greenhalgh et al., 2019 ‘build your own framework’ framework template, 4 our co-design approach was principally compatible with their proposition to utilize a menu of bespoke, fit-for-purpose, locally-suitable co-design activities. We developed a range of involvement activities, each selected in response to contextual constraints, such as the feasibility of sustained involvement for many patients. Simultaneously, we sought to cultivate a stakeholder group who deeply understood all aspects of our platform, and the layered decisions and compromises inherent within digital health development. In co-designing a digital solution with patients with brain tumours and their carers, we had to be sensitive to their unique needs, possibilities of rapid changes in prognoses and resulting attrition from co-design participation. Furthermore, with our desire to build a comprehensive platform that could meet unmet needs of patients and carers in a variety of life circumstances, we required consideration of personalisation features in the proposed platform. Our methodologically pluralist approach served to mitigate inequity issues in our co-design approach, maximise participant involvement, and minimise participant burden. Thus, based on our experience, we recommend methodological pluralism to other digital health practitioners who wish to co-design solutions with vulnerable communities where continuity of participation from a single set of end-users throughout the project lifecycle may be infeasible, and indeed in certain cases, even unethical.
Recent literature identifies similar contextual considerations which should shape the ‘co-design of co-design’: the underpinning values of the stakeholder group, past involvement experience—whether successful or otherwise—concerning practical set-up, stakeholder preferences for involvement, and the anticipated scope of research activities which place effective bounds on
Our findings echo the importance of deliberate and context-sensitive bespoke co-design processes that reflect logistical realities. Our experience highlights specific pragmatic considerations such as funding and contracting constraints that should be factored in, as relevant concerns in discussions, when setting up co-design activities. Our findings also highlight the importance of considering potential co-design strategies through multiple lenses. For example, directly elicited stakeholder preferences for involvement can be augmented by asking whether and how condition trajectory might affect potential for participation in the longer-term. Our results also highlight the possibility that initial inputs, such as research scope, will continue to evolve. The need for an ongoing process of adaptation suggests that ‘building your own framework’ for co-design may not be a linear one-shot process. It also suggests a specific responsibility for co-design governance groups to support active review and refinement of the selected approaches.
Digital health projects are almost always multi-disciplinary. Our project team comprised a range of professionals from neurosurgeons to sociologists to software developers, with each disciplinary expert contributing to a different aspect of the proposed solution's design and features, and thus having different information needs for end-user input. Therefore, allowing project partners the opportunity to curate bespoke co-design activities was critical for enabling meaningful, fit-for-purpose, and democratic engagement with informants. Setting up a series of working groups that have distinct and clearly articulated functions and responsibilities may provide digital health practitioners a useful avenue for meaningful reflection and facilitate in-depth exploration of sensitive subjects in a safe environment. Notably, Greenhalgh et al. similarly advocate for balancing democratic values with good governance and continual organisational learning. 4 Our fortnightly DRG meetings served the aforementioned functions of enabling continuity and governance. Additionally, in a multi-disciplinary digital health project team, knowledge-sharing and sense-making of inter-disciplinary know-how is key. The DRG meetings also provided a forum for the above purpose. Thus, digital health projects may benefit from multi-perspective co-design opportunities with a suitable linking forum to ensure continuity of project knowledge and good governance to achieve the desired project outcomes.
When presenting our contextual challenges, we previously noted the need to balance our project partners’ disparate operating models, financial constraints, and the need to practice epistemic flexibility in our co-design approach. Based on our experience of co-designing this digital health solution, we posit that a pragmatist epistemological position may sometimes be needed for digital health projects. Digital health practitioners may need to steer clear of metaphysical debates about the nature of reality, in favour of seeking to crystallise insights from multiple perspectives to generate actionable advice for real-world translatable digital health solutions. In particular, based on our experience, transparency about financial incentives and operating models helped to appropriately address emerging tensions and safeguard project progress. The uncertainty inherent in co-designed projects is further compounded in digital health where product specifications are not known at the start, and thus solution development cost cannot be accurately predicted upfront at contract signing stages. In turn, this may have implications for contractual arrangements. Primary research investigators may opt to hire software developers on milestone-based or time-and-materials payment schedules, rather than through lump-sum fees tied to fixed deliverables—especially as project requirements are likely to evolve in response to co-design findings. This proverbial need to “fly the plane while building it” should thus be acknowledged, even embraced in digital health projects in favour of creative and collaborative problem solving towards shared project goals.
