Developing and evaluating a child-centred intervention for diabetes medicine management using mixed methods and a multicentre randomised controlled trial

Aim

The overarching aim of the multiple strands of work was to develop and evaluate an individually tailored, age-appropriate diabetes diary and information pack for children and young people aged 6–18 years with T1D to support decision-making and self-care with a specific focus on insulin management and electronic blood glucose monitoring, compared with available resources in routine clinical practice.

Objectives

The EPIC study objectives were as follows:

1. to review gold standard diabetes clinical guidelines, currently available diabetes information, including findings from the linked qualitative IMP study, to identify best practice, and types/formats of information most likely to assist age-appropriate decision-making and choices concerning blood glucose monitoring and insulin management

2. to develop an age-appropriate diabetes information intervention (child-centred diabetes diary and information pack) for children and young people to support the appropriate use of blood glucose monitoring to optimise management of and concordance with their insulin regime

3. to explore the utility of the child-centred diabetes diary and information pack (in this context utility refers to ease of use and fitness for purpose) within different contexts in which children and young people manage their routine diabetes care (home, school, community), with and without support from parents or health-care professionals (HCPs), and in alternative settings

4. to explore how children and young people with and without their parents, teachers, nurses and doctors use (or not) the diabetes diary and information pack to support decision-making, in particular how children and parents ‘self-prescribe’ the correct (or incorrect) dose of insulin

5. to identify similarities and differences between the diabetes diary and information pack developed for adolescents and those available within adult diabetes services

6. to evaluate the diabetes diary and information pack within the context of routine diabetes care in relation to patient outcomes (diabetes-specific health-related quality of life, generic health-related quality of life, medicine and treatment concordance, acceptability, ease of use and glycaemic control)

7. to identify gaps in knowledge to inform a future research agenda.

In the following sections we outline the complex legal framework that is applied to children’s health care, which was at the forefront of our minds when developing age-appropriate diabetes information packs and diabetes diaries for children and young people aged 6–18 years. Having explained the legal framework, attention is then turned to describing the theoretical frameworks and proposition underpinning the development and evaluation of the children’s diabetes information packs and diabetes diaries.

Children aged < 16 years

Children aged < 16 years are not deemed to be automatically legally competent to give consent. The courts have determined that such children can be legally competent if they have sufficient understanding and maturity to enable them to understand fully what is proposed. This concept – now known as Gillick competency – initially arose in the case of Gillick v. West Norfolk and Wisbech Health Authority in 1985. 37 The term ‘Fraser guidelines’ is also used in this respect (Lord Fraser was the judge who ruled on the case). Decision-making regarding consent will depend on the relationship of the HCP with the child and the family and also the kind of consent being sought. Even if a child cannot give consent in law he or she will need appropriate information to give his or her assent. Children may be competent to take some decisions even if they are not competent to make others. For example, even young children may be competent to make some health-care decisions such as their preferred choice concerning whether to take the liquid or the tablet form of a medicine, whereas young children would not be expected to understand or weigh up information concerning the need for an operation. If a child is aged < 16 years and deemed not legally competent, consent will need to be obtained from someone with parental responsibility, unless it is an emergency.

In the context of the IMP and EPIC study, which focus on long-term conditions, competency is something that can be developed over time by presenting children with information appropriate to their age and level of education and understanding. The emphasis in DH guidance38 is that families of children in this age group should be involved in decisions about their care unless there is a very good reason for not doing so. If, however, a competent child aged < 16 years is insistent that his or her family should not be involved, the child’s right to confidentiality must be respected, unless such an approach would put the child at serious risk of harm.

Young people aged 16 and 17 years

Once children reach the age of 16 years they are presumed in law to be competent unless there is evidence to the contrary. Importantly, children aged 16 or 17 years cannot refuse treatment if it has been agreed by a person with parental responsibility or the court and it is in their best interests. Therefore, they do not have the same status as adults aged ≥ 18 years. Nonetheless, in a diabetes context, young people in this age group frequently do not follow advice or maintain optimal glycaemic control, but this is considered lack of concordance, adherence or compliance rather than outright refusal to treatment. The Mental Capacity Act 200539 applies to people aged 16 years and over. The DH38 recommends that it is nonetheless good practice to encourage young people of this age to involve their families in decisions about their care, unless it would not be in their interests to do so. Evidence suggests, however, that parents often find it difficult to keep track of what their teenage children are doing in a diabetes self-care context.

Adults aged > 18 years

Once a person has reached the age of 18 years no-one can give consent on his or her behalf. If he or she is not competent HCPs can provide treatment and care providing that this is in his or her best interests, which may be determined by the court. 38

Assessing competence

Even if a young person is aged > 16 years this does not, as with adults aged ≥ 18 years, necessarily mean that the person is competent. A competent person is able to:

• understand and retain information pertinent to the decision about their care, that is, the nature, purpose and possible consequences of the proposed investigations or treatment, as well as the consequences of not having treatment

• use this information to consider whether or not he or she should consent to the intervention offered and to communicate his or her wishes.

The DH guidance38 advises HCPs that they should not assume that children with learning difficulties are unable to make competent decisions, which can be aided by presenting them with information in an appropriate way. If, however, any child aged < 18 years is deemed not competent a person with parental responsibility would need to give consent. Similar provision is made in Scotland by the Age of Legal Capacity (Scotland) Act 1991. 40

Parental responsibility

Parental responsibility includes the right of parents to consent to treatment on behalf of a child or young person aged < 18 years when he or she is unable to provide valid consent, provided the treatment is in the best interests of the child.

Devolving parental responsibility

Provision in law is made for parents who are not with their children 24 hours a day. There are times when parents can devolve the responsibility for consent to care and treatment to others, for example grandparents, child minders and school teachers, for certain situations such as emergency care, management of minor illness and support with ongoing treatment of long-term conditions such as diabetes.

The DH guidance38 outlines that parental devolved consent does not need to be in writing if consent is devolved to close family members. It is, however, best practice for professionals such as school teachers and employed carers to develop a shared care plan with parents that outlines an agreed approach in specific circumstances (such as administering medicines in a school setting to a child experiencing an acute hypoglycaemic attack).

In the EPIC study we were particularly interested in finding out children’s experiences of managing their diabetes away from their families and in situations such as at school where parental responsibility is devolved and they are required to self-care with minimal or no adult supervision.

Having outlined the legal framework, attention is now turned to the theoretical frameworks.

Biopsychosocial and anthropological model of childhood and illness management

As with the IMP,1 from conception we were influenced by the imperative, drawn from social anthropology, to elucidate and critically analyse context and content, diversity of viewpoint and ideas driving motivation and actions. 44,45

We recognised that policies need to be placed within the context of children’s lives and the illnesses they experience, and acknowledged that our research should take the anthropological context into account. Inspired by Jordan’s46 account of birth in four cultures, we also recognised that, although biomedically conditions may be considered universally the same (e.g. childhood diabetes), how conditions are experienced and managed may be strikingly different for different children and their families and different groups. Nevertheless, despite differences in orientation and interpretation, certain ideas about how care is managed are shared. Levine,47 an anthropologist writing in 1986, writes that the shared organisation of ideas includes the intellectual, moral and aesthetic standards prevalent in a community, and the meanings of communicative actions (pp. 66–77). It is, he suggests, important to take account of general rules, concepts or assumptions that generate the particulars readily accessible to the researcher.

The social anthropological perspective acknowledges that individual children’s experiences of having diabetes are unique. They have to manage diabetes within the context of their everyday lives, which includes being part of a family within society, attending school and participation in out-of-school activities. Families need to take responsibility for diabetes management, with support from specialised diabetes teams. When younger children are diagnosed parents assume responsibility for their child’s diabetes, but there is a gradual shift in responsibility from parent to child as children and young people become more knowledgeable about diabetes and develop the necessary cognitive skills to make decisions about, for example, insulin titration according to blood glucose levels.

Children also have different experiences as they grow and develop and, although there are periods of equilibrium, they need to constantly consider and manage their diabetes as they encounter changes in their lives. In adolescence puberty is known to increase insulin requirements and has a negative effect on glycaemic control. In addition, teenagers are more likely to ignore diabetes management because of, for example, increased independence from parents, increased risk-taking behaviours and a desire for peer conformity.

This approach to conceptualising the child, and in particular their long-term diabetes management, primarily within the family and society, is fundamentally different from the acute medical model whereby children enter the health-care system for diagnoses and treatment under the direction of a lead doctor and a team of HCPs who direct and initiate care (e.g. Kennedy17).

The context of children’s and young people’s lives in relation to their health, medicines management and self-care is an important consideration involving the interplay between their medical condition, how they orientate themselves to their condition, their values and expectations, as well as the settings in which they find themselves. In Figure 1 we have summarised the key concepts that we have taken into consideration when collecting and analysing data to support this approach.

FIGURE 1.

Biopsychosocial model for families and children with acute and long-term conditions.

The Promoting Action on Research Implementation in Health Services framework

In addition to using the MRC framework for designing and evaluating complex interventions, the PARiHS framework will be used as the framework for the translation of evidence into practice evaluation. 41,48 The framework has been theoretically and empirically developed to represent the interplay and interdependence of the many factors influencing the implementation of evidence into practice. This is explained by a function of the relation between evidence, context and facilitation. 41,49,50 The hypothesis offered is that for implementation of evidence to be successful there needs to be clarity about the nature of the evidence being used, the quality of context and the type of facilitation needed to ensure a successful process. The framework has been used by others to inform the design and evaluation of evidence into practice initiatives. 51–53 The PARiHS framework is particularly relevant to this study because:

1. The study aims to introduce new diabetes diaries and assembled diabetes information packs (evidence) into children’s self-care regimes and health-care practice to improve blood glucose monitoring and insulin management. Understanding the factors that influence its implementation and use will be important in determining the acceptability and feasibility of the information pack (facilitation) – this framework will provide a conceptual guide for mapping these issues.

2. Understanding how the information pack is used in different contexts where children and young people manage their diabetes will be key in the evaluation of its utility and contribution. Applying the framework will allow a focus on the key contextual variables mediating the implementation and use of the information pack.

3. It facilitates the gathering of individual (e.g. child/practitioner/carer) experiences as well as appreciating the fit with the broader context of care delivery.

Medical Research Council framework for developing and evaluating complex interventions

Children’s health information is part of a complex intervention, or a number of different complex interventions, that contribute to care pathways and clinical management plans in different settings for different tracer conditions. Our approach to exploring the use of children’s health information in routine practice settings follows the theoretical phase of the MRC framework42 to understand better child, family, HCP and organisational behaviours concerning the use (or not) of children’s health information, and informal evidence regarding organisational, child and family, and professional constraints or beliefs that may promote or inhibit behavioural change. This initial phase of the MRC framework to assess theory and evidence was considered to be most useful when identifying preliminary ideas about the types, formats and facilitation of health information most likely to help children and young people manage their illnesses in different settings in age-appropriate ways.

The 2008 updated MRC guidance43 provides further clarity on the complexity of interventions and how to consider the relative roles of active ingredients, interacting components and the relative contribution of children’s health information when used within a larger complex system of children’s service delivery, and beyond that when used by families and in different community contexts such as home and schools. We have used the updated framework to consider different dimensions of complexity, such as the number and difficulty of behaviours, and age-appropriate behaviours, required by those delivering and receiving children’s health information in various settings, and the degree of flexibility or age-appropriate tailoring of children’s health information to specific children and contexts. We were also interested in developing our understanding of how children’s health information was used by children, parents and HCPs in different settings in order to establish its practical effectiveness and where strong and weak links occurred that could potentially impact on outcomes and quality of family and child-centred care.

An integrated model of complex interventions, implementing and using evidence in practice and understanding the family and societal context

Figure 2 shows how we have integrated the three frameworks into a single model to guide the systematic review, interpretation of the critical discourse analysis (CDA), intervention development, delivery of the trial and analysis of the process evaluation.

