1. Background

The burden of living with long-term diseases such as Type 2 diabetes mellitus (T2DM) and its complications is growing dramatically worldwide [1, 2]. Owing to its heterogeneity and multifactorial nature, a complex continuum of care, including diet management, physical activity, self-monitoring and medication adherence, is needed [3, 4]. It requires individuals to take an active and ongoing role in their self-care [5, 6] and make behavioural changes that allow them to improve the management of their condition. Additionally, they must develop health literacy skills to effectively comprehend written and spoken health information, use digital tools and health technologies and navigate health systems to access necessary services [7]. Emphasised as having a major role in improving self-management, adherence to therapeutic regimens and reduced complications, behaviour change theories such as the Behaviour Change Wheel (BCW) framework are used in the health intervention design and implementation [8–10]. This study adopts the core COM-B model of the BCW, which facilitates the development of health interventions aimed at behaviour change, addressing key factors such as capability, motivation and opportunity that can hinder behaviour change [10]. It also incorporates, as a resource, behaviour change techniques that were used as support tools for informed decision-making and to encourage people to adopt new habits [11].

The numerous challenges imposed by T2DM can lead to feelings of overload, stress, anxiety or even depression [12]. Although recognised as a public health concern, efforts in recent decades to reduce its incidence, prevalence and impact on quality of life and well-being have met with limited success. The World Health Organisation (WHO) accentuates key points in the prevention and promotion of T2DM management, through the conception and development of integrated health interventions in primary healthcare with the collaborative involvement of stakeholders and people living with diabetes [13].

Social prescribing (SP) has been highlighted for its role in an interconnected response in preventing and managing noncommunicable diseases, such as T2DM, through activities developed by the voluntary and community sectors [14–16]. Among the different definitions, SP can be defined as a complex intervention that, through community referrals between health professionals and social prescribers, links patients to different nonclinical activities in the community that aim to improve health and well-being outcomes [17]. According to previous studies, from a holistic perspective, the individual needs of T2DM patients are positively related to quality of life, mental and social well-being (depression, anxiety and loneliness), metabolic control (based on HbA1C levels), a reduction in sedentary lifestyles and cardiovascular risk by reducing weight, abdominal circumference and cholesterol levels [16, 18–24]. Broadly implemented across different countries but with several pathways between contexts and stakeholders, heterogeneous intervention components and diverse-related outcomes exist [25–27]. To address the following research question—which is the most suitable model and multicomponent of a complex SP intervention to enhance self-care behaviours and health literacy in patients with T2DM?—a study was conducted using two consensus techniques to identify and refine the optimal SP model and its multicomponents.

Structured and systematised consensus techniques, such as the nominal group technique (NGT) and Delphi panels, are essential for bridging the gap between research findings and practical application in healthcare, ensuring that evidence-based insights are effectively integrated into clinical practice [28]. They are the most widely used consensus techniques in medical and health research [29, 30]. Traditional consensus techniques have been modified to optimise their structure, such as the number of rounds or delivery modes, making them more flexible, dynamic and suited to the needs of both the research and participant groups [31]. Whereas the NGT typically begins with a brainstorming session where participants share their ideas on a theme, in this study, a systematic literature review was carried out as an initial step to identify SP models within the context of T2DM [26]. A modified NGT used in the discussion of the model, provided an open, structured forum for discussion, encouraging each participant to reflect on and share their perspectives. It enabled prioritisation of key elements, fostering agreement on complex aspects of the model [31]. The Delphi panel originally involved in-person meetings and began with open-ended questions, evolving into a structured, iterative process with multiple rounds and controlled feedback [28, 29]. A modified online Delphi panel was conducted, involving rounds of questionnaires and controlled feedback to discuss and analyse the design of the proposed intervention components based on the literature review [26] and an advisory group. This modified Delphi panel was chosen to provide a flexible but rigorous framework for collecting expert input, ensuring balanced contributions, minimising the influence of dominant voices on the expert panel and supporting a more equitable consensus-building process [29, 30]. The modified Delphi panel facilitated an organised systematic approach, allowing experts to prioritise intervention components, provide feedback and reach a supported consensus on the multicomponent intervention [32, 33]. However, this consensus technique was not chosen to explore the more suitable SP model because it is less interactive due to its asynchronous, questionnaire-based structure, losing the opportunity to explore nuances or clarify points in real time, which can limit the depth and fluidity of responses [32]. Using structured methods collects the expert’s opinions and systematically organises the individual ideas and comments that emerge in a group [30]. Composed of heterogeneous expert groups during consensus building, they can capture different backgrounds, experiences, needs, concerns and perspectives. By leveraging the strengths of both consensus techniques, we aim to develop an SP model, and a multicomponent intervention grounded in expert consensus.