Challenges and limitations
Each of our co-design methods came with their unique set of challenges and limitations. First, as the project began during the pandemic, the project team members primarily collaborated virtually and did not have a chance to meet in-person until almost 18 months after project start. Limited opportunities for in-person collaboration may impact team rapport and efficiency. Furthermore, our co-design activities included data collection from multiple sites. There were also significant delays in obtaining the relevant ethical and governance approvals across the participating health services, which impacted timelines for subsequent project milestones.
We conducted our one-on-one interviews via telephone to minimise participant burden. However, this meant that opportunities for reading of facial cues and other non-verbal communication were limited, compared to in-person or video interviews. The co-design workshops and usability testing activities were also conducted via Zoom. While the research team made themselves available to participants who may need extra technological support to participate, digital literacy may still have been a barrier for some people who may have otherwise wished to participate in the co-design activities.
As part of our Design Reference Group (DRG) forum, consumer experts participated regularly. However, the DRG meetings had a significantly higher number of academic researchers and healthcare professionals than consumer experts, which would potentially lead to power asymmetries. Thus, some consumer representatives may have felt hesitant about sharing their views freely in front of other stakeholders who they perceived as being ‘the experts’. Separately, across all the co-design activities, while we sought to engage participants from a variety of socio-demographic groups including culturally and linguistically diverse patients and carers, a vast majority of the participants who volunteered were English-speaking metropolitan residents. The co-design and implementation of digital health solutions generally present complexities such as resource limitations, meaningful incorporation of consumer preferences, and problem-system fit. We also discuss such implementation challenges associated with our project in a separate paper by Schadewaldt et al. 20 Furthermore, there are indeed distinct challenges associated with the evaluation and impact assessment of digital health solutions such as ours; these issues will be the subject of a future commentary.
Conclusion
Through a bespoke multi-modal co-design approach, our team was able to combine strengths of various stakeholders and mitigate organizational barriers to working across sectors, characteristic of digital health initiatives in healthcare. In the absence of an off-the-shelf framework, amid a variety of logistical and other project constraints, our team was able to creatively problem-solve and seek meaningful, translatable insights from a vulnerable end user community in a timely manner to inform relevant decisions at appropriate times in our project lifecycle. Importantly, our process of ‘co-design of co-design’ was influenced by considerations not discussed elsewhere relating to the underlying condition and (digital health) intervention type. Future work should continue to evaluate how these, and related themes, can be applied most profitably within ‘build your own framework’ approaches to co-design in digital health and elsewhere. The synergies among the research partners’ academic rigour, the software developer's technical expertise, clinicians’ domain knowledge, and consumers’ perspectives on priority needs provided a fertile ground for germination of the Brain Tumours Online supportive care platform.
Footnotes
Acknowledgements
We thank consumer experts Diana Andrew, Julia Fin, Samantha Wright, and Janet Micaleff for providing their expertise and perspectives to inform this work. We would also like to thank the members of our Design Reference Group and the other lived experience experts who participated in our co-design activities. The project described in this paper is supported by the Digital Health Validitron, a collaborative and interdisciplinary research group that assists digital health innovators from healthcare, academia, and industry to accelerate the creation of evidence that proves the real-world value of their ideas and products.
Author contributions
Mahima Kalla: Conceptualization, Methodology, Formal Analysis, Writing—Original Draft, Writing—Review And Editing, Project Administration
Kit Huckvale: Conceptualization, Methodology, Writing—Original Draft, Writing—Review and Editing
Ashleigh Bradford: Formal Analysis, Data Curation, Writing—Review and Editing, Project Administration
Verena Schadewaldt: Funding Acquisition, Conceptualization, Methodology, Formal Analysis, Writing—Review and Editing
Sarah Bray: Writing—Review and Editing, Project Administration
Ann Borda: Conceptualization, Methodology, Writing—Review and Editing
Kara Burns: Writing—Review And Editing
Heidi Mcalpine: Writing—Review and Editing
Joseph Thomas: Writing—Review and Editing
Daniel Capurro: Writing—Review and Editing
Richard De Abreu Lourenco: Funding Acquisition, Writing—Review and Editing
Sarah Cain: Writing—Review and Editing
Wendy Chapman: Funding Acquisition, Writing—Review and Editing
James Whittle: Funding Acquisition, Writing—Review and Editing
Kate Drummond: Funding Acquisition, Writing—Review and Editing, Supervision
Mei Krishnasamy: Conceptualization, Methodology, Funding Acquisition, Writing—Review and Editing, Supervision
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research received grant funding from the Australian Government through the Medical Research Future Fund 2020 Brain Cancer Survivorship Grant Program (MRFBCII000014).
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.