CGM Co-Design Targets a High-Risk Diabetes Population Left Out of Routine Care
Adults living with serious mental illness and type 2 diabetes carry a double burden that routine diabetes services often fail to meet. Schizophrenia and bipolar disorder are linked to a two- to threefold higher risk of type 2 diabetes, and the clinical consequences extend well beyond glycemia. Lower life expectancy, fragmented care, and persistent barriers to self-management create a setting in which standard diabetes education can feel misaligned with daily reality. Continuous glucose monitoring has the potential to make glucose patterns visible in real time, but adoption remains limited in this group, in part because the technology itself does not solve the practical and psychological barriers surrounding its use.
This co-design study did not test glucose outcomes. Instead, it built the scaffolding for a future structured CGM intervention by working directly with service users, carers, and healthcare professionals to define what a usable program should look like. The aim was pragmatic: identify the intervention ingredients, arrange them into a logic model, and create a programme theory that can guide later development and evaluation. In a population where disengagement is often assumed rather than examined, that design choice matters. The work focused on how support must be delivered if CGM is to be more than a device handed to people who already face multiple layers of exclusion.
Workshop Input Shaped the Intervention Around Everyday Barriers
Experience-based co-design sessions brought together 10 participants: four healthcare professionals, four service users, and two carers. Their discussions centered on the practical barriers that make diabetes self-management difficult when mental health symptoms, treatment burden, and social instability intersect. Limited technical support emerged as a major obstacle. Accessibility issues also surfaced, along with the absence of integrated physical and mental healthcare, a gap that can leave glucose management disconnected from the realities of psychiatric care.
These barriers are not abstract. CGM depends on understanding data, responding to trends, and staying engaged with a device over time. For someone coping with fluctuating mood, cognitive burden, or distrust of healthcare systems, the technology can become unusable without guidance that is calm, repeated, and adaptable. The co-design process therefore emphasized not only what the intervention should contain, but how it should be delivered. Participants helped develop candidate components that were then translated into a logic model and programme theory, with the outline refined iteratively across multiple sessions.
A concise signal emerged from the discussions: the intervention must fit the person, not the other way around.
Three Phases Emerged: Set-Up, Early Days, and Living With CGM
The proposed structured intervention was organized into three phases: set-up, early days, and living with CGM. That sequencing reflects the reality that technology adoption is not a single event. It begins with preparation, moves through the fragile period when alarms, data, and wearability may feel overwhelming, and continues into the longer phase in which CGM becomes part of daily diabetes care.
During set-up, participants favored personalized CGM training rather than a generic orientation. Training needs varied, and the intervention was shaped to allow adjustment rather than imposing a fixed script. Flexible delivery methods were also prioritized, which is especially important in a population that may have difficulty attending standard appointments or processing large amounts of information in one sitting. In the early days, ongoing support was viewed as essential. That support is likely to be the difference between a device that informs behavior and a device that gets abandoned after an initial trial period.
The living-with-CGM phase captured something equally important: long-term use depends on sustained reassurance, troubleshooting, and peer contact. Participants identified peer networks as a key component, suggesting that shared experience may help normalize the learning curve and reduce isolation. For people with serious mental illness, peer support may also reduce the sense that diabetes technology belongs only to a narrow, more stable patient group. Embedding peers into the intervention design acknowledges that motivation is not the only issue; confidence, trust, and habit formation are central too.
Why Integrated Physical and Mental Health Support Matters Here
The study repeatedly returned to one central theme: CGM will not work well if physical health care is isolated from mental health care. Adults with serious mental illness often encounter services that are split by discipline, with separate pathways, different priorities, and limited coordination. Diabetes treatment, however, is cumulative and behavioral. Monitoring, interpreting, and acting on glucose data all require continuity. If psychiatric symptoms, medication effects, or social stressors are ignored, even a well-designed CGM rollout can fail at the bedside level.
By designing the intervention around both physical and mental health needs, the study took a more realistic view of adherence. The practical value of CGM in this population lies not just in identifying hyperglycemia or hypoglycemia earlier, but in creating a feedback loop that can help users and clinicians see how eating patterns, medication changes, stress, and sleep may influence glucose variability. That kind of visibility can be useful only when support is available to interpret the information without blame or overload.
Feasibility Was Demonstrated, but Clinical Testing Still Lies Ahead
The strongest contribution of this work is methodological. It shows that people who are often labeled “hard to reach” can, in fact, contribute meaningfully to intervention design when the process is flexible and respectful. Service users and carers did not simply endorse a prewritten plan; they shaped the intervention content and delivery model. Their input helped establish a structured framework that can now be carried forward into future co-production and formal testing.
What this study does not yet answer is whether the resulting intervention improves glycemic control, patient experience, engagement, or cost-effectiveness. Those endpoints will require a different study design and a larger, more diverse population. The current work instead establishes the logic for future evaluation: if CGM is to benefit adults with serious mental illness and type 2 diabetes, it will likely need personalized education, accessible delivery, continued support, and a care structure that treats mental and physical health as inseparable parts of the same clinical problem.
The intervention is still a draft, but the design principles are already clear: flexibility, repetition, peer support, and joined-up care will determine whether CGM can move from promising technology to usable diabetes support in this vulnerable group.