2. Methods

Due to the limited information available, an initial systematic literature review on the effectiveness of SP interventions in achieving health outcomes for T2DM patients was conducted. This review also helped identify the existing SP models, as well as the providers and facilitators involved [26]. A descriptive cross-sectional study was subsequently carried out to identify the specific self-care needs of the target population [34]. Based on evidence from the literature and the identification of T2DM individual needs, the research team developed a preliminary design of the SP model between primary healthcare and the community and the SP intervention multicomponents in T2DM patients. Following a person-centred approach and recognising the importance of early involvement of the person and communities, a diverse advisory group was engaged to discuss face-to-face and coproduce the future SP T2DM intervention collaboratively [35]. The contents of the different intervention categories were based on evidence from the literature and international recommendations for self-care in T2DM [36–38]. Once the first draft of the model and components had been planned, we used a modified NGT and a modified Delphi panel for four months developed remotely, between June and September 2023 (Figure 1). The NGT was used as an initial step to facilitate open online discussion among experts, allowing for collaborative brainstorming and prioritisation of key elements in the model. Following the NGT, a modified Delphi panel with a different sample of NGT was conducted to refine the specific components of the intervention. The experts that joined consensus groups were characterised by their heterogeneity, knowledge in the discussion areas or context, and contributions from different perspectives (as users and or providers) [39].

3. Study Design

4. Data Analysis

The quantitative data from both techniques were analysed via the score obtained in the vote by the participants, and the mean, standard deviation (SD) and percentage were calculated using the JAMOVi software package 2.3.26 a free and open statistical software designed for analysing descriptive and inferential statistical data. During the voting process in the modified NGT, the poll was anonymous, and each expert did not have access to the vote of another expert participating in the meeting or questionnaire. The expert’s comments and responses were introduced and discussed during the first and second meetings of the modified NGT. The qualitative data were transcribed from the NGT meetings and, in the Delphi Panel, extracted from Google Forms® into an Excel spreadsheet. The data were analysed following the steps of deductive thematic analysis [47, 48]. The analysis was carried out independently by two members of the research team who coded the opinions of each participant according to the themes previously coded in the modified NGT (Role of the diabetes specialist nurse, Role of the social prescriber, Role of the community intervener, Prescription of the HP intervention and Articulation process) and the categories of the modified Delphi panel (cross-cutting components of the intervention, physical activity, nutrition, medication management, self-control and well-being). A third research team member reviewed the coding carried out by the other two research members and conducted the final analysis. Coding the qualitative data made it possible to analyse the experts’ perspectives and incorporate the intervention functions of the Behaviour Change Techniques [11] into the modelling of pre-established interventions or the creation of new ones. In this way, the integrated analysis of quantitative and qualitative methods aimed to ensure that each component was closely aligned with the established objectives and followed the experts’ perspectives. The endorsement or rejection of a statement was defined as a score on a Likert scale above 3.

5. Results

6. Delphi Panel

The first phase comprised quantitative data analysis and the second phase qualitative analysis of the comments corresponding to each item. Individually, every comment was analysed and considered for integration or modelling of the items listed under study and subject to expert evaluation in the next round. After round one and the comment analysis, before round two, some points were clarified with the experts through feedback on the results of the first round.