FIGURE 2.

Single theoretical model: IMP conceptual model.

Theoretical models of partnership and participation and the critical success factors for successful translation of children’s health information into routine practice

As in the parallel IMP,1 the EPIC project is anchored to the idea that partnership between the public and the UK NHS has the potential to build a healthier society. 10 Information on medicines management is regarded as critical to effective partnership. 11,12 If ‘partnership’ is to be promoted then it is important to recognise that children and young people are capable of being partners in the approach to their treatment. A detailed description of how the theoretical models were developed is reported in the IMP. 1 The theoretical model shown in Table 1 represents high and low levels of partnership and participation in NHS service delivery and organisational contexts.

Translation, implementation and use of children’s health information

The theoretical model of the critical success factors for translation and use of children’s health information in routine NHS care (Table 2) uses the conceptual domains of the PARiHS framework to show high and low levels of ‘partnership and participation’ in care and decision-making between children, families and HCPs, and successful implementation of children’s health information in routine clinical practice. Process evaluation findings will be mapped against the theoretical models to identify where barriers to successful implementation and use of the EPIC packs have occurred.

The EPIC project proposition

More broadly, the theoretical models developed at the conclusion of the IMP1 helped us consider a theoretical proposition to explore further in the EPIC project. We proposed that for optimal translation of children’s health information and participation in care:

High quality, age-appropriate and accessible children’s health information, valued by health care professionals, children and parents, and shared using child-centered, facilitative and partnership approaches to care and service delivery, will be more highly used and a more effective component of complex interventions to optimize children’s long-term condition management.

p. 2071

Stage 1: context

We conducted multiple strands of investigation to focus our understanding on the context of children’s diabetes care, including:

• a review of reviews and mixed-method systematic review of the barriers to and enablers of optimal diabetes self-care by children and young people in educational settings (see Chapter 3)

• further searching for currently available children’s diabetes information to add to the database constructed for the IMP

• an extension and update of the IMP comparative analysis to establish whether key health messages in children’s diabetes information matched with diabetes clinical guidelines (see Chapter 2)

• an extension of the CDA refocusing on diabetes to investigate the content and meaning of selected currently available children’s diabetes resources (see Chapter 2).

Stage 2: intervention development – diabetes diaries and information packs

In addition to incorporating evidence from the IMP1 and stage 1 of this study, we conducted further strands of investigation and consultation, including:

• convening an expert clinical advisory group to advise on intervention development and clinical risk management

• undertaking qualitative interviews and focus groups with children and young people to ascertain their information preferences and self-care practices with an emphasis on children who had spent time away from their families

• obtaining children’s and young people’s perspectives on various iterations of the age-appropriate diabetes diaries and information packs (web consultation and stall at diabetes charity family days).

Stage 3: randomised controlled trial to evaluate the diabetes diaries and information packs in routine practice

We designed and carried out a pragmatic RCT to assess the clinical effectiveness and cost-effectiveness of the diabetes diaries and information packs in routine practice (see Chapters 5 and 6).

Stage 4: process evaluation

As the trial did not detect an intervention effect, we have used the process evaluation to help explain the outcome. For the process evaluation we collected the following stands of evidence:

• interviews with HCPs in each site to document current routine practice and local clinical care pathways for children with T1D

• a baseline survey to determine the ethnic profile of trial participants

• baseline postcode analysis to determine the index of deprivation of trial participants

• qualitative interviews with children and parents to ascertain their experiences and perspectives

• a questionnaire completed by HCPs at the conclusion of the trial to ascertain their perspectives.

The process evaluation is reported in Chapter 7.

Searching for children’s diabetes information

The scoping exercise to identify children’s diabetes information conducted for the IMP1 identified 69 paper-based resources, 40 web pages and eight multimedia resources up until August 2008 (p. 57). This work was extended commencing in August 2008 and continued intensively for approximately 6 months. Examples of current diabetes information for children and young people were sought from charities, pharmaceutical companies, health and medical institutions and some international sources that could be accessed in the UK through the internet, for example sources in Canada, the USA and Australia. Since the initial period of intensive searching, update searches have been conducted on a monthly basis throughout the life of the project. However, as we cannot claim to have a comprehensive record of all such information sources, our aim has been to provide a snapshot of information that is currently available around the UK. The methods followed the same approach as in the related scoping exercise in the IMP. 1

In addition, we also contacted selected NHS organisations providing children’s diabetes care, including children’s diabetes services that were being considered as sites for the EPIC RCT, those known to have already produced children’s resources and a well-known children’s hospital in the UK. An audit trail of organisations and individuals contacted throughout our search can be found in Appendix 2 (see Audit of all organisations and individuals contacted for diabetes information resourcing). The websites of all NHS organisations, pharmaceutical companies and charities were first searched and then contacted by e-mail or telephone to request copies of their most up-to-date information.

Children’s diabetes information resources found

In total, 142 children’s diabetes information resources (leaflets, booklets, bespoke locally produced information, teaching aids, DVDs, CD-ROMs) were found and a database was established (see Appendix 2, Diabetes information database). Details of all of these resources including source, format, age range for which the information was written (as determined by two EPIC researchers) and date of publication were recorded on this database. Of these, 120 were currently available and made reference within the text to blood glucose monitoring and insulin management. A separate list of books (n = 10) (see Appendix 2, Resources: books) and a separate list of interactive websites (n = 36) (see Appendix 2, Resources: websites) were also produced.

Addressing diversity

The 110 paper-based children’s diabetes information resources (referred to henceforth as leaflets) were explored for issues of diversity (sex and ethnicity). It was found that there is a lack of availability of leaflets for children and young people whose first language is not English as all of the diabetes leaflets were written in English. However, in some centres in Wales, translations of selected leaflets were available in the Welsh language.

Images in the leaflets were also explored for issues of sex and ethnicity. Photographs, cartoons, clip art and pencil drawings were critically analysed. The results of this analysis of the visual representation of diversity are provided in Table 3. If a leaflet contained images of both males and females (but not necessarily in balance) it was recorded as representing a gender balance. If a leaflet contained images of children and young people from different ethnic backgrounds it was recorded as representing ethnicity. Overall, we found ethnicity to be well represented in leaflets for children and young people with T1D, with the images representing a wide mix of cultures and races. The gender representation of children was usually balanced; that said, boys and girls tend to be portrayed as having interests conventionally or stereotypically attributed to their respective genders. In several leaflets, however, there were more images of boys than girls. This over-representation of boys is similar to that seen in mainstream children’s media. 54,55 When animals and inanimate objects were used as the main characters/providers of information, these were almost exclusively male (Lenny the lion, Hu-Mee the frog, Frankie the frog, Desmond the dragon).

Ethnic and gender differences, although clearly present (e.g. different ethnicity is represented, especially in drawn illustrations; boys and girls are portrayed as having interests conventionally attributed to their respective genders), are not linguistically addressed in the texts. In other words, these texts construct boys and girls and minority ethnic and white children as essentially the same. As such, they do not take into consideration how T1D might be differentially experienced by boys and girls, or by those children and young people living within non-Westernised, non-white cultures within the UK. Normative maleness and whiteness shape the mode of address employed, which may have certain consequences for children, young people and their families in terms of the management of their condition. In other words, what is not taken into account, if normative maleness and whiteness largely shape these texts, is how diabetes is experienced differently by different genders, ethnicities, social class backgrounds, etc. Thus, for example, the healthy diet suggested might assume a standard white British diet and not take into account children from different ethnic backgrounds in which to eat healthily means eating different foods from those suggested by the text. Or it might mean that the reader judges his or her diet as somehow lacking or ‘unhealthy’ because it does not resemble what is presented, typically, in these texts.

Although this work enabled a greater understanding of how gender was represented in texts, when developing the EPIC diary and packs we were not able to specifically address gender differences linguistically. We had insufficient resources to produce additional gender-specific, age-appropriate diaries and we selected currently available children’s health information for inclusion in the pack.

Aim

The aim was to establish whether a sample of children’s diabetes health information resources on self-management of blood glucose monitoring and insulin management was consistent with clinical standards and best practice in selected diabetes clinical guidelines.

Methods

The methodology used was the same as in the IMP1 and is reported in the following sections.

Identification and selection of children’s diabetes information for quality analysis and critical discourse analysis

The search strategy for locating children’s diabetes information is detailed in the IMP report (p. 56). 1 The IMP scoping exercise had identified 56 individual items of diabetes information for children and young people with T1D aged between 6 and 18 years and 13 for their families and friends (69 in total). Books and websites were not included in this sample. We adopted a purposive and pragmatic approach to selecting examples of children’s diabetes information for quality analysis of the content and subsequent CDA based on available time and resources to conduct the analysis. The process of selection is shown in Figure 3. Children’s diabetes information resources were excluded for a variety of reasons. The final sample included only those resources that focused on blood glucose monitoring and insulin management. Selection of resources was prioritised according to whether contextual information about the resource, determined as part of the IMP, was available (see Appendix 2, Discourse analysis: contextual information), NHS staff reported that the resource was commonly used in the NHS (see Appendix 2, Information used within NHS trusts for children and young people with diabetes), the resource was described as ‘popular’ in IMP fieldwork, the resources had been subject to CDA in the IMP and the final selection represented pharmaceutical companies, charities and health and medical institutions. In total, 19 children’s diabetes resources focusing on blood glucose monitoring and insulin management were analysed, as shown in Box 2.

FIGURE 3.

Flow diagram of selection of children’s diabetes resources for quality analysis and CDA. T2D, type 2 diabetes.

BOX 2 - Children’s diabetes information used in quality analysis and CDA

6–10 years

Just for You – Diabetes UK, 2004/5.

Diabetes Made Simple – Novo Nordisk, 2005. ^a

Hangin’ with Hu-mee – Eli Lilly, 2007.

Tadpole Times – Diabetes UK, 2010. ^a

T1 – Juvenile Diabetes Research Foundation, 2010. ^a

11–15 years

T1 – Juvenile Diabetes Research Foundation, 2010. ^a

Managing your Diabetes – Roche Diagnostics, 2005.

On the Level – Diabetes UK, 2010.

Growing up with Diabetes – BD Medical – Diabetes Care, 2009.

Home Blood Glucose Testing with Diabetes, Streetwise collection – Eli Lilly, 2008. ^a

High Blood Glucose with Diabetes, Streetwise collection – Eli Lilly, 2008. ^a

16–18 years

Home Blood Glucose Testing with Diabetes, Streetwise collection – Eli Lilly, 2008. ^a

High Blood Glucose with Diabetes, Streetwise collection – Eli Lilly, 2008. ^a

Living with Diabetes – Abbott Diabetes Care, 2008. ^a

Joe’s Rough Guide to Diabetes – Sanofi Aventis, 2009.

Living with Diabetes – BD Medical – Diabetes Care, 2009. ^a

Parents/families

From Hospital to Home – Waltham Forest Primary Care Trust, 2006.

Tots to Teens – Diabetes UK, 2007/8.

A Parent/Carer Guide. Bringing up a Child with Diabetes – Eli Lilly, 2007. ^a

Caring for a Child with Diabetes – BD Medical – Diabetes Care, 2009. ^a

Type 1 Diabetes – Information for Families – Royal United Hospital Bath NHS Trust and Wiltshire Primary Care Trust, 2009. ^a

Salisbury Child & Adolescent Diabetes Clinic Information Folder – Salisbury NHS Foundation Trust, 2004. ^a

Adults^b

Managing your Diabetes, an Introduction – Roche Diagnostics, ND.

Testing, Testing. The Balanced Guide to Blood Glucose Testing – Diabetes UK, 2011–12.

Complications – Eli Lilly, 2010.

ND, not dated.

