Sustained type 1 diabetes self‐management: Specifying the behaviours involved and their influences

Abstract Aims Sustained engagement in type 1 diabetes self‐management behaviours is a critical element in achieving improvements in glycated haemoglobin (HbA1c) and minimising risk of complications. Evaluations of self‐management programmes, such as Dose Adjustment for Normal Eating (DAFNE), typically find that initial improvements are rarely sustained beyond 12 months. This study identified behaviours involved in sustained type 1 diabetes self‐management, their influences and relationships to each other. Methods A mixed‐methods study was conducted following the first two steps of the Behaviour Change Wheel framework. First, an expert stakeholder consultation identified behaviours involved in self‐management of type 1 diabetes. Second, three evidence sources (systematic review, healthcare provider‐generated ‘red flags’ and participant‐generated ‘frequently asked questions’) were analysed to identify and synthesise modifiable barriers and enablers to sustained self‐management. These were characterised according to the Capability‐Opportunity‐Motivation‐Behaviour (COM‐B) model. Results 150 distinct behaviours were identified and organised into three self‐regulatory behavioural cycles, reflecting different temporal and situational aspects of diabetes self‐management: Routine (e.g. checking blood glucose), Reactive (e.g. treating hypoglycaemia) and Reflective (e.g. reviewing blood glucose data to identify patterns). Thirty‐four barriers and five enablers were identified: 10 relating to Capability, 20 to Opportunity and nine to Motivation. Conclusions Multiple behaviours within three self‐management cycles are involved in sustained type 1 diabetes self‐management. There are a wide range of barriers and enablers that should be addressed to support self‐management behaviours and improve clinical outcomes. The present study provides an evidence base for refining and developing type 1 diabetes self‐management programmes.


| INTRODUCTION
Self-management is recognised as a central component of diabetes care. It has been estimated that people with diabetes are responsible for 95% of their care. 1 Self-management is a broad concept involving cognitive, emotional and behavioural self-regulatory processes. 2 Type 1 diabetes is behaviourally demanding and complex, with self-management involving calculation and administration of insulin doses informed by blood glucose monitoring and review, estimated carbohydrate intake and consideration of other factors such as physical activity levels (e.g. 3,4,5). Sustained enactment of self-management behaviours is recognised as critical to optimising clinical outcomes. 3,4 While previous work has described behaviours involved in type 1 diabetes self-management (e.g. 6), relationships between the behaviours and the order in which these need to be carried out to increase likelihood of improvements in blood glucose levels are undefined.
In the UK, National Institute for Health and Care Excellence (NICE) guidance recommends that adults with type 1 diabetes are offered self-management support in the form of a structured education programme. 5 One recommended programme is Dose Adjustment for Normal Eating (DAFNE), which has been widely adopted nationally and internationally, and is described in detail elsewhere (see 8,9). It is a skills training course that advocates a regimen of flexible intensive insulin therapy, promoting freedom in dietary choices. While DAFNE is effective in improving quality of life and reducing HbA1c in the months following participation, improvements in HbA1c are not typically sustained beyond six to 12 months, 6-8 although see. 9 This suggests there may be scope to promote sustained self-management by refining DAFNE and other similar programmes, drawing on behavioural science. [10][11][12] Intervention development frameworks from behavioural science offer a systematic approach to refining interventions. The Behaviour Change Wheel is one such framework, which is based on a synthesis of intervention frameworks and has the COM-B (Capability-Opportunity-Motivation-Behaviour) model at its hub. 13 The framework has been applied to refine existing interventions aiming to change health-related behaviours. [14][15][16] The first step in the Behaviour Change Wheel process is identifying and precisely specifying the behaviour(s) underpinning the outcome of interest. The second step is identifying the key influences on these behaviours, and what it would take to bring about change. 17 The COM-B model 13 presents an integrated way to systematically explore behavioural influences. It specifies three necessary conditions for a behaviour to occur: capability (i.e. knowledge and skills), opportunity (i.e. physical and social environment), and motivation (i.e. reflective, or conscious, and automatic reactions).
In this study the Behaviour Change Wheel framework was applied to 1) specify the behaviours involved in sustained type 1 diabetes self-management, and 2) identify the influences on these behaviours. This a pre-requisite for subsequently identifying behaviour change techniques and intervention components that address identified barriers and enablers and facilitate sustained self-management behaviours. This study was part of a wider programme of research, DAFNEplus, in which behavioural science was used to optimise the DAFNE self-management programme to improve clinical outcomes and quality of life for people with type 1 diabetes. 18

| Design
A two-part mixed-methods study was conducted corresponding to the first two stages of the Behaviour Change