7. Discussion

Consensus techniques, such as the NGT and the Delphi panel, are important in comprehensive design and effective health programs, as they bring together different perspectives and generate agreement on complex issues [29, 30]. This study thus aims to synthesise the results of expert consensus by integrating the two consensus techniques, which allowed for a deeper understanding and greater refinement of the SP intervention model and its multicomponents. The decision to use both methods was based on the essential qualities that each offers which facilitate an in-depth exploration and analysis of the main points of discussion under study. Through their knowledge and experience, the participants contributed significantly to reaching a group consensus coproducing an SP intervention to promote self-care behaviours, empowerment and autonomy, commitment to health goals and informed decision-making. A person-centred SP model development is essential to ensure that health interventions align with what matters to people. At the same time, it proposes to improve health literacy, encouraging each patient to take a more active role in promoting their health and well-being. Improving health literacy through the intervention of SPs can help them better understand their condition, adopt healthy practices and use health services more efficiently. Through networking with voluntary and community sectors, the model allows participants to adapt and build a path to a healthier and more balanced life, aligned with their values and needs. It also provides time for the person to reflect, ask questions and be actively involved in their health journey. Living with a chronic and complex disease such as T2DM requires an approach aimed at improving the person’s involvement in their daily self-management. We are aware of the burden of T2DM and the different self-care behaviours that are required. Due to its complexity, this SP intervention has several multicomponents that make it up, which can lead to overburdening. However, we would like to emphasise that each person will define their objectives in a personalised and independent way, deciding which areas they wish to improve.

A clear structure with defined roles for primary care professionals, social prescribers and community partners improves the cohesion and effectiveness of the intervention. This SP approach ensures that people’s needs are accurately identified and addressed, with input from different health and community stakeholders [41]. Previous research suggests that limiting the number of clinical and nonclinical professionals involved in the SP intervention can avoid role confusion [50]. Thus, resulting from the consensus this model adopts an optimised and simplified process flow structure with clearly defined responsibilities aligned with the objectives of each participant. Using a flowchart for the model facilitates the understanding of the steps of a process, their sequence, responsible parties, scope (start and end points) and decision points with their possible outcomes [51]. Due to its simple and organized presentation, it can be thought about the transferability to different areas or in areas inside diabetes mellitus. In this model, the nurse plays the role of social prescriber by referring patients to community resources, assessing health needs and addressing the psychosocial factors that influence self-care and diabetes management. As in other countries, nurses have been chosen as social prescribers because of their proximity, ability to identify individual needs and ease of setting goals with the person [52]. Nurses are key to connecting patients with community activities, reinforcing the success of the intervention by aligning behaviour change plans with the motivations of each patient [53]. This model allows for integration into the national public health system, enabling it to be used by primary healthcare professionals and recorded in the patient’s medical record, allowing for greater reliability and evaluation of results, supporting the development and impact of the intervention. As an evolution of a deeper understanding of the model, participants gained a deeper comprehension of each stage of the process, allowing for continuous evaluation and refinement. Ultimately, this iterative approach led to a consensus among the experts, achieving an agreement rate of 93.52%.

Furthermore, this study reached a consensus on a set of multicomponent interventions that address multiple needs of T2DM patients and domains of self-care behaviours, providing a holistic and customisable intervention. Composed of six categories covering different areas such as physical activity, nutrition, medication management, self-monitoring and well-being, the expert panel mostly received the interventions with a consensus rate of 88.9%. Participants noticed the relevance of integrating different areas the person can develop according to their preferences and specific needs. Except for one intervention related to religion in the last category, this intervention was mostly rejected by the person with T2DM and some of the professionals involved. The research team therefore opted to remove it. However, the literature shows that religion and spirituality can play important roles in coping with the disease process and influencing self-care behaviours in the person with T2DM [54, 55]. It is sometimes reported that professionals find it difficult to approach religious activities and consider that they do not respond to the person’s needs in managing T2DM [55].