Abbott Diabetes Care – glucose meter company; BD Medical – Diabetes Care – medical technology company; Diabetes UK – charity; Eli Lilly – pharmaceutical company; Juvenile Diabetes Research Foundation – charity; Novo Nordisk – pharmaceutical company; Roche Diagnostics – medical technology company; Sanofi Aventis – pharmaceutical company.

a Children’s diabetes information used in CDA.

b Adult information used only in CDA.

Key findings from the comparison of children’s diabetes guidelines and children’s diabetes information resources

Key findings from the comparison of children’s diabetes guidelines and children’s diabetes information resources were as follows:

• Children’s diabetes clinical guidelines were broadly consistent, with a small degree of variation regarding recommended blood glucose levels and varying levels of detail (see Table 4).

• Children’s diabetes information resources were far less specific than clinical guidelines about self-monitoring of blood glucose, and presentation of key information and messages were sometimes vague and open to interpretation (see Table 7).

• At the time of the analysis conducted for the IMP1 (2008), children’s diabetes information was not always consistent with clinical guidelines for self-monitoring of blood glucose and did not appear to be consistently based on gold standard evidence-based guidelines. However, on extending this analysis to a wider selection of resources published since the gold standard clinical guidance was published in 200426 (see Table 7), a greater degree of consistency was found.

• Insulin management varies for different children and young people dependent on factors such as lifestyle (e.g. exercise, diet), cognitive ability, culture, age, weight, school issues, adherence/concordance with treatment, child and parent choice and consultant’s preference. Approaches to insulin administration include multiple daily injection (MDI); insulin injections once, twice or three times a day; and continuous subcutaneous insulin infusion (CSII). Insulin injections may be administered by pen or syringe. Given these varied approaches and multiple factors that influence decision-making, it is not surprising that insulin management is mentioned only in vague terms in all children’s diabetes information (see Table 6).

• Most children’s diabetes resources make reference to record keeping. Children and young people are encouraged to record their blood glucose readings in their diary and this is strongly emphasised in the information targeted at parents (see Table 5).

• The risk of long-term complications is not mentioned in diabetes information for younger children aged 6–10 years and only brief mention is made in the 11–15 years’ resources.

• There is a need to update the content of children’s diabetes health information to reflect the updated gold standard clinical guidance published in 2004. 26

• NICE60 provides a lay version of its clinical guidelines but this is written in the third person and is not specifically written for children and young people living with the condition. However, in the further information section it does state that ‘If you need further information about any aspects of type 1 diabetes or the care that you or your child is receiving, please ask a member of the diabetes care team.’

• Children’s diabetes health information did not usually carry a quality badge indicating that it was suitable for use in the NHS and it was unclear what (if any) quality assurance processes had been followed.

Evidence to inform intervention development

The following important findings from this work have implications for EPIC intervention development:

• the diabetes diary is conceived as an important element of diabetes self-management

• children and young people should be developing self-management skills from an early age to titrate their insulin by daily recording of their blood glucose levels in their diary

• children’s diabetes information varies in quality and some resources better align with children’s diabetes clinical guidelines than others

• some deficiencies in children’s diabetes information are the consequence of commercial pressures and conflicts of interest from their sources

• some children’s diabetes information is obsolete and should be removed from circulation

• lack of NHS quality approval of diabetes resources needs to be addressed.

In the next section, a CDA of children’s diabetes information resources will be presented.

An introduction to critical discourse analysis

A discourse is a ‘way of knowing’ or ‘way of talking about’ some aspect of reality. 61 Two main discourses have been identified in contemporary health information: ‘patient education’ and ‘patient empowerment’. 62 We can identify the presence of both of these in contemporary health information for children; however, in line with our study aims and objectives, our focus was a CDA of empowerment in relation to managing medicines, as reflected, or not, in the texts sampled from our database of predominantly paper-based health information for children, young people and their parents.

Critical discourse analysis is concerned with the textual relationship between, on the one hand, meaning making and, on the other hand, social structure and change. In other words, what critical discourse analysts are trying to make clear is how the hierarchical power relations of particular societies shape what is, and can be, said and how it is said. Examining textual features of diabetes texts produced for children and young people thus allows us to lay bare the power relations underlying the messages found in texts. For example, by analysing such diabetes texts we are able to better understand how certain ways of talking about the condition are used and their potential consequences for the child or young person with diabetes (and those around him or her). From this we might ask how do such texts construct the relationships between the child or young person with diabetes, HCPs and parents? To what extent do such texts reproduce existing power relations between these actors? If HCPs and parents are constructed as those who have the greatest power and authority to make decisions about health-care management, is that an issue or a problem? If so, how might future diabetes texts challenge these power relations in ways that help to encourage and empower children and young people to take greater control over their own care? As such, CDA offers a powerful tool for illuminating such power relations and in so doing provides insights that may be used to rethink those relations through the range and types of information provided for children and young people with diabetes. Of course, CDA never simply examines texts in isolation from the contexts of their production, and therefore it is always important to analyse them in relation to the wider processes in which children and young people with diabetes are enmeshed – with HCPs (including doctors, nurses, pharmaceutical companies and diabetes charities), parents or carers, siblings and extended family, and teachers and peers.

Silverman63 presents the case that we should not think of texts as if they were correct or incorrect written representations of reality, but rather as accounts that help to construct the reality, and Grime and Ong take this further, stating that ‘the language of the patient information leaflet is not simply an intermediary which enables an understanding of the “truth” about health and disease but actually helps construct particular understandings of health and disease’ (p. 2). 64

Institutional discourse appears to be especially in need of critique and as particularly amenable to a CDA approach. What is meant by ‘institutional discourse’ is the ways of knowing and talking about diabetes as it is defined by the health-care profession through its policies, practices and processes in relation to children and young people with the condition. The strictly bounded and hierarchical structure of institutional organisations, and the largely textual nature of their internal mediation/communication, suggests that the production and maintenance of power relations may be to some extent ‘recovered’ by means of engagement with these texts. In the context of this study, we were interested to pose questions about the institutionalised social relationship between health-care practitioner and child or young person. Following Fairclough (p. 226),65 the kinds of questions that we wanted to explore through CDA were:

• How are children discursively positioned with regard to decision-making?

• What are the constraints on their choices?

• How are these represented?

Critical discourse analysis tools and mechanisms

Mechanisms at work in the data identified as relevant to the CDA of health information resources for children, young people and their families include intertextuality and recontextualisation. Intertextuality is concerned with describing and explaining how specific meanings are made in and across specific texts and how such texts are formally linked to each other. Intertextuality may be realised by several modes of communication– for example language or visual layout.

Recontextualisation is a useful way of understanding intertextuality as the recontextualisation of social practices. There are two social practices at issue in the analysis of health information texts aimed at children, young people and their families: (1) the practice of promoting health practices and (2) the practices surrounding the management of an illness, such as the administration of medicines, and the decision-making process. We attend to both in our analysis. With the concept of recontextualisation, we assume that text producers must use information existent in an adult context (broadly speaking) and reformulate it with a child reader in mind. They must try and understand that information in the context of childhood and construct texts accordingly.

We define a social practice as having the following characteristics: activities, participants, performance indicators, times, tools and materials, dress and grooming, and eligibility conditions. 61 In the process of recontextualisation the characteristics of social practices are subject to a variety of ‘transformations’. Box 3 provides the mechanisms of transformation that are suggested as analytical tools, that is, as ways of understanding how health information is recontextualised for children readers.

Our study of children’s diabetes information produced for children, young people and their families aims to consider how power and choice are represented, primarily in relation to medicines management (blood glucose monitoring and insulin management). A key focus for us is the legitimation of certain discourses and social practices.

Summary of the critical discourses analysis from the Information Matters Project

The CDA in the IMP1 was conducted to better understand the types and formats of health information resources using text and pictures that could empower the decision-making and choices about care of children, young people and their families. It was observed that health messages conveyed in information for children use fantasy, and in some cases medicines and/or doctors were represented as having magic powers. Children were discursively positioned with friends or knowledgeable adults and then families. As well as gaining insight into discursive mechanisms to promote empowerment, we found evidence of processes at work to create choice where choice may not seem to exist. For example, in the diabetes texts medicines are presented as necessary to the management of the condition – and in this sense there is no choice. But at the same time the illusion of choice is made available to the child in a consumer context. For example, the child with diabetes may appear to be able to choose an insulin pen from those available, some of which also allow a choice of colour. The pen may be referred to as a ‘special pen’. In reality, the insulin appropriate to the child’s treatment regimen and, in younger children, the ability to dial up insulin in half units are often determined by diabetes HCPs and will limit the child’s choice of available pens. The findings also demonstrated the importance of involving children and young people in the production of health information to establish whether key messages conveyed are received and interpreted as intended (see p. 67 of the IMP report1 for further details).

The role of the critical discourse analysis in informing EPIC intervention development

The intention was to use relevant findings from the CDA to inform the EPIC intervention development. The key issue for intervention development was to gain a better understanding of discourses used in children’s diabetes information in order to develop a programme theory. The understanding that children and young people will get about their T1D from the written component of the EPIC information pack will depend on the facts included and on the discourse employed. In addition, the shared understanding of empowerment will depend not only on how the child or young person understands the information that he or she receives but also on the ‘voice’ used to convey this information. It is with these important intervention development issues in mind that we undertook the CDA.

Research questions for the critical discourse analysis

The research questions posed for the purposes of the CDA element of the EPIC study related specifically to the discursive positioning of children, young people and their families within the children’s diabetes health information texts and to the idea of ‘empowerment’, which, in turn, related to our stated aim of identifying types and formats of information that could empower the decision-making and choices about care of children, young people and their families. The research questions posed are:

1. What does ‘empowerment’ mean in contemporary child-centred diabetes texts?

2. What textual means are deployed to appeal to the child reader?

3. How are children and/or their families discursively ‘positioned’, especially with regard to decision-making and medicines and how this changes throughout childhood?

The first step was to finalise the sample of children’s diabetes information resources to be drawn on and the processes to be put in place to undertake the CDA.

Sample of children’s diabetes health information for the critical discourse analysis

The process of selecting a subset of diabetes information for the CDA is described in Identification and selection of children’s diabetes information for quality analysis and critical discourse analysis. The final selection of children’s diabetes leaflets (agreed by the CDA working group) can be found in Box 2 (see box notes). Not all of the resources identified the target age range for which they were intended. The research team therefore classified the resources into three age categories (6–10 years, 11–15 years and 16–18 years) based on the age group of the images on the cover and within the texts. A small sample of generic ‘adult’ information from the same text producers (charities and pharmaceutical companies) was obtained to address the research objectives for the CDA.

Processes involved in critical discourse analysis

The data for CDA are the words and images that are used in children’s diabetes health information. Each piece of information was examined to further explore the issues identified in the CDA within the IMP and to identify any new areas in order to answer the research questions. Quotations and examples from the texts and images were used to further illustrate the analysis when appropriate.

Finding 1: what does ‘empowerment’ mean in contemporary child-centred health texts?

Identity empowerment

Three types of identity empowerment were found in the IMP, with texts variously discursively constructing the child who has diabetes as (1) ‘normal’ – so like other children; (2) ‘special’ – different from other children in a positive sense; and (3) ‘same but different’ – like other children but with different needs from those who do not have diabetes.

Child with diabetes as ‘normal’

This particular discourse of empowerment constructs the notion that the child and young person living with T1D may live a ‘normal’ life, in other words the life of a non-‘sufferer’ (Box 6). This discourse can be established through the images contained within the texts. This can be through the use of real people or cartoon images. For older children and young people (11–18 years), the photographs are of young people at clubs, partying, listening to music and skateboarding (Living with Diabetes, BD Medical – Diabetes Care). In a text for younger children (6–10 years) and their parents by the same company, the use of photographs of happy healthy children being active and sporty gives the impression that children with diabetes can do anything they want, just like any other child.

BOX 6 - Examples of child with diabetes as ‘normal’

Diabetes doesn’t stop you doing stuff. You’ll still be doing pretty much everything you like.