Novelty statement
• Type 1 diabetes self-management programmes improve HbA1c in the short-term, but participants may struggle to maintain behavioural changes. • This is the first study to map key diabetes selfmanagement behavioural processes using the Behaviour Change Wheel. • Three cycles of self-management behaviours were developed: Routine, Reactive and Reflective. 34 barriers and 5 enablers of these behaviours were identified, relating to Motivation (e.g. 'Discomfort managing diabetes in public'), Capability (e.g. 'Lack of skills to apply principles') and Opportunity (e.g. 'Group experiences adding credibility'). • This work will inform refinement of self-management programmes to support long-term behaviour change by incorporating techniques targeting barriers/enablers of maintenance. Wheel process for developing or refining behaviour change interventions. 17 Part 1 involved expert stakeholder consultation to identify the behaviours involved in self-management advocated by the DAFNE programme. Part 2 involved identifying barriers and enablers to enacting these self-management behaviours across multiple evidence sources and categorising these according to COM-B (see Figure 1 for an overview). Ethical approval for this study was granted by the University College London Division of Psychology and Language Sciences Ethics Committee (CEHP/2013/508).

| Identifying type 1 self-management behaviours according to DAFNE
A face-to-face consultation meeting held in London to identify the behaviours involved in the self-management of type 1 diabetes was attended by 16 stakeholders: five clinical diabetologists with experience of delivering or developing the DAFNE programme, three specialist nurse and dietitian DAFNE educators, six health and clinical psychologists and behavioural scientists with expertise in type 1 diabetes (two of whom had type 1 diabetes) and two independent representatives of the type 1 diabetes patient advisory group who had previously attended a DAFNE course. Prior to the stakeholder meeting, a patient advisory group meeting attended by five people with diabetes and two clinicians was held in Sheffield to discuss the DAFNE programme and the 'fundamental principles of self-management' (defined as things people had to do or know in order to be able to self-manage effectively). These principles were added to the actions generated by the stakeholder meeting. Stakeholders were asked to review the most recent DAFNE course curriculum prior to the meeting and to suggest the actions (i.e. behaviours) involved in self-management of type 1 diabetes, based on their review of the curriculum and their diabetes experience. After the face-to-face meeting, stakeholders were sent this list for review and asked to add any actions that might have been missed. Four behavioural scientists subsequently met to group the behaviours generated into themed topics. For example, 'calculating carbohydrate content by weighing' and 'Using resources to identify carbohydrate content of foods' were put into the group 'Carbohydrate-Counting'. In recognition of the interdependent nature of the behaviours involved in self-management, and high level of self-regulation required to maintain them (i.e. monitoring behaviour/outcomes in relation to a goal and acting to correct discrepancies), these topics were then organised into flow diagrams representing cycles of behaviours. Stakeholders were invited to comment on the diagrams via email or in face-to-face meetings, and suggested refinements to ensure the cycles represented the self-regulatory processes involved in diabetes self-management as accurately as possible. The final self-management behaviours and diagrams were circulated to all stakeholders for approval.

| Identifying influences on selfmanagement behaviours
Influences on the target self-management behaviours were extracted by two researchers (KH and SSF) from three sources. The sources were: We extracted the reported barriers and enablers to self-management following participation in DAFNE from the findings of a systematic review and meta-ethnography, 19 which reviewed 18 articles reporting on six qualitative studies of people with type 1 diabetes who had taken part in a DAFNE course.

