Top ten unanswered questions in adults living with type 1 diabetes – a UK and Ireland James Lind Alliance Priority Setting Partnership

Across Ireland and the UK, nearly 300,000 adults and children live with type 1 diabetes (T1D), a chronic metabolic condition defined by hyperglycaemia and associated with multiple complications.^1,2,3 While T1D, its complications, and treatment options are the subject of a huge number of studies, there is strong evidence to say that investigators and researchers do not always understand and share the priorities of those living with T1D every day.⁴

To rectify this, the James Lind Alliance (JLA), a non-profit organisation, aims to bring together people living with health issues, their families, carers, and healthcare professions to take a multi-stakeholder approach to determining the most important questions in health research.⁵ This process, called a priority setting partnership (PSP), brings people together as equal partners, firstly to identify questions through public surveys, and then to prioritise those questions.

The final questions are then brought to research funders to ensure that research which is relevant and important to patients is prioritised. In 2011, a UK-only PSP was completed in T1D; however, since that time, the landscape of diabetes management has changed significantly.

Positively though, many of the 2011 top 10 questions have not only been answered, but have become part of standard clinical care for those living with the disease.^6,7,8 The aim of this PSP was to review and refresh the priorities for adults living with T1D across Ireland and the UK, as both offer free healthcare and medications for adults living with T1D, and have access to similar ranges of diabetes technology.

JLA process and results

To complete this PSP, we adhered to established JLA methodology.⁹ Step one involved forming a steering committee and outlining the scope of this PSP. Our steering group was made up of four people living with diabetes and one parent of a child living with diabetes, six clinicians (four endocrinologists, one consultant clinical psychologist, and one senior dietitian), and three members of diabetes charities from across Ireland and the UK. We invited two members of the 2011 PSP to join the steering group to ensure their experience was represented.

In order to ensure that the JLA process was followed, the steering group was overseen by an impartial JLA adviser who chaired all steering group meetings. Key members of the steering group also included an information specialist, who collated survey data and performed evidence checking (discussed below), and the project co-ordinator, who organised steering group meetings, managed communication with the steering group and other stakeholders, and oversaw the survey software. The co-ordinator and information specialist also worked together to target different, under-represented groups during the course of the survey and promoted engagement through social media channels and emails.

The steering group was balanced in terms of gender and had 10 UK- and four Irish-based members. The scope of the PSP included adults living with T1D, their families, and carers, and welcomed all questions about living with the disorder, its complications, prevention, causes, co-morbidities, and treatment, as well as its physical, social, cultural, economic, and psychological impacts.

We excluded questions about paediatric care, questions or priorities about other forms of diabetes including maturity-onset diabetes of the young (MODY), type 2 diabetes and gestational diabetes, T1D in pregnancy (due to recent completion of a PSP in the area of diabetes in pregnancy), and cystic fibrosis-related diabetes. We also excluded questions that did not have an Irish or UK focus.

The second step was to establish the uncertainties. The JLA process has a set pathway for determining uncertainties⁹ which includes designing and disseminating a primary survey that asks adults living with T1D, their families/carers, and healthcare professionals to list up to three questions they would like to see answered about the condition.

We also asked participants to provide some basic demographic details. The survey was available in English only and ran for three months. It was promoted predominately through social media networks, however, targeted interventions included emails to healthcare professionals in endocrinology, nephrology, cardiology, ophthalmology, and general practice throughout Ireland and the UK. Printed posters with QR code links to the survey were placed in diabetes clinics and general practice offices.

To ensure a diverse range of voices were included we also ran a phoneline for two hours each week for individuals with visual impairment or who preferred spoken to written English, emailed organisations working with under-represented patient groups, and contacted professional networks. We also utilised social media platforms (Facebook, X (formerly Twitter), and Instagram) to contact local community champions to promote this survey in different localities.

In the survey, all participants were provided with an explanation of how their data and information would be used and participants gave consent before completing the survey. Next, the information specialist manually reviewed each individual question and ruled them in or out of scope. From the ‘in scope’ questions, indicative summary questions were formed in a PICO (population, intervention, comparison, and outcome) structure.

The steering group reviewed all questions deemed out of scope by the information specialist. Three steering committee sub-groups reviewed each of the indicative summary questions (with the original questions) to minimise individual bias and ensure the summary questions were truly reflective of the questions asked. The steering sub-groups were free to edit the summary questions provided the group was in agreement and all changes were tracked and recorded.

Once the steering group was satisfied with the summary questions, evidence checking commenced. This involved the information specialist checking all summary questions against evidence sources (Ireland and the UK guidelines, PubMed, and the Cochrane database for systematic reviews published from 2020 onwards).

The initial survey received 13,387 views. In total, 1,050 individuals filled at least one question box and the average number of questions asked was 2.9. Respondent demographic details can be found in Table 1. When all text boxes were analysed by the information specialist, there were 3,457 individual questions.