Growing up with Diabetes, BD Medical – Diabetes Care, Spring 2000, pp. 6–7

With diabetes properly managed your child can lead a perfectly normal life.

Caring for Children with Diabetes, BD Medical – Diabetes Care

An example of a text aimed at younger children (6–10 years), which uses cartoon images, is Tadpole Times, the quarterly magazine produced by the charity Diabetes UK (Figure 4). Cartoon tadpoles are portrayed as enjoying different everyday ‘normal’ activities and having fun, for example cooking, gardening, playing, practising for the school Christmas play and make-believe (pretending to be astronauts or pirates).

FIGURE 4.

Front covers of Tadpole Times (images reproduced with the kind permission of Diabetes UK).

However, in the effort to include a child who is living with diabetes in the larger group of ‘normal’ children, a very circumscribed notion of ‘normality’ is constructed. In other words, a narrow and, realistically, unachievable notion of what constitutes ‘normality’ is discursively constructed by many of the texts examined. That is, because diabetic children have diabetes, they can never be exactly the same as, and thus ‘normal’ like, others who do not have the condition. Although the social and physical benefits of controlling the effects of a condition in children are undeniable, the representations of normality may be too strongly emphasised as a positive and something that children and young people with diabetes ought to try to achieve as, inevitably, in discursive and real-life terms they will always be left out of any construction or experience of ‘normality.’

Child with diabetes as ‘special’

In one text (On the Level, Diabetes UK) the focus is very much on the achievements of children and young people (11–18 years) who just happen to have diabetes. Individual stories explore the challenges and problems that these people face whilst accomplishing these achievements, often portraying a heroic narrative. It is assumed that, if one works hard enough, one can also accomplish such things (see the mountain climbing example in Figure 5).

FIGURE 5.

On the Level, autumn 2010. Young person trekking in the Himalayas (image reproduced with the kind permission of Diabetes UK).

It is hoped that these images would be a source of inspiration; the message that is given is that even this is possible. There is no mention of inhibitors or barriers that might prevent them from doing such things, for example cost. There may actually be a risk of excluding those children living with diabetes who are not inclined to participate socially in the ‘ideal’ way.

It was found that in an effort to present children and young people with diabetes as ‘normal’ there is sometimes a degree of overcompensation in their activities and accomplishments. In these instances the bar for personal physical achievement is set very high, such as in the On the Level magazines from Diabetes UK (Box 7).

BOX 7 - Examples of activities and accomplishments from On the Level magazine

Winter 2008 – boy cycling from coast to coast across the UK.

Winter 2008 – girl backpacking – Brilliant Backpacking! ‘Diabetes didn’t stop me!’

Autumn 2008 – boy doing the Snowdon horseshoe walk.

Spring 2008 – boy climbing the Alps.

Summer 2009 – boy playing rugby.

Summer 2010 – wakeboarding – girl aged 14 years.

In this sense, then, ‘special’ is discursively constructed in the texts in very specific ways, that is, children and adults with diabetes are textually positioned as participating in very challenging activities that demand high levels of physical endurance that would be extremely demanding for anyone. In discursive terms such exceptional children, young people and adults are constructed as role models for what can be achieved, and in so doing this often neglects to acknowledge that it is a rare individual, with or without diabetes, who can undertake such activities. This form of identity empowerment may, in some cases, disempower child readers who may have limited physical abilities not usually associated with diabetes.

Child with diabetes as ‘same but different’

Within each issue of Tadpole Times magazine (Diabetes UK magazine for children aged 6–10 years) there is a section called ‘Cartoon time’ (see link to intertextuality in Critical discourse analysis tools and mechanisms) that follows the life of a girl called Emma and how she manages her diabetes in a variety of everyday situations, for example during a school trip, going to the dentist, at the Christmas disco, at sports day. In these situations the child is portrayed as being the ‘same but different’ as there are extra things that she needs to do to effectively look after her diabetes. The cartoon character Emma is displayed as having some knowledge and looking after her diabetes, for example knowing when to check her blood glucose levels and how to correct them (Box 8)

BOX 8 - Examples of child with diabetes as ‘same but different’

Before sports day:

Emma does a quick blood test to check it’s ok to exercise.

It’s a bit lower than usual, so Emma has an apple to bring her levels back up.

Later on at home, Emma has an extra snack before bedtime because she’s been so active during the day.

Tadpole Times, summer 2010, pp. s5–8, Diabetes UK

In the spring 2010 edition of On the Level from Diabetes UK there is an article about a 17-year-old boy (Joe) with T1D who went to the Glastonbury music festival:

It was good to know we were all the same, returning home with wrecked tents and rucksacks of dirty washing. In my case, there was also two sorts of spare insulin, spare insulin pens, a box of needles, a blood sugar monitor, test strips, multiclix lancets, glucose tablets and a sharps bin . . . the same but a bit different.

On the Level, spring 2010, ‘Festival Fever’, pp. 6–8, Diabetes UK

Finding 2: what textual means are deployed to appeal to the child reader?

Use of images

Images are used to expand meaningfully on written language. The authors of the texts produced ‘in house’ attempt to engage the reader through the use of clip art images whereas those produced by the diabetes charities and pharmaceutical companies have the resources to use full colour photographs (Box 11).

BOX 11 - Examples of use of images

A photograph of a cereal bar to highlight:

Sometimes when I test my glucose levels they aren’t what I expect . . . that’s OK, I just use my insulin or a cereal bar to move my level back to where it should be.

Living with Diabetes. A Guide for Teenagers, Abbot Diabetes Care

A photograph of a teenage girl eating a bar of chocolate positioned in proximity to the text ‘Common reasons for high blood glucose’, followed by such reasons.

High Blood Glucose with Diabetes, Streetwise collection, Eli Lilly (Figure 6)

FIGURE 6.

High Blood Glucose with Diabetes (image reproduced with the kind permission of the Royal College of Nursing and Eli Lilly).

Finding 3: how are children and/or their families discursively ‘positioned’, especially with regard to decision-making and medicines?

Selecting appropriate existing high-quality diabetes information

The CDA provided a clear steer that the type and format of text and images are important in terms of establishing and supporting children and young people’s empowerment to self-manage their T1D:

• Diabetes texts have been constructed with the aim of empowering and motivating children and young people to manage their diabetes on a daily basis as they take the ‘journey’ of living with T1D through childhood and into adulthood.

• Understanding the ‘empowerments’ inherent in children’s diabetes information, and how ‘empowerment’ is variously defined and represented, are essential to developing a robust intervention programme theory.

• Achievement of empowerment depends on how well the child or young person understands the information that he or she receives through the presentation of facts and advice. As such, the selection of appropriate medically informed discourses used to convey these facts is paramount. 91

• If children and young people are to engage and subsequently be empowered and motivated to manage their T1D then their ability to understand the materials given to them will also depend to a large extent on the visual presentation of these texts.

• Age-appropriate as well as visually and verbally appealing resources are an important consideration for intervention development because they draw children and young people to them, thereby encouraging their use in the development of T1D self-management.

• Use of colour, language and typography appear key because these are features of a text that interpolate or call to the reader to engage with age-appropriate diabetes leaflets, books and other texts. Bright, primary colours are often associated with children’s story books and connote excitement, fun and something of interest to young children. Language that is easily understood by specific age groups of children also facilitates interest and engagement. Likewise, typography is important as large, varied fonts are more easily engaged with by young readers, whereas older children typically seek text that connects to their interests and exposure to other texts which may use stylised fonts, for example.

• Combining texts and pictures enhances understanding, particularly for young children. Visual representations support understanding of written text with younger age groups. With older children this combination draws the reader in through a connection with representations of young people of a similar age, encouraging them to read the written text.

• Images on the front covers of the diabetes texts for children were not necessarily related to diabetes but drew on texts already familiar to children and young people (‘intertextuality’), as they connect the children and young people to familiar images appropriate to their age group – cartoons for younger children (6–10 years) and photographic images for young people (16–18 years), with a combination of both for older children (11–15 years).

• It is important to understand the importance of reconstitutions and substitutions and additions/deletions as they are sometimes used to make, for example, insulin injections seem less frightening and more familiar (the use of the term ‘special’ or ‘whizzy’ before the word ‘pen’, for example).

• Use of the authoritarian voice mirrors the approach to management of diabetes professionals. Across the texts that we examined this mode of address is typically associated with more traditional forms of top-down communication from medical professionals who are discursively positioned as diabetes ‘experts’ in relation to their ‘patients’. Knowledge flows from those who have ‘expertise’ to those who do not but who have T1D. Such discourses are very different from those in which expertise is more diffused across medical practitioners (including doctors and diabetes nurses), children with T1D and their families.

• Finally, quality assurance of children’s diabetes information is essential and branding and legitimisation for use in the NHS is important. We found that various drug companies discursively linked their products to organisations such as the NHS, as well as to diabetes charities such as Diabetes UK, to help establish their brands as having been endorsed by such organisations. By association, then, their brands garnered esteem and legitimacy.

Evidence to inform development of the EPIC intervention programme theory

• Identity empowerment of diabetes as ‘normal’, whether a cartoon image of an animal or a person, a pencil drawing or a photograph, is used to connect to the intended audience. These images are in the forefront of the majority of texts, which strongly portray identify empowerment.

• Children and young people with T1D are represented as wanting to be seen as ‘normal’. Not wanting to be different from their peers was also a finding highlighted across all of the interviews conducted for the EPIC study (see Chapters 4 and 7); this point is backed up by other empirical research92–96 and is a major factor that influences self-care abilities at school as reported in the systematic review (see Chapter 3).

• Insulin is represented throughout the texts as a social enabler – ‘normalcy’ is presented as being achievable if a strict diabetes self-management regimen is adhered to. Empowerment to be normal is the dominant discourse in children’s diabetes information.

• Real-life stories are shared of children and young people in everyday situations, and in some instances empowerment is represented via the form of the child with T1D as ‘special’, which is sometimes accompanied by a degree of overcompensation through the representation of extraordinary activities and accomplishments of some children and young people with diabetes (mountain climbing, extreme sports, etc.).

• Some resources reflect that empowerment is also about being the ‘same but different’ as there are extra responsibilities that go hand in hand with diabetes self-management. As such, normalcy is discursively delimited, albeit not ultimately undermined, by such responsibilities.

• The importance of children’s age and developmentally ‘appropriate’ journey towards adult diabetes care: child working with parent with responsibility shifting to the child as he or she feels able to be more autonomous – assumptions of control, moving forward, you are not alone, transitions into adulthood.

• Family-orientated paradigms locate children and young people primarily within the family (particularly preteen children); it is in the social context of the family that age-appropriate diabetes management is seen to be best undertaken.

• The need for active promotion and facilitation of children’s diabetes resources by diabetes professionals in routine care.

• Diabetes risks and diabetes-related complications are presented differently in children’s and adult diabetes information resources. In children’s information resources risks do not feature prominently whereas in adult information they are highly prominent. It seems reasonable to presume that outlining risks, especially for young children with diabetes, is not regarded as desirable in terms of the type of information that they need to have to manage their condition. The absence or marginalisation of risk discourses is founded on prevailing Western conceptions of the child and his or her vulnerability and thus need for protection from difficult facts, situations and knowledge. How this might impact on children’s and young people’s conceptions of their condition and empowerment to self-manage is worthy of further exploration in future research.

The next chapter reports the systematic review.

Policy and best practice context in the UK and Europe

In the UK the DH120 recommends that all children should be supported to manage their diabetes according to their individual health-care plan and that school and early years settings should be encouraged to offer effective levels of support so that parents do not have to attend school to administer medication. In the UK the use of individual health-care plans in educational settings is recommended to ensure that school staff are sufficiently informed about a pupil’s medical needs and, in relation to T1D, they should describe all parties’ responsibilities regarding diabetes supplies and provisions. 121 Common elements that current guidelines and policy documents27,33,56,120–125 recommend for the management of children and young people with T1D in school are shown in Box 26.