| Red flags
Seven DAFNE educators from three NHS Trusts (with on average 16 years' experience delivering DAFNE courses) were asked to list factors (characteristics, situations, course experiences etc.) based on their clinical experience that represent 'red flags' that might indicate that a participant may need additional support with self-management at three time points: before, during and after participation in a DAFNE course.

| FAQs
The DAFNE educators were also asked to create a list of frequently asked questions (FAQs) that participants typically ask after attending a DAFNE course, drawing on their clinical experience and by asking participants via the DAFNE Online forum to report aspects of self-management they were unsure of or struggled with post-course.

| Data analysis
Extracted barriers and enablers from the systematic review, red flags and FAQs were categorised according to the COM-B model using a combined deductive framework analysis and inductive thematic analysis approach (following the methods of Graham-Rowe et al. 20 . Deductive framework analysis: Extracted data were deductively coded to the COM-B domain(s) they were judged to represent. For example, 'Repeatedly failing to achieve blood glucose targets is demoralising and results in feelings of failure' was coded under the domain 'automatic motivation' (i.e. emotions). Coding was conducted in MS Excel by two researchers independently (KH/SSF) and any disagreement or uncertainty resolved via discussion with a third researcher (FL).
Inductive thematic analysis: Each COM-B domain was considered in turn. Similar items coded to the same domain were grouped together and a summary theme label was inductively generated, with any disagreements resolved as above. For example, the items 'Do I need to retest after a hypo?' and 'If my BG is below 3.5 at my mealtime, should I eat first then inject?' (both FAQs) were both coded to the domain Psychological Capability, grouped together and summarised using the inductively generated theme label 'Lack of skills to apply DAFNE principles' within the subtheme 'Lack of knowledge/skill about how to treat a hypo'. Resulting themes were classified as either a barrier or an enabler by looking at the qualitative meaning of the data points contributing to that theme, and whether the data suggested that this factor hindered or sustained self-management. Themes were then mapped to the corresponding behaviour(s) identified in Step 1. For example, the theme 'Perception of hypoglycaemia as an opportunity to overindulge in unhealthy foods' was classified as a barrier and mapped to the behaviour 'Carry out and record hypoglycaemia treatment' in the Reactive cycle (see Figure 1 for an overview of methodological steps).

| Identifying self-management behaviours
Stakeholders identified 150 behaviours needed for optimal and sustained self-management of type 1 diabetes. These were combined into three behavioural cycles reflecting the different temporal and situational aspects of diabetes selfmanagement (see Figure 2).

| Cycle one: Routine self-management
This cycle describes the adaptive self-management behaviours required to keep blood glucose within target on a daily basis, influenced by food intake, physical activity and alcohol consumption. The steps involve daily checking, planning and implementing a response, based on regular data gathered. This cycle may need to be enacted several times a day in response to fluctuating parameters.

| Cycle two: Reactive self-management
This cycle describes the self-management behaviours required to manage the impact of unanticipated events that have an impact on blood glucose regulation, e.g. hypoglycaemia or suspected illness or presence of ketones. The steps involve checking, rechecking, problem solving, reflecting and planning. This cycle is in response to an atypical event and incorporates actions to reduce the incidence of these events where possible.
3.1.3 | Cycle three: Reflective selfmanagement This cycle describes the longer-term review of blood glucose data (ideally on a weekly basis) to evaluate the impact of daily self-management behaviours. The steps involve planning, implementing, reviewing, monitoring and problem solving. This cycle supports the fine-tuning of habitual selfmanagement behaviours by identifying patterns in blood glucose data, and the instigation of more long-term adjustments such as changing background insulin dose or mealtime ratios.

| Identifying influences on self-management behaviours: Barriers and enablers
The stakeholder consultation produced 97 FAQs and 69 red flags; 45 barriers and enablers were extracted from the literature review. This resulted in 211 data points overall. These were synthesised into 39 themes representing barriers and enablers to self-management behaviours, coded to COM-B. Thirty-four of these represented barriers to sustained selfmanagement and five represented enablers. These are presented in Table 1 according to each domain of COM-B and summarised below. All barriers and enablers mapped to target behaviours can be found in Appendix 1.