After excluding spoiled or non-sensible boxes, 2,937 questions were individually reviewed and 82 indicative summary questions were formed with questions on cause and aetiology (28 per cent), cure (19 per cent), treatment and prevention (18 per cent), long-term outcomes and complications (11 per cent), questions regarding sexual health and fertility (5 per cent), service access (5 per cent), psychological queries (4 per cent), weight management (4 per cent), assessment and initial diagnosis (2 per cent), the impact of Covid-19 on T1D (2 per cent), and communication and education (1 per cent). Of the 82 indicative summary questions, 17 were answered by recent evidence, leaving 65 questions unanswered.

The fourth step was interim priority setting and this involved ranking the 65 unanswered questions in order of importance. This interim survey was distributed via the same media as before; however, participants of the first survey who gave their consent and contact details were emailed directly to complete the second survey. In this survey, participants were asked to select up to 10 questions which they felt were important for research to answer. This survey ran for a total of eight weeks and received 497 replies. The demographic make-up of respondents closely matched those from survey 1 (Table 1).

While analysing the replies from the interim survey, we reviewed the results from the general population and separately reviewed responses from subgroups, including healthcare professionals, people living with diabetes and their carers, ethnic minorities, and those who identified as being on a lower income.

We also reviewed the answers from male participants as they accounted for only one third of replies. Significant differences were noted, for example, between healthcare professionals and those with lived experience, only five out of 10 priorities overlapped. Similarly, when the replies from ethnic minorities were compared to replies from the general population, only five out of 10 priorities overlapped.

To ensure that the priorities carried to the final workshop adequately reflected the priorities of the entire community, we followed the JLA process of granting priorities ranked highly by underrepresented groups higher scores to ensure they are fairly represented and not diluted by over-representation of a certain cohort (eg, this survey had a large response rate from white female participants).

We then combined the scores from the general population and each of the sub-groups mentioned (healthcare professionals, people living with diabetes and their carers, ethnic minorities, those who identified as being on a lower income, and male participants) to determine the most popular priorities. By giving equal weight to all sub-groups and combining these weights to achieve an average score, we were able to ensure there that was a fair representation from all under-represented groups.

Lastly, we held a final workshop to decide the final top 10 priorities. To ensure a diverse and representative panel, the steering group reviewed the final workshop invitees to ensure ethnicity and clinical and lived experience were balanced. Where deficits in diversity and expertise – either by profession or lived experience – were identified, targeted interventions were undertaken to ensure fair representation.

The final workshop was held virtually with 27 participants (14 healthcare professionals – two dietitians, two advanced diabetes nurse practitioners, two podiatrists, one general practice nurse, five endocrinologists – four adult and one paediatric –, one psychologist, and one mental health nurse who is also trained as a behavioural scientist – and 13 individuals living with T1D and their family members).

A total of 15 members of the final workshop were female (55.5 per cent), 11 were male (40.7 per cent), and one member identified as non-binary. Two members (7.4 per cent) of the final group were from black or minority ethnic groups. Three members of the steering group also sat on the final workshop panel.

The final workshop was led by the JLA adviser and three other JLA moderators were present to facilitate four small group discussions. Prior to the workshop, each participant was asked to rank all 19 questions from 1-19. During the workshop, participants were split into four sub-groups. At the end of the workshop, all four groups’ rankings were combined to create a final 1-19 ranking, with the focus being on the top 10 most important priorities.

The four JLA facilitators oversaw small sub-group discussions over the course of the two days. There were three sessions for small group discussion which were held in virtual ‘break-out rooms’. During these discussions, the facilitators oversaw conversation and debate between groups, and each group was balanced in numbers of healthcare professionals and those with lived experience. Equal input was sought from each member of the group.

During the first sub-group session, each participant identified their highest and lowest priority and gave a brief explanation of why they ranked the question in that way. The JLA facilitator kept a record of how each question was ranked and placed each question in either a ‘high’, ‘middle’, or ‘low’ priority category.

In the second session, the sub-groups remained the same and the group reviewed the questions using visual aids and ranked the questions in the ‘high’ and ‘middle’ categories to create a top 10. At the end of day one, the rankings from each sub-group were combined to create an overall ranking.

At the start of day two, these results were shared with the entire group and new sub-groups were formed, again made up equally of healthcare professionals and those with lived experience to ensure participants heard different perspectives. These groups reviewed the priorities and debated the order of questions to create a top 10 list. Once again, the results of each of the sub-groups were combined to form the final top ranking and top 10.

During the break-out groups, there was discussion on what forms of technology are currently available versus how reliable certain forms of technology – including artificial intelligence – currently are, and how they will progress in future. Other topics of discussion included the importance of ensuring treatment accessibility for all people living with T1D, and if this was research or a policy question. In particular, the differences in priorities between healthcare professionals and those living with diabetes were discussed at length, and careful consideration was given to the final ranking of questions.

The results of the ultimate top 10 priorities from the final workshop and their ranking in the interim survey can be found in Table 2. It is notable that the top 10 questions identified by the interim survey overlapped significantly with the final top 10 priorities (seven out of 10 carried forward). As the workshop was held in a virtual format there was no follow-up discussion on the top 10 in the plenary setting, as this would have given greater influence on more confident participants. Instead, participants were invited to give formal feedback on both the top 10 results and their experience of the process.