In the UK there is no legal duty requiring school staff to administer insulin to children and young people with T1D. However, in an emergency situation or under certain circumstances school staff might be expected to administer insulin or take other appropriate action. This is seen as a voluntary role, but school staff who are responsible for children and young people with T1D on a daily basis have a responsibility to ensure that children and young people remain safe and healthy while on school premises. 126

A DH working group examined the current challenges surrounding children and young people with diabetes, producing the report Making Every Young Person with Diabetes Matter. 120 The need for effective, integrated and accessible services for children and young people and their families was set out in this document.

Policy and best practice context in the USA

In the US, federal law requires schools to have a trained nurse available to manage children with T1D, and the most recent guidelines for diabetes management state that at least one adult should be available for all diabetes management needs if a school nurse is unavailable. 127

In the US a number of key publications set out the components of diabetes care at school. 128,129 The American Diabetes Association128 has produced a position statement to ‘provide recommendations for the management of children with diabetes in the school and day care setting’ (p. S76) and the National Diabetes Education Program publication Helping the Student with Diabetes Succeed. A Guide for School Personnel129 is designed ‘to educate school personnel about effective diabetes management and to share a set of practices that enable schools to ensure a safe learning environment for pupils with diabetes’ (p. 1).

The recommendations within these documents are not necessarily required by the federal laws enforced by the US Department of Education for each student with diabetes. State and local requirements in the USA vary from state to state and from school district to school district. But it is recommended that they be used in conjunction with federal as well as state and local laws.

In the USA a Diabetes Medical Management Plan (DMMP) outlines how each student’s diabetes will be managed. The DMMP should be used as the basis for the development of an Individualized Health Care Plan (IHP) and Emergency Care Plans for Hypoglycemia and Hyperglycemia within a school setting. Common elements that the plans are likely to address are shown in Box 26.

Objectives

1. To undertake a review of the effectiveness of interventions that are conducted within an educational setting that seek to improve the care and management of children and young people with T1D (stream 1).

2. To explore the experiences of children and young people with T1D and those involved with their care and management in an educational setting (stream 2).

3. To conduct an overarching synthesis of the first and second objectives to determine the extent to which interventions address the barriers identified by children, parents and teachers and build on the facilitators for providing optimal care and management of children and young people with T1D in educational settings (stream 3).

Inclusion criteria

A modified version of the SPICE framework132 was used to inform the systematic review. For stream 1 this was Setting, Population, Intervention, Comparison and Evaluation (Table 9) and for stream 2 this was Setting, Perspective/People, Issues of interest, Context and Evaluation (Table 10).

For stream 1 (intervention studies) any RCTs were considered. In the absence of RCTs other research designs, such as non-RCTs, clinical trials, cohort studies, experimental and non-experimental studies, observational studies, descriptive studies and before-and-after studies, were considered for inclusion.

For stream 2 (non-intervention studies) all studies reporting the views of participants (includes both surveys and qualitative studies) were considered.

Types of participants

Studies were included if they focused on children and young people with T1D within an educational setting and included those aged 3–18 years in preschool or formal education and those aged 18–25 years in post-compulsory education. A search was conducted to determine the age range of young people with diabetes who were participants within studies conducted in post-compulsory educational settings to inform the inclusion criteria of this review. The minimum age for these studies was 18 years and the maximum age reported was 24 years. 133–138 In addition, studies including or focusing on parents, school personnel and school health professionals that related to this age group were included.

Exclusion criteria

Studies were excluded if there were no before-and-after measures (stream 1); they did not directly obtain the views of children and young people, parents, peers or professionals (stream 2); the children and young people were aged < 3 years or > 30 years (streams 1 and 2) or they were not in the English language (streams 1 and 2).

Study selection

An initial scoping search was conducted in MEDLINE and Scopus using keywords drawn from the natural language of the topic. The preliminary keywords that were searched are displayed in Box 25.

The databases searched for relevant studies over the last 15 years (1996–2011) were:

• Cumulative Index to Nursing and Allied Health Literature (CINAHL)

• MEDLINE

• Scopus

• British Nursing Index

• The Cochrane Library

• EMBASE

• PsycINFO

• Web of Science.

The final search strategy is summarised within the modified SPICE table (see Appendix 3, Search terms presented within the SPICE framework).

The search terms included medical subject heading and ‘free-text’ terms in combination and was adapted according to the particular database (for a sample of the searches see Appendix 3).

All studies identified were assessed for relevance to the review based on the information provided in the title and, when available, the abstract. For studies that appeared to meet the inclusion criteria, or in cases in which a definite decision could not be made based on the title and/or abstract alone, the full paper was obtained for detailed assessment by two researchers against the inclusion criteria. A screening tool to aid this process was used (see Appendix 3, EPIC screening tool). Any disagreement was resolved by referring back to the protocol and, when necessary, by consultation with a third independent reviewer.

In addition, reference lists of retrieved papers and published reviews were searched for potentially relevant papers. References were managed using EndNote version X1 (Thomson Reuters, CA, USA). Restrictions were not applied in terms of research design or methods, as a single search was used for both stages of the review. Unpublished data were not sought from authors.

A flow chart showing the number of studies/papers at each stage of the selection process is provided in Figure 8. In total, 71 papers (describing 66 studies) were included in the review.

FIGURE 8.

Flow chart of the study selection process.

Quality assessment

The decision was made to include all studies unless they were fatally flawed and a discussion with regard to quality assessment would be reported for each stream.

For stream 1 (intervention studies), randomised intervention studies were assessed on the following quality elements:139

• adequacy of randomisation

• concealment of allocation

• blinding of outcome assessors/data analysts

• use of power calculations

• comparability of groups at baseline

• level of detail provided about the intervention

• intention-to-treat (ITT) analysis

• use of validated outcome measures

• length of follow-up

• identification of confounding factors.

Key aspects of quality for non-randomised intervention studies in stream 1 were based on the work of Deek et al. 140 and are summarised below (see p. 39 of the Centre for Reviews and Dissemination, University of York, guidance on undertaking reviews in health care141):

• how the treatment groups were created (how allocation occurred and whether the study was designed to generate groups that are comparable on key prognostic factors, e.g. by ‘matching’ participants in each group)

• the comparability of intervention and comparison groups at the analysis stage, for example whether prognostic factors were identified and whether case mix adjustment was used to account for any between-group differences

• blinding of participants and investigators

• the level of confidence that the participants received the intervention to which they were assigned and experienced the reported outcome as a result of that intervention and the adequacy of the follow-up

• the appropriateness of the analysis.

For studies in stream 2 (non-intervention studies) that used a survey design we used the checklist designed by Rees et al. 142 and for qualitative studies we used the appropriate checklists available from the Critical Appraisal Skills Programme. 143 These were then incorporated with the following quality criteria that were adapted from Kirk et al. 139 to provide a summary of quality assessment:

• clear statement of the aims of the study

• adequate description of the context for the study

• clear specification of the research design and its appropriateness for the research aims

• clear reporting of details of the sample and method of recruitment/sampling

• clear description of data collection

• clear description of the data analysis

• attempts made to establish rigour of the data analysis

• discussion of ethical issues/approval details

• inclusion of sufficient original data to support interpretations and conclusions.

Data extraction

For stream 1 data were extracted directly into electronic tables and followed the format recommended by the Centre for Reviews and Dissemination141 as outlined in Box 27.

For both streams one researcher extracted the data and a second researcher independently checked the data extraction forms for accuracy and completeness. 141 Any disagreements were noted and resolved by consensus among the researchers. A record of corrections was kept.

Data synthesis

To answer the study objectives three types of syntheses were performed. First, for stream 1 (intervention studies) meta-analysis was inappropriate for the review because of the heterogeneous nature of the studies in relation to populations, interventions and outcomes. Instead, the results from the studies are summarised and reported in a narrative summary within and across studies. Second, the process for stream 2 was a narrative analysis and synthesis using Ritchie and Spencer’s144 thematic framework analysis approach of non-intervention studies. An index ladder of codes aligned with issues of interest from the review questions and objectives was developed, refined and adapted for the different streams. Third, a final overarching synthesis of intervention and non-intervention studies was conducted. For this final synthesis a matrix was constructed that mapped the barriers and facilitators identified by children and young people, parents, school personnel and school health professionals alongside descriptions of the interventions and outcomes included in stream 1.

Design

Only three145–147 of the 11 studies were RCTs, one155 was a controlled trial, three148–150 were before-and-after studies, two151,152 were analogue experiments, one153 was a cohort study and one154 was a programme evaluation. Sample sizes were small and ranged from 20–156 participants with the exception of the cohort study153 in which the number of nurses attending the continuing education program was 417. Follow-up periods ranged from 3 months to 1 year. Studies were all published after the year 2000.

Settings

The majority of the studies (n = 10) were conducted in the USA, with one study conducted in Canada. 147

Participants

Four studies involved children and young people, which included children from kindergarten to sixth grade150 and kindergarten to eighth grade,146 those aged 5–19 years148 and those aged 10–17 years. 145 Of these, one study148 focused on children with T1D who were struggling academically or who were having difficulty managing their illness at school and was part of a wider study on chronic illness, and another focused on those with poorly controlled T1D with a HbA_1c ≥ 9%. 145 Only one study155 focused on students within a higher education setting; the intervention was delivered to three cohorts of students aged 18–27 years recruited from a university health centre. Three of the studies145,150,155 reported sex. The study by Faro et al. 150 was the only study to report ethnicity; this study included an ethnically diverse sample (55% African American, 25% Hispanic, 18% white and 1% other). A further six studies involved school personnel and the samples included both school nurses and school teachers,149 elementary school teachers,147 regular and special education elementary teachers,152 continuing education and preservice teachers151 and school nurses. 153,154

Interventions and outcomes assessed

Theoretical underpinnings

Of the 11 interventions reviewed, only two150,155 were explicitly reported as theory based.

The study by Wdowik et al. 155 utilised the theory of reasoned action and social learning theory and developed an expanded health belief model. The conceptual frameworks for the pilot study by Faro et al. 150 were based on both social learning theory and developmental theory.

Children and young people with type 1 diabetes

Three studies145,146,148 described child-focused interventions (one used a case management approach, one targeted blood glucose management and insulin dose adjustment and the last targeted school-based care and diabetes-related education) and were facilitated by the school nurse. Of these, two showed improved quality of life146,148 and in the short term all three appeared to have an effect on glycaemic control. A fourth study, facilitated by a paediatric nurse practitioner, showed no change in self-efficacy as self-reported by parents and care providers. 150 One study155 investigated the diabetes knowledge of university students using a 12-item measure specifically designed to test components of the intervention. However, the authors reported only on the content validity. The authors reported that knowledge was improved as a direct result of the intervention and was maintained at follow-up.

Quality of life was measured in two studies146,148 using the Paediatric Quality of Life Inventory (PedsQL) 3.0 T1D module. 156 This is a 28-item instrument with five subscales: diabetes symptoms, treatment barriers, treatment adherence, worry and communication. Both studies found significant improvements on the treatment barriers subscale, which assesses the extent to which children experience pain during finger prick or insulin injections, embarrassment about having diabetes, arguments about patient care and difficulty complying with their diabetes plan. The study by Izquierdo et al. 146 also found a significant improvement in the treatment adherence subscale, which assesses the extent to which children experience difficulty with taking blood glucose tests, taking insulin injections, exercising, tracking carbohydrates/exchanges, wearing their medical alert bracelet, carrying a fast-acting carbohydrate or eating snacks.

Three studies145,146,150 measured HbA_1c levels. Two studies145,146 showed significant improvements in the HbA_1c readings over a 3-month122 and 6-month146 period following the intervention, whereas the other150 showed no change. A further study155 showed a significant increase in the number of university students who knew their recent HbA_1c results. The study by Faro et al. 150 showed limited success in trying to achieve a HbA_1c value of < 7%.