| Physical capability
All three themes within this category were barriers, and related to physical symptoms. For example, 'impaired awareness of hypos' is a barrier as the lack of physiological cues does not provide a trigger for checking blood glucose and treating of hypoglycaemia.

| Psychological capability
Six themes within this category were barriers, and included difficulty adapting self-management routines in the face of changing life events such as new caring responsibilities. There was only one enabler: establishing and maintaining routines, which diminishes the cognitive burden of behavioural change and reduces the likelihood of behavioural sequences being disrupted. Many of these themes influenced all self-management behaviours as they related to sub-optimal DAFNE knowledge and skills, for example assessing and adjusting meal-time ratios or managing physical activity. (See the Appendix for a full list of behaviours).

| Automatic motivation
All six themes within this category were barriers, for example anxiety about managing diabetes and acute symptoms, or a lack of acceptance, both leading to disengagement with self-management behaviours in order to reduce emotional discomfort associated with diabetes. Again the majority of these barriers related to all self-management behaviours, as they could contribute to a global withdrawal from self-management.

| Reflective motivation
Of the 14 themes in this category, 12 were barriers, such as a lack of trust in guidelines, leading to people modifying their goals (e.g. blood glucose targets). Two were enablers, relating to feeling empowered by new knowledge and skills, which leads to confidence in independent decision-making, and having clear guidelines and targets to aim for. The majority of themes related to all behaviours, such as a lack of prioritisation of diabetes self-management when faced with competing demands, or a lack of readiness to change, causing disinclination to enact certain principles.

| Physical opportunity
This category contained four barriers, such as inadequate access to monitoring equipment (e.g. blood glucose meter strips), which impairs ability to follow blood glucose monitoring guidelines. One enabler was the presence of bolus advisers facilitating data recording, which increases the likelihood of timely recording by reducing the associated cognitive burden. Several themes, such as lack of a stable home environment, related to all self-management behaviours.