A preliminary mapping of voluntary and community sector resources was important, as it helped to understand the offers made by these resources, to establish channels of communication between contexts and to co-design the intervention. Each domain of the multicomponent SP intervention was aligned to support the specific self-care behaviours of this model, supported using behaviour change techniques [11], which included strategies such as encouragement, support, motivation and resources to support the interventions developed. The use of resources is considered important in involving the person in the decision-making process, improving health behaviours and self-care management [44]. The information provided through resources should involve verbal, written or digital forms [44]. Among the cross-cutting components, the first category, different approaches have been combined, such as SMS messages, face-to-face and telephone follow-up, as well as resources such as pedometers and screening booklets. Each plays an essential role: face-to-face follow-up strengthens the bond between health professionals and patients and allows for more detailed assessments, while telephone follow-up keeps the person connected and offers ongoing support [56]. During the first round, the expert panel suggested the creation of an SP kit in T2DM, which would bring together the resources made available for the development of the interactive components. After analysing the perspectives presented, a kit consisting of a backpack, SP passport, pedometer and water container was included as a new intervention to be carried out and explained by the diabetes nurse during the counselling and referral process. Together, these approaches form an integrated SP intervention to empower people in self-care, promote health literacy and improve the quality of life and well-being of the person with T2DM. A final set of 30 SP interventions established the multiple complex components of the SP intervention in T2DM. The design for and with the person living with T2DM integrates a wide range of community resources and activities, promoting self-care, health literacy and physical, emotional and social well-being.

8. Strengths and Limitations

The expert group of clinical and nonclinical professionals, community stakeholders and the person who lives with T2DM collaborated in a cocreation process to refine the assessment of how specific SP interventions could impact the health behaviours of the person with T2DM. Their feedback, which drew from their expertise in areas such as diabetes, SP, nutrition, physical activity and social environments, played a crucial role in shaping the model and the components of the SP intervention future. The active involvement of T2DM patients, as future users of the intervention, was vital for conceptualising solutions that meet their specific needs and are feasible in terms of location, frequency and duration of use. Involving stakeholders from various backgrounds guarantees a more holistic perspective, addressing a wide range of needs and concerns that a more homogeneous group might otherwise be overlooked. As a key strength, a participatory approach increases the likelihood of the SP intervention acceptance and engagement, fostering trust between patients, healthcare providers and intervention developers.

One of the study’s limitations lies in the facilitating role of the modified NGT process. The facilitator, as a research team member, may inadvertently influence how the content is presented and discussed in group settings. This close relationship between the facilitator and the subject matter could introduce bias, steering discussions in a way that might foster specific ideas or perspectives, thus compromising the neutrality of the facilitation. This influence could impact group dynamics, potentially skewing the outcomes and undermining the objectivity of the results. Another limitation may stem from the small number of participants in the Delphi panel. The literature presents varying definitions regarding the ideal number of participants, with the adopted number being described as valid [53]. Although all relevant parties were represented on the panel, the number of participants, especially the number of T2DM patients included, could be improved in future research. However, the consensus achieved through the combination of both techniques involved a group of 22 experts, enhancing the robustness of the data obtained.

9. Conclusion

In managing chronic diseases such as T2DM, SP plays an essential role by providing support beyond traditional approaches, promoting lifestyle changes and encouraging self-care behaviours. A person-centred approach creates the opportunity to use community resources tailored to individual needs, empowering them to take an active role in their health and well-being. By building a consensus among various experts, it was possible to outline a solid and structured model and intervention components to support and respond to the specific self-care and management needs of the T2DM patient. The different perspectives, whether from health professionals, patients or community stakeholders, were important pillars in identifying gaps, building new interventions and consolidating the proposed interventions. The active inclusion of the person and the involvement of the community from the earliest stages of development is fundamental to guaranteeing the adaptability of interventions to local realities. Adopting these approaches increases suitability and adherence to future interventions, ensuring that SP interventions meet the diverse needs of the person and their community.

Ethics Statement

The development of this study is part of a research study on self-care and health literacy among persons with T2DM, approved by the ARSLVT Health Ethics Committee under registration number 5078/CES/2022. In both consensus techniques, ethical principles were respected, and all invited participants took part based on their informed consent.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding

This research does not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

This study is part of a larger project based at the University of Lisbon on Community health promotion through nursing intervention in Social Prescribing. The Lisbon Centre for Research, Innovation, and Development in Nursing (CIDNUR), ESEL’s research unit, supports the publication of this paper.

Supporting Information

Additional supporting information can be found online in the Supporting Information section.