The effect of the intervention on health service use was measured in two studies. This section of the analysis for the study by Izquierdo et al. 146 was poorly reported but indicated that urgent visits to the school nurse for diabetes-related problems and urgent calls to the diabetes centre decreased significantly and that there were fewer hospitalisations and emergency department visits. In contrast, the study by Faro et al. 150 did not show any significant difference in the frequency of hospitalisation or emergency department visits.

Only one study155 was conducted with young people studying at higher education establishments. The findings showed that education sessions that were based on an expanded health belief model and social learning theory had an impact on self-management behaviour, knowledge and attitudes.

School personnel

There is evidence that education-based interventions for school personnel can improve knowledge and confidence although, of the two studies in this review on education-based interventions,153,154 one showed no effect, which could be attributed to the fact that all of the school personnel in the intervention already had experience of caring for a child and young person with T1D and therefore already possessed a good level of knowledge of diabetes.

Two studies147,149 conducted their interventions with school personnel using validated, purpose-designed measures. One study149 showed a significant improvement in diabetes knowledge (only 38% of teachers had experience of children with T1D) and another147 showed no change in general diabetes knowledge or in knowledge of hypoglycaemia (100% of teachers had experience of children with T1D).

Children, young people and parents (n = 27)

Nine studies focused on children and young people,157–164,174 eight focused on parents94,165–171 and 10 focused on children, young people and parents. 172,173,175–182

Students (n = 9)

School personnel (n = 12)

School health-care personnel (n = 7)

Barriers and facilitators associated with diabetes self-management in school

Children and young people with T1D spend most of their time during the day at school. Best practice guidance26,128,129 sets out optimal ways for insulin administration and blood glucose monitoring to be integrated into children’s daily school life, as well as allowing snacks to be eaten during class time so that pupils with T1D can participate fully in all school activities. Despite clear policy and best practice guidance on self-management in schools, some children found that managing their T1D in school was stressful and that there were unnecessary barriers to self-management,177,181 especially when experiencing low blood glucose levels (hypoglycaemia)161,162,177 or high blood glucose levels (hyperglycaemia). 162

Barriers and facilitators associated with school culture

It is recommended that all children and young people with diabetes should have an agreed individualised care plan (ICP) in the UK and an IHP and Emergency Care Plans for Hypoglycemia and Hyperglycemia in the USA. Such plans, when in place, can act as important facilitators for diabetes self-care. Issues such as providing suitable locations for blood glucose monitoring and insulin administration, allowing the pupil to leave the classroom to access the restroom and water, provision for appropriate storage of medical supplies, nutritional needs including provision for meals and snacks and participation in physical activity programmes and extracurricular and social activities should all be laid out in such documents. However, between 31% and 46% of pupils did not have a written care plan. 167,168,178

School policies generally apply to the entire student body within a particular school and do not often consider the child and young person with T1D and their need to perform diabetes self-management at school. 176,180 In some schools there are generic policies for all children and young people with diabetes that fail to take into account individual needs. For example, one parent said

Not every kid will be comfortable testing in an open classroom. Not every kid will feel comfortable going to the nurse’s station to do that . . . I think kids should have more freedom to test, to manage their diabetes according to their comfort level.

p. 168176

School nurses felt that a care plan for emergencies was important for facilitating the care of a pupil with T1D in the school environment. 204 The majority of school nurses reported that they had developed written care plans and those who had worked for longer as a school nurse were more likely to report that they had done so. 204

Barriers and facilitators associated with school-based facilities

Barriers and facilitators associated with the role of school health-care personnel

The main difference between the UK and the USA is the provision of a school nurse. There are no specific recommendations regarding the role of the school nurse within the UK guidelines. The US guidelines recognise the school nurse as the most appropriate person in the school setting to provide care for the student with diabetes. Each pupil with T1D should have permission to see the school nurse and other trained school personnel on request. 128

Barriers and facilitators associated with the role of school teachers

Barriers and facilitators associated with self-management at college/university

Just over half of students (55%) reported that having diabetes interfered with being a student,135 with the majority (71%) of students finding it more difficult to manage their diabetes in college than in high school,133 although they learned to overcome these difficulties with time and experience. 186 Students’ own perception of blood glucose control at college was variable, ranging from ‘better’ (36%) to ‘worse’ (33%), ‘no change’ (26%) and ‘didn’t know’ (5%). 133 Clinicians’ rating of metabolic control was also variable, ranging from ‘improved’ (30%) to ‘worse’ (30%) and ‘no change’ (39%). 133 Assessment of HbA_1c readings between high school and college133 and between the first year and the last year of attendance at university138 showed no significant change. However, the prevalence of retinopathy was found to have increased from 10.9% to 14.5% by the time that students had finished university. 138

The reasons selected for any type of change (positive or negative) in college students’ diabetes control were diet, exercise, frequency of blood glucose monitoring, increased responsibility, irregular schedule, fear of hypoglycaemia, alcohol use, no parental involvement and contact with health-care provider. 133 Students had difficulties balancing the effects of having diabetes with the student lifestyle. 135 Some students reported reduced participation in social events189 whereas others felt that having diabetes did not affect their social life. 135 The lack of a perceived routine in university was considered a barrier to effective self-management,136 with students reporting little or no time to engage in practices such as blood glucose testing,135,136,184,187,189 exercising,135,136,187 eating snacks during the day187,189 and injecting in a suitable environment. 189 Inadequate finances was also cited as a barrier to successful diabetes management135,136 and depended on how much extra support students received from parents or whether they lived in halls of residence of not. 135 Psychosocial issues were also identified as barriers to diabetes management. These included the inconvenience of diabetes management, motivators for managing diabetes, and social support issues. 136 Students with a positive attitude and good intentions may be unable to engage in desired self-care behaviours when significant barriers or negative emotions are present. 187

In a desire to appear normal136,185,189 students try to avoid hypoglycaemia in a social situation183,189 and often make a conscious decision to ‘run high’ rather than keep their blood glucose at the recommended level. 189 Many students (especially first-year students) drank alcohol in a way that was risky for their diabetes control in order to identify as ‘normal’ young students. In a further attempt to appear normal most students acknowledged that they were not as adherent to diet recommendations as they thought they should be. 136,207 It was felt that being a student could affect the type of food eaten, as eating a few more pizzas and take-away meals was all part of being a student. 135 Furthermore, many T1D students did not plan ahead for meals and snacks. 136

The majority of students reported that they engaged in alcohol consumption practices whilst at university,135,183,188 often consuming five or more drinks in one sitting. 188 Drinking perceptions included ‘students with diabetes can drink if they are careful’, ‘drinking is the primary social activity at this university’ and ‘the peer pressure to drink is strong’. 190 There was evidence that as students experienced transitions within university their attitudes towards the risks of drinking changed and, in many cases, their drinking decreased substantially after the first year. 183

Students reported the following strategies when engaging in alcohol consumption practices: eat before and/or during drinking,136,183,188,190 keep track of the number of drinks consumed,188 determine in advance not to exceed a set number of drinks,136,188,190 avoid drinking games,188 have a friend let you know when you’ve had enough,188 alternate non-alcoholic and alcoholic beverages,188 pace drinks to one or fewer per hour,188 choose not to drink alcohol,190 drink an alcohol lookalike (non-alcoholic beer, punch, etc.)188 to avoid or diffuse peer pressure,190 limit the frequency of going out to parties and bars,190 drink with trusted friends136,190 who understand the symptoms of low blood sugar that would mimic intoxication136 and check or monitor blood glucose levels. 184,190 Students reported that they usually tested at home, in more private spaces, before and after going out, rather than while they were out. 184

Younger students mainly engaged in alcohol consumption practices that were risky for their diabetes control in public spaces where they felt that there would be a risk to their identities as normal young people if they did not engage in these practices. 183 Younger students were more likely to rely on a friend to let them know when they’d had enough and to keep track of how many drinks they were having. 188 In contrast, the strategies of pacing drinks and avoiding drinking games were more common among older students. 188

Students felt that they did not receive adequate support from their college or university and/or their diabetes team to enable them to balance the demands of further education and the management of their condition. 189 Some students preferred to continue to receive care from their diabetes team in their home town. 135

Best practice guidance and empirical evidence on producing health information

We also consulted best practice guidance and empirical evidence on producing health information and, although little evidence was located specifically on producing children’s health information, some general best practice principles were helpful in conceptualising the design features that constitute a ‘good’ information resource64,74,215–221 (Box 28). We used these principles in age-appropriate ways to develop the EPIC intervention (see Figure 10).

Qualitative fieldwork with children and young people and their families

We sought to conduct approximately three focus groups (see Focus groups) and 20 interviews (see Semistructured interviews with children, young people and parents) to explore children’s perspectives on currently available children’s diabetes information resources and to ascertain the diabetes information needs of children and their families and how and when they use diabetes information resources to support decision-making concerning choice and self-care. We also sought to interview children and young people with T1D who live away from their families in the short, medium or long term (the ‘hard-to-reach’ population). We had previously conducted 11 interviews with parents and eight interviews with children and young people with T1D for the IMP. 1 For the purposes of intervention development, a further 12 interviews with children and young people, with or without their parents, were undertaken to gain additional perspectives on diabetes information use and preferences.

Various approaches were used to identify children and young people with T1D and their parents and recruit them into the study. The study was registered with the MCRN, the Diabetes Research Network (DRN), the Children and Young People’s Research Network and the Clinical Research Collaboration Cymru (CRC Cymru) (now the National Institute for Social Care and Health Research Clinical Research Centre) to utilise their resources and expertise in gaining access to study sites and facilitating recruitment. Children and young people with T1D aged between 6 and 18 years were recruited from hospital outpatient clinics in sites not involved in the subsequent RCT and in locations that could potentially provide an opportunity to recruit from a range of socioeconomic and ethnic groups. This process was facilitated by the MCRN and CRC Cymru.

Focus group method

Age-appropriate focus groups222 were facilitated to gain an in-depth understanding of what and who influence children and young people in terms of conveying T1D information and health messages and to capture their views on types and preferred formats of diabetes information resources.

Sampling strategy for focus groups

Children and young people with T1D aged between 6 and 18 years were recruited to participate in the focus groups. Our sampling strategy was to identify 8–10 children for each focus group across three age groups: 6–10 years, 11–15 years and 16–18 years. Local diabetes HCPs, with the support of MCRN nurses, identified children and their families, who were then sent a study information pack on behalf of the EPIC research team. Children and their families who were interested in participating completed contact forms and returned them to the research team who then made contact and arranged attendance at the focus groups. Focus groups were conducted in suitable, local, child-friendly venues, for example leisure centres.

It is widely recognised that engaging children and young people with T1D in group activities related to diabetes, particularly outside the clinic setting, is notoriously difficult. Children live with diabetes on a day-to-day basis and dislike it spilling over into their social life. Recruiting them to participate in a research study is even more challenging given the need to work through third parties, for example HCPs, whose enthusiasm, commitment and goodwill are imperative to facilitate the process. 223 The final participant numbers are described in Focus group findings.

Facilitation of focus groups

Focus groups were structured around a range of age-appropriate and child-centred activities to facilitate interactions with and the contribution of children (see Appendix 4, Example of focus group schedule: children aged 6–10 years). The focus groups were video recorded and lasted for approximately 2 hours, excluding time for refreshments/debriefing. Photographs were taken of data collected on wall charts. Two researchers were present; one facilitated the discussion whilst the other operated the video recorder. MCRN nurses supported administration and registration. Participants received travel expenses and a £20 high street voucher as a ‘thank you’ for their time.

Resource selection for focus groups

Media sources of information (e.g. DVDs, television, internet, books, leaflets) were used to stimulate discussion with children (the scoping exercise of currently available children’s health information has been reported in Chapter 2). A selection of available age-appropriate diabetes information resources was chosen to provide a contrast between different types of format (see Appendix 4, Information resources used as part of the focus groups).