Discussion

This PSP, led by the Diabetes Collaborative Clinical Trials Network and the JLA, brought together over one thousand individuals with lived experience of, or working with, T1D to identify the most important unanswered questions in this area. Following evidence checking, 65 summary questions were carried forward to a second-round survey in which a top 19 were selected. A final workshop, held online over two days, facilitated detailed and thoughtful discussions between all stakeholders and resulted in the final top 10 priorities.

During this PSP, we adhered to the JLA’s five-step process (establishing steering group, gathering uncertainties, data process and evidence checking, interim priority setting, and final priority setting). The ultimate aim of this robust, multi-step pathway is to identify key areas for research that can improve outcomes and quality of life for those living with T1D and to work with funders to direct researchers and resources towards these topics.

This was a refresh PSP updating the 10 priorities outlined in 2011⁷ and is one of only four refresh JLA PSPs to date. A refresh PSP is conducted when there has been considerable research developments or changes in clinical practice in an area.¹⁰ The change in diabetes management is evident from the differences in the topics covered by the 2011 and 2024 PSPs and limited overlap. In 2011 there was a focus on insulin pumps, continuous glucose monitors, and closed loop therapy. Most, if not all, of these treatments are now part of routine diabetes care in the UK and Ireland.

Current priorities include the use of artificial intelligence; the use of newer technologies to identify foods and accurately count carbohydrates; and access to therapies including stem cell transplants and immunotherapy.

There was also a new focus on the impact of hormonal changes, menstruation, and menopause on glycaemic management, and the impact of T1D on cognition in older adults, while the impact and the prevention of hypoglycaemia; the psychological burden of living with T1D; and for healthcare professionals, the importance of a person-centred approach during clinical interactions, are of ongoing importance in T1D.

During the final workshop there was enthusiastic debate about the ranking of questions. Different viewpoints were exchanged, and this had a meaningful impact on the ranking of questions. While seven of the final top 10 questions also ranked in the top 10 on the interim survey, three questions moved up from positions 11, 18 and 21 to occupy a top 10 space. These questions focused on diabetes distress, insulin resistance, and women’s health. This reflects the importance of group interactions and meaningful exchange in the JLA process.

There were also some geographical differences in priorities. Questions regarding the use of hybrid closed loop (HCL) therapy were common to participants from both the UK and Ireland. However, while questions from respondents living in Ireland focused on its safety and efficacy in optimising glucose management, multiple questions from UK participants focused on geographical limitations in technology access – the so-called technology ‘postcode lottery’.¹¹

Differences between diabetes care in Ireland and the UK were also illustrated by an announcement from the National Institute of Clinical Excellence’s (NICE) technology appraisal between the initial survey closing and the launch of the interim survey. This approved HCL therapy for any adult with suboptimal diabetes management (ie, a haemoglobin A1c of 58mmol/mol or higher or with disabling hypoglycaemia) despite the use of continuous subcutaneous insulin infusion (CSII) or continuous glucose monitoring.¹²

This led to a challenge during the evidence-checking phase as questions about availability were comprehensively answered by UK but not Irish guidelines. Due to the number of queries received about HCL therapy and the likelihood that challenges will remain in the roll-out of HCL therapy in the UK for some time due to capacity issues, the decision was made to carry this question forward to the interim survey.

We received nearly 300 individual queries on questions like “when will there be a cure”, “what are the barriers to cure”, and “when a cure is available what will be done to make sure it is offered to everyone with T1D”. While accepting that a cure for T1D is the ultimate goal for all individuals impacted by the disease, the decision was made by the steering group to rule these queries ‘out of scope’ as there is currently no strict definition for cure.¹³ For example, some people living with T1D might be satisfied with a once monthly infusion that eliminates the need for insulin delivery and blood glucose monitoring; however, for others this still represents a burden of disease and does not meet the threshold for cure.¹⁴

This process had a number of strengths, including its robust methodology and the efforts made to establish a diverse range of opinions. Despite this, we also identified some limitations. Although the steering group was balanced based on gender, it was exclusively made up of individuals from white Irish and UK backgrounds.

Despite multiple targeted efforts to engage respondents from ethnic minority groups, this population was under-represented at both survey stages. Moreover, the surveys were available in English only, which may have also have limited the range of respondents.¹⁵ Finally, although another media for survey completion was offered, the predominant use of social media and online platforms to promote the survey may have disadvantaged individuals exposed to digital poverty.¹⁶

Conclusion

Further research is needed to respond to and answer adequately the top 10 important questions identified by this JLA refresh PSP. This rigorous process has highlighted ongoing inadequacies in T1D care in the areas of hypoglycaemia prevention and treatment, holistic care and psychological health, and the need to harness advancements in technology, including artificial intelligence, to improve outcomes for people living with T1D. The publishing of this top 10 priority should serve to motivate researchers and direct funders towards the most pressing unanswered questions in type 1 diabetes.