| Social opportunity
Three themes in this category were barriers relating to social influences beyond the DAFNE setting, such as the perception of external criticism, influencing behaviours such as injecting insulin in public before eating. The single enabler related to Group experiences adding credibility Enabler All SR The accumulation of experiences which patients bring to the group helps to add credence and credibility to key teaching points on the curriculum. The findings from this study highlight a number of potential ways to optimise type 1 diabetes self-management programmes. First, conceptualising self-management behaviours in terms of functionally connected cycles recognises the complexity of type 1 diabetes self-management in the moment (within the Routine and Reactive cycles) as well as acknowledging the importance of reflecting on self-management over a longer period (within the Reflective cycle). The high number of distinct behaviours identified, reflects the notion that sustained self-management in type 1 diabetes is a complex set of behaviours, consisting of multiple sub-behaviours. People may struggle with certain behaviours more than others. Evidence reported from other studies of people with diabetes 23 as well as representatives and clinicians in our stakeholder group suggests that the Reflective cycle is the most challenging for people. This points to a need for a more targeted and tailored approach to designing self-management programmes in type 1 diabetes. This should begin with identifying which specific behaviours the individual is finding challenging, setting behaviourally focused goals (rather than goals around improving self-management more broadly), and selecting strategies to address the barriers and enablers uniquely associated with selected behaviours. Although current DAFNE teaches pattern recognition, this study highlights the need to promote the benefit of reflective over spontaneous decision-making and furthermore, to provide skills in impulse control and emotion regulation that would enable this process to occur in real life. The behavioural cycles developed as part of this study could also be used in a clinical context during consultations with people with diabetes to facilitate consideration and identification of target behaviours and to provide a temporal order of skill acquisition, starting with routine daily adjustments and moving onto less frequent and more specific blood glucose management scenarios.
Second, the large number of themes within the COM-B domain of Motivation suggests that self-management of type 1 diabetes longer-term may be driven predominantly by emotional and cognitive influences. In particular, anxiety in relation to diabetes was identified as a key influence in automatic motivation (i.e. anxiety related hypoglycaemia, general health, 'getting it wrong'). While fear of hypoglycaemia has received much research attention, the impact of anxiety about complications has been less studied, despite research suggesting it may be at least as significant a concern. 24 It is possible that fear of complications could prompt withdrawal from diabetes self-management through reducing acceptance, or conversely that it could act as a motivating factor to keep blood glucose in range. This highlights the need for self-management interventions to extend beyond educational and training strategies focused on short-term knowledge and skill acquisition, towards inclusion of psychological support to help manage anxiety, bolster resilience, and enhance and maintain motivation longer term (e.g. habit formation, persuasion, enablement). This will involve addressing the identified cognitive and emotional experiences (beliefs, perceptions, anxieties, concerns, and acceptance) associated with living with the condition and affecting engagement in self-management behaviours.
Ensuring that the philosophy underpinning self-management programmes, such as DAFNE, are person-centred, flexible and tailored to the needs, values and motivations of the individual participants is critical to ensuring successful delivery and continued engagement. 25 Multi-disciplinary working and adhering to principles of care planning support this approach. Ongoing training and support for health care professionals in these approaches is also required and an area for future research.
Use of the Behaviour Change Wheel framework 17 provides a method for systematically selecting strategies to address the key identified barriers and enablers to self-management. The COM-B model is mapped to the Behaviour Change Wheel which specifies nine broad ways to change behaviour (i.e. persuasion, enablement, environmental restructuring), as well as a taxonomy of 93 more granular behaviour change techniques (BCTs), in order to suggest the types of interventions likely to be effective in addressing different types of barriers and enablers 17 Potential BCTs that could be included in future or revised interventions to support sustained self-management in type 1 diabetes are summarised in Table 2.
Such strategies will be evaluated in the DAFNEplus research programme, 18 which aims to refine the existing DAFNE self-management programme to include BCTs targeting the behaviours and influences to sustained self-management identified in this study (Stanton-Fay et al., under review).
This work could also help strengthen self-management research by feeding into the refinement of diabetes specific measures such as the Summary of Diabetes Self-Care Activities measure 23 or the Diabetes Self-Management Questionnaire. 4 As yet, there is no tool available for assessing engagement in self-management behaviours specific to type 1 diabetes. However, work to develop such a questionnaire is currently underway: Self-Care Behaviours in Type 1 Diabetes (SCB-T1D, Cooke et al., in prep.).
In addition to the grounding in behavioural science theories and frameworks, a key strength of this study is the use of triangulation, incorporating perspectives from clinicians, people with diabetes and researchers. People living with diabetes and clinicians are able to offer different insights into challenges and factors enabling self-management; a review of relevant studies concluded that co-production (involving users in the evaluation or design of services) is particularly valuable when generating ideas for healthcare change. 26 One limitation, is that this study involved a secondary analysis of the published findings of a previous review on barriers and facilitators for sustaining self-management skills after attending a type 1 diabetes structured education programme. 19 This introduces a source of potential bias, in that the original data had already undergone a degree interpretation and synthesis by the review authors. However, involvement of the senior author of that review (JB) in the current study's analysis allowed opportunity for discussion and clarification to ensure similarity in data interpretation. Furthermore, none of the studies included in the review explored barriers and enablers to self-management using a theory of behaviour change. It is therefore possible that they did not explore the full range of potential influences on self-management behaviours, highlighting the need for more theory-based studies in this field.
In conclusion, this study identified three cycles of self-management behaviours: Routine, Reactive and Reflective, translating the complexity and self-regulatory nature of self-management into a useable framework. We found enactment of these cycles to be influenced by 39 themes relating to Capability, Opportunity and Motivation; 35 of which acted as barriers to self-management and five of which acted to enable self-management. This work is the first to the authors' knowledge to articulate the influences on self-management of type 1 diabetes using the Behaviour Change Wheel framework. It provides a model for considering the behavioural processes underpinning long-term conditions that require enactment of self-management behaviour, with a view to developing targeted interventions, including that used in the DAFNEplus intervention development programme. 18