Thermometer rating scale task

During the focus group session, the facilitator asked a question about what kind of diabetes information children and young people would like to receive in a specific context and how they would like to receive it. The thermometer rating scale task involved showing children and young people a large printed thermometer rating scale, which was placed on the wall, and a range of laminated images of possible providers and types of diabetes information. The resource images included leaflets, books, DVDs, a support group, a home telephone, a mobile telephone, a photograph of a nurse, a photograph of a doctor, an internet messaging board, a web forum and an e-mail.

Following presentation of the ‘provider’ image, children and young people had to say or demonstrate what they thought of the method of communicating the specific type of diabetes information in a specific context. Children would shout a number from the thermometer rating scale or demonstrated with laminated ‘thumbs up’ or ‘thumbs down’ signs to indicate whether they liked getting information in that way or not. Consensus agreement was reached before the facilitator placed the appropriate image on the wall-mounted thermometer. At the end of the provider rating task the facilitator summarised the findings and asked the children whether they wanted anything changed. The facilitator then displayed the final agreed image position on a thermometer rating scale using sticky tack (Figure 11).

FIGURE 11.

Wall-mounted ‘thermometer’ for rating scale task.

Qualitative analysis of focus group data

Focus groups were video recorded and the audio data transcribed verbatim. Ritchie and Spencer’s144 thematic framework analysis approach was used to analyse the data using ATLAS.ti™ software (version 6.2.27, ATLAS.ti, Berlin, Germany). This involved the following five steps: familiarisation (reading transcripts and listening to tapes in detail to gain an overview of content), thematic analysis (developing a coding scheme), indexing (applying the codes systematically to the data), charting (rearranging the data according to the thematic content to allow comparative analysis) and mapping and interpretation (defining key concepts, delineating the range and nature of phenomena, creating typologies, finding associations, providing explanations and developing strategies). Guided by this approach, coding of focus group data was conducted by two researchers and discussed with other members of the team. The a priori coding framework of interest was broadly information types, characteristics and preferences.

Real narratives and images

A preference for information presenting real narratives and realistic (as opposed to cartoon) images of children and young people was a strong theme. For example, when referring to the book Becky has Diabetes for the 6–10 years age group and Getting a Grip on Diabetes for the 11–15 years age group, participants stated that:

It was good, because it was about a day, and she injects her insulin . . . . It shows you people injecting insulin and it’s real.

6–10 years, girl 1

It’s like showing a real girl with diabetes.

6–10 years, girl 2

It’s helpful for someone who’s new to diabetes, because it tells you how she does it at home and tells them how they could possibly do it at home.

16–18 years, boy 1

It might be a story but it’s actually some information . . . . I like looking at stuff like, like stories, like you can get like, from other people sort of thing.

11–15 years, girl 1

Information tailored to age and lifestyle of the individual

The importance of tailoring health information resources to age and lifestyle emerged as a result of showing children and young people resources intended for use by their age group:

Because it’s [Streetwise leaflets] more like a booklet, it’s not thin and . . . it’s got a lot more information in it [smoking, drugs, solvents, help advice, contact numbers] . . . yeah I think I’d pick that one up.

16–18 years, girl 1

you know as soon as you see it [Streetwise leaflets], what it’s about really, and also it’s not like filled full of cartoons all over it, it’s got one or two pictures, which is okay, so you sort of know it’s aimed around you know what age level sort of thing that’s . . . well, teenage years.

11–15 years, boy 4

When reviewing the resources shown, participants often identified resources as being ‘too young’, ‘too old’ or ‘adult’. It was clear that children’s and young people’s preferences in the 6–10 years and 11–15 years focus groups did not always ‘fit’ with the ubiquitous target age bands set by the information producers:

I would have when I was younger but not now no [T1 magazine, JDRF].

11–15 years, boy 5

plain and boring and no pictures . . . For adults [named NHS trust leaflet].

11–15 years, girl 5

Like six or seven . . . cartoony . . . When I was little but not as much now [Getting Started with Diabetes, BD Medical – Diabetes Care].

11–15 years, girl 1

But the cartoons, err, are well drawn but seem to be for younger children. And they distract you, and you’re looking more at the cartoons rather at what’s actually in the book [Managing our Diabetes, Roche].

11–15 years, boy 4

Children appeared very conscious of the connection between age group and what would be considered appropriate in terms of information, for example the balance of text and words to pictures, drawings or cartoons running through the text, with younger children preferring less text:

Well, especially on some of these like, something like that which has pictures, which makes you understand it quite a lot more.

6–10 years, boy 3

I would read it [Type 1 Diabetes in Children, Adolescents and Young Adults, Ragnar Hanas] but I wouldn’t really enjoy it because it’s just basically lots of text and not much pictures.

6–10 years, girl 2

One boy described being given a diabetes diary in which to record his blood glucose readings but found that it was not geared to his age and as a result he didn’t understand it and stopped using it:

Well, I tried it, but it was, erm, it was like, I ended up just. I filled it in, but I didn’t really understand it, I just saw it as I was just putting numbers on a piece of paper really. So it was more. I still think it was very, err, mature and adult, it was going into all these like . . .

Technical details was it?

Yeah, which I was all right with, but it just, as soon as you start hearing hyperglycaemic and all that you know, you tend to switch off.

11–15 years, boy 4

Younger children (6–10 years) commented that a diabetes record diary that was suitable for their age group needed to have spaces that were big enough to write in:

I don’t think that you could write a lot in it because it is so small. It’s too small gaps [Hangin’ with Hu-mee, Eli Lilly].

6–10 years, girl 1

I had one of those. Err, but not like the book, but a diary type of thing, but the gaps were bigger on mine.

6–10 years, boy 1

Information organised into sections

Older children (11–15 years) wanted to be able to turn directly to information of interest through the use of an index page, and having clearly labelled sections was a useful feature of thicker information booklets and resource folders:

It’s [resource folder developed by Gwent Healthcare NHS Trust] actually pretty good. Um, it’s very well organised, yeah, everything has got nice sections in it, so like, travelling with diabetes, um, then it’s got an index which tells you what it’s about, before you actually even look at it.

11–15 years, boy 4

I thought this book [resource folder developed by Gwent Healthcare NHS Trust] was quite useful to me, it’s in sections but you can always find what you want.

11–15 years, boy 3

Overall likes and dislikes with regard to diabetes information

A summary of the overall likes and dislikes of children and young people with T1D by age category is presented in Table 22.

Thermometer rating scale task: information type and provision preferences

Findings from the thermometer rating scale task are presented by age category. Children in the 6–10 years focus group identified a preference for receiving information about T1D from their diabetes HCP. They wanted to know more about what caused T1D and how injecting insulin helped to make them feel better. The order of preference for other forms of information provision was the internet, leaflets, books, support groups, e-mail, web forums, DVDs, message boards and by telephone.

Similar patterns of information type and provider preference were identified by young people in the 11–15 years focus group. This age group wanted to have more information about growing up and their changing body and things that happen to them as they grow up. The consensus was that they preferred to obtain information about T1D from their diabetes HCP followed by the internet and web forums, leaflets, books, internet message boards, by e-mail and by telephone.

As previously reported, individual interviews were undertaken with a 10-year-old boy and a 16-year-old girl because of their lone attendance at separate focus groups. They also participated individually in the thermometer rating scale task and presented different preferences from those identified by children and young people in the focus groups. The 10-year-old boy preferred to obtain information about T1D from books, closely followed by diabetes HCPs and support groups. His order of preference for other forms of information provision was the internet, leaflets, DVDs, by home telephone and by mobile telephone. He said he would not use e-mail, web forums or internet message boards as sources of information. In line with the 6–10 year olds in the focus group, he wanted more information on what actually happened inside the body, especially what the pancreas did.

The 16-year-old girl preferred to obtain information about T1D from leaflets, followed by support groups and DVDs. Her order of preference for the other forms of information provision was websites, by e-mail, diabetes HCPs, books, by home telephone and by mobile telephone. She said that she would not look for information on web forums. She specifically wanted more information about carbohydrate counting.

Findings from these two single interviews emphasise that children and young people have very individual and age-related preferences concerning the type and mode of information provision. Choice may also depend on their relationship with their PDSN and engagement with their diabetes team.

Summary of the focus group findings relevant to EPIC intervention development

• Children wanted diabetes information resources that are organised into age-appropriate packs and that reflect how diabetes is experienced in ‘real’ life.

• Children wanted age-appropriate and interesting images/formats.

• Children wanted to be able to choose from a wide selection of resources that reflect individual interests and lifestyles beyond the specific diabetes focus.

• Children wanted information to be organised into sections.

• Those aged 6–10 years wanted information on diagnosis and basic diabetes management and how the body worked.

• Those aged 11–15 years wanted information on growing up and their changing body and things that happen to them as they grow up.

• Children aged 6–15 years primarily wanted to receive information from their PDSN.

• Some wanted information on carbohydrate counting.

• Diabetes diaries need to use appropriate non-technical language with sufficient space to record information.

These key findings are incorporated in the summary table of evidence informing EPIC intervention and programme theory development (see Table 20).

Interview method

A generic qualitative approach224 with in-depth semistructured face-to-face interviews with children and key family members was used to gain a greater understanding of their diabetes information needs, with an emphasis on medicines management and service delivery. We arranged interviews in a suitably convenient location for the family, for example at home, on a NHS site hospital ward, in an outpatient or other clinic or in a primary care setting. Interviews were conducted individually or jointly with the parent of a child or young person with T1D.

Sampling strategy for interviews

Children and young people aged 6–18 years with T1D were recruited to take part in the face-to-face interviews.

For the recruitment of the 11 parents and eight children and young people interviewed as part of the IMP, see Chapter 4 (p. 107). 1 The sampling strategy for the EPIC interviews was to identify 10–12 children and young people across three age groups: 6–10 years, 11–15 years and 16–18 years. Children and young people and their families were identified by local diabetes HCPs with support from CRC Cymru nurses. Children and young people were sent a study information pack on behalf of the EPIC research team and those who were interested completed contact forms and returned them to the research team. Interviews were arranged in a convenient location for the family, which in all but one case was the family home (see Appendix 4 for an example of an interview schedule).

An additional five parents were recruited when their child attended one of the diabetes focus groups. These parents were interviewed either on the day of the focus group at the location where the focus group was held or in their home on a separate day.

Children and young people classified as ‘hard to reach’

‘Hard-to-reach’ children and young people with T1D who lived away from their families in the short, medium or long term were conceptualised in the funder brief as those who were homeless, who had experience of places such as youth offender institutions and who were socially marginalised. Additional recruitment strategies were employed to try and access such children and young people. One national press release, co-written by the Involving People representative on the study, was picked up by the Big Issue, teletext and other regional newspapers. We also used local press releases (one in South Wales and one in North Wales) and advertisements on electronic university noticeboards (Cardiff University and Bangor University). Both Welsh- and English-language press releases were distributed to press agencies in Wales. The research team also asked PDSNs at EPIC trial sites if there were any patients on their caseloads who fulfilled the criteria of ‘hard to reach’. Children and young people who were interested were asked to contact the research team and study information and consent forms were sent to those who responded. Once completed consent forms had been received, arrangements were made to conduct a face-to-face or telephone interview at a mutually convenient time.

Resource selection for interviews

A selection of age-appropriate paper-based diabetes health-related information that had been used in the focus groups was taken to the interviews to prompt discussion (see Appendix 4, Information resources used as part of the focus groups).

Qualitative analysis of interview data

Interviews were digitally recorded and transcribed verbatim. Ritchie and Spencer’s144 thematic framework analysis approach was used with ATLAS.ti software. Initially, three transcripts were read and reread by the two researchers and the major themes that were related to the research questions were noted. A coding frame was then developed in discussion with the research team. This initial coding frame was applied to five different transcripts and amended during this process to include new emergent issues. Using the revised coding framework, all transcripts were then coded or recoded.

The coding framework represented three overarching areas of investigation, closely aligned to the study objectives, namely information needs, information characteristics and information preferences (Box 29). Interview transcripts were coded by two researchers and discussed with other members of the research team.

Diabetes information needs

Most diabetes information was given to children and young people at diagnosis, and if children were very young diabetes information was primarily aimed at their parents.

In the following sections quotes are designated a code indicating the unique number of the child, his or her age group, if interviewed for EPIC, whether he or she had been diagnosed for > 2 years or < 2 years and sex.

Information characteristics

Children, young people and parents received information about T1D from:

• PDSNs

• child-focused age-banded magazines (Diabetes UK, JDRF)

• peer support groups

• parental support groups

• leaflets (NHS trust in-house leaflets, charities, pharmaceutical companies)

• DVDs

• the internet

• social networking sites

• specialised books (e.g. carbohydrate counting)

• insulin injection teaching aids [e.g. JRDF teddies (Ruby and Rufus), Medtronic lion (Lenny the lion) or an orange]

• educational diabetes camps and activity fun days.

Information preferences

Summary of interview findings relevant to EPIC intervention development

• Diabetes information provision was patchy with most information received around diagnosis.

• Children and young people receive/like to receive most of their information from PDSNs.

• Parents took primary responsibility for the diabetes management of younger children.

• Children were taught to take more responsibility as soon as appropriate and especially when attending school.

• Younger children wanted information about the causes of diabetes.

• All children and young people expressed the need for individually tailored, age-appropriate and updated information as needs changed over time.

• The size of a diabetes record diary is important; children need to be able to write in it and young people want to be able to put it into their back pocket.

• Young people needed information on a wide variety of lifestyle issues such as alcohol consumption, pregnancy and taking part in sporting activities.

• Information leaflets for quick access to relevant information were preferred to more comprehensive textbooks.

• Most children and young people and their families were not aware that information was available in other formats, for example DVDs, and were not confident about knowing where to look on the internet for accurate information.

These key findings are incorporated in the summary table of evidence informing EPIC intervention and programme theory development (see Table 20).

The next section moves on to report the methods and processes used in the design and production of the EPIC children’s diabetes information packs and diaries.

Ongoing consultation with a clinical expert group

A clinical expert group (for details of members see Members of the children’s diabetes clinical expert group for the EPIC trial) was convened to advise on the development of the EPIC diaries and to clarify:

• strengths, weaknesses and gaps in current children’s diabetes information provision and service delivery

• the need for a new set of age-appropriate diabetes diaries and quality-assured information packs to fill known gaps

• the aims, purpose and anticipated outcomes of the EPIC intervention

• options for type, structure, format and clinical content

• current best/standard practice in children’s diabetes documentation

• risk management and clinical governance

• implementation, facilitation and delivery of the intervention in routine clinical practice

• training needs in use of the packs and diaries through launch events.

The group confirmed the need for diabetes diaries and the proposed size of diaries. Members made suggestions about the key areas that information packs should cover and felt that lifestyle issues should be included in the 11–15 years pack as well as in the 16–18 years pack. Separate information was required for children and young people on different types of insulin regime (e.g. insulin pumps).

Development of the EPIC intervention content and associated programme theory and logic

The core EPIC research group met regularly to develop the EPIC intervention and associated programme theory and logic. The group comprised the two principal investigators (PIs), a media specialist, a children’s diabetes specialist nurse, a sociologist, two research officers and a research assistant.

As reported in Chapter 1, two theoretical models of age-appropriate partnership and participation, and the critical success factors for implementation of children’s health information in routine clinical practice were developed at the end of the IMP1 (see Tables 1 and 2). A theoretical proposition was also refined, that for optimal translation of children’s health information and participation in care:

High quality, age-appropriate and accessible children’s health information, valued by health care professionals, children and parents, and shared using child-centered, facilitative and partnership approaches to care and service delivery, will be more highly used and a more effective component of complex interventions to optimize children’s long-term condition management.

p. 2071

With reference to the study proposition and the theoretical models of partnership and participation and the critical success factors for implementation of children’s health information in routine practice, the group refined the scope and purpose of the EPIC intervention, which took the form of a children’s diabetes information pack in a folder and a diabetes diary. The diabetes diary (to be produced by the team) was conceived as the central component to optimise children’s blood glucose monitoring and insulin management. At the same time, thoughts were focused on how the intervention was intended to work (programme theory) and the sequence of events and ‘fit’ of the intervention with routine clinical care (programme logic).

The core research group commenced intervention development with a set of evidence-based principles for developing patient information (see Table 20) and initial elements of programme theory and logic (Box 31) derived from key contextual work and selected evidence (summarised in Table 20) and in ongoing consultation with the expert group. The EPIC diabetes information packs and diaries were intended to increase children’s self-efficacy in diabetes self-management with a specific focus on blood glucose monitoring and insulin management in the following ways:

• active promotion of the pack and endorsement of children’s ownership, and ongoing active referral to the EPIC pack and diabetes diary by diabetes professionals and parents would indicate to children and young people the high value that professionals and parents placed on the diary and pack

• by owning and personalising an individually tailored and age-appropriate EPIC pack, children and young people would be more likely to access it and use the quality-assured information to inform their decision-making about daily self-management of their diabetes, and would refer to this information in the first instance rather than waiting to speak with a PDSN or waiting to raise a question about self-management at the next clinic appointment

• by recording their blood glucose readings regularly on a daily basis in their age-appropriate diabetes diary children and young people would be able to identify trends and manage their diet and lifestyle by titrating their insulin dose accordingly to achieve optimal glycaemic control

• the dominant discourse of ‘normalisation’ of diabetes and use of insulin as a social enabler would motivate children to optimally manage their diabetes

• age-appropriate behaviour change towards optimal glycaemic control was viewed in the context of the lives of children and their families at home and school, etc.

Once clarified, a logic model was developed and refined (see Chapter 7, Figure 35, which reports the process evaluation).

Further, more nuanced age-appropriate aspects of the programme theory were articulated once the EPIC packs and diaries had been developed; these are presented in Further refining the age-appropriate elements of the programme theory and logic of the EPIC children’s diabetes diaries and information packs.

The next task was to decide what the EPIC packs should look like, what they would contain and what information resources would need to be produced in-house.

Ongoing consultation with children, young people and their families

Members of the core EPIC study research team attended a number of events organised by diabetes charities, Diabetes UK and the JDRF and had permission from the organisers to seek children’s and young people’s opinions regarding the draft images and sizes of the EPIC diaries. A sheet was also distributed to ascertain what information (apart from space to record blood glucose readings and insulin dosage) they would like in the diaries and what top 10 tips they would give children and young people with T1D to help them look after themselves.

Additional consultation and advice was provided by a young university student with T1D (an advisor to the EPIC project) and a mother of a teenager with T1D (on the EPIC project steering group committee).

Development of artwork for the diabetes diaries and sticker sheets in conjunction with the medical illustrator

In the interviews and focus groups the two younger age groups, 6–10 years and 11–15 years, expressed a preference for cartoon characters that related to them. The medical illustrator also had experience of developing artwork for children’s health-related projects and designed the cartoons for the youngest age group to cater for their preference for round shapes and bright colours and fun characters that looked like soft toys, which children felt were reassuring and safe, and so that they did not reflect ‘diabetes’ as the discourse on the outer cover.

The cartoons and colours for the 11–15 years age group were designed to appear more complex and grown-up whilst still looking fun and being in accordance with popular television cartoons for a similar age range.

The 16–18 years age group considered cartoons to be childish and irrelevant to them as young adults. They were interested in the opinions of other teenagers and considered photographs to be more appropriate. Photographs of students were taken, which conveyed camaraderie and support for one another. The models were local teenagers studying for their A levels, who were paid a £10 fee. Each individual photographed signed a level 2 consent agreement to allow us to use the photographs for the EPIC project only.

Diabetes diaries

Three age-appropriate blood glucose monitoring and insulin management diaries [age ranges 6–10 years (Figure 12), 11–15 years (Figure 13) and 16–18 years (Figure 14)] and a specific diary for all age groups for children and young people on insulin pumps (Figure 15) were developed.

FIGURE 12.

EPIC diary for children and young people aged 6–10 years.

FIGURE 13.

EPIC diary for children and young people aged 11–15 years.

FIGURE 14.

EPIC diary for children and young people aged 16–18 years.

FIGURE 15.

EPIC for children and young people on insulin pumps.

Production of text for diabetes diaries

Continuing the discourse of ‘normalisation’ and using an authoritarian voice, clinical information (e.g. sick-day rules) for the diaries was developed in conjunction with the members of the clinical expert group.

‘Top tips’ for optimal diabetes self-care, provided by children and young people with T1D, were highlighted on the bottom of the diary pages. Development of the pump diary necessitated additional support from children’s diabetes specialist nurses from two local NHS trusts.

The final EPIC diaries contained key health and self-management messages that were intended to motivate positive diabetes self-management behaviours in children and young people. The language used in each diary reflected the age of the target audience. It was anticipated that children and young people would carry the EPIC diary with them at all times and use it on a daily basis to optimise self-management. A reminder to this effect was added to the diary covers: ‘Take me with you wherever you go’ was printed on the front of the 6–10, 11–15 and pump diaries, and ‘Always take your diary with you’ was printed on the back of the 16–18 diary.

Personalisation

Sticker sheets that matched the images on the front of the EPIC diaries (Figures 16 and 17) were added to the 6–10 pack and the 11–15 pack. The 11–15 pack also contained a coloured permanent marker pen.

FIGURE 16.

Sticker sheet for the EPIC diary for children and young people aged 6–10 years.

FIGURE 17.

Sticker sheet for the EPIC diary for children and young people aged 11–15 years.

The purpose of the stickers and pen was to encourage children to personalise the pack, hence giving a sense of ownership. The assumption was that, if children felt ownership of the EPIC pack, they would be more likely to use it routinely when they needed to check out information and learn about diabetes.

Signposting to further information, branding and carbohydrate counting sheets

The EPIC team developed age-appropriate further information sheets, which contained details of books and websites. An EPIC logo sticker was placed on the front of each folder.

A carbohydrate counting sheet for children and young people who undertake carbohydrate counting as part of their daily regimen was developed by the EPIC research team in conjunction with the medical illustrator.

Production of text for further information and carbohydrate counting sheets

The content and text for the further information and carbohydrate counting sheets was developed in conjunction with the members of the clinical expert group.

EPIC children’s diabetes information pack folders

Three age-appropriate children’s diabetes information packs (age ranges 6–10 years, 11–15 years and 16–18 years) were prepared for children and young people on insulin injections and pumps.

Purchasing EPIC age-appropriate pack folders

Resources for the 6–10 years and 11–15 years age groups were presented within a transparent folder comprising a number of different compartments with colourful tabs to label each compartment (Figure 18a and b). The folder for the 16–18 years age group was a grey portfolio with a number of different compartments with tabs to label each compartment (Figure 18c).

FIGURE 18.

EPIC packs. (a) 6–10 years; (b) 11–15 years; (c) 16–18 years.

Selection of age-appropriate diabetes resources in conjunction with paediatric diabetes specialist nurses

The scoping exercise of currently available children’s health information, reported in Chapter 2, identified 120 different resources for children and young people with T1D aged from 6 to 18 years. The resources were organised into the three age ranges used in this study and, taking into account quality assurance and children’s and young people’s preferences and information needs, a selection of age-appropriate resources was purposively chosen for each information pack in consultation with a PDSN. These selected resources included age-appropriate paper-based information, DVDs and an additional information sheet listing age-appropriate books and websites. The contents of the three age-appropriate packs are shown in Boxes 32–34. These resources were quality assured against NICE guidelines (see Chapter 2) and approved by members of the clinical expert group.