Management of diabetes mellitus (DM) during adolescence is challenging, affecting social and emotional well-being.1 Glycaemic control and quality of life (QoL) typically decline during this period.2^–7 Based on findings from the Diabetes Control and Complications Trial (DCCT), intensive therapy is recommended for all patients.8 However, tight glycaemic control is difficult to achieve and moving to multiple daily injections (MDI) does not necessarily improve control9 unless it is introduced in parallel with structured education.10^–12

DAFNE (Dose Adjustment For Normal Eating) is a 5-day out-patient course for adults with type 1 DM where participants are taught the skills of carbohydrate counting and insulin dose adjustment. In a randomised controlled trial (RCT), significantly improved glycaemic control, satisfaction with treatment and QoL were found in those who completed the course compared to those receiving standard care, and these improvements were sustained at 12 months after intervention.13

Using DAFNE principals of care we have produced the Kids in Control of Food (KICk OFF) training course for children and adolescents. Paediatric specialist nurses, children, parents and secondary school teachers contributed to the development of an age-appropriate curriculum which employs a range of teaching methods and a modular approach, building on skills acquired during the week. Development and evaluation of the programme has been guided by recommendations of the Medical Research Council (MRC)14 and are described in previous reports.15^–17 In this paper, our aims were to assess (i) the quality of the curriculum through independent evaluation, (ii) acceptability to families (willingness to take part, have time off school or work) and (iii) impact on outcome measures (HBA1c, weight, diary reports of hypoglycaemia, QoL).

METHODS

Procedure

Ethics approval was obtained from the Multicentre Research Ethics Committee and each participating trust. Patients aged 11–16 years with type 1 DM of ⩾1-year duration from three UK centres (Sheffield (n = 70), Manchester (n = 58) and Derby (n = 60)) were invited to take part by letter. Fifty five families agreed to participate (34%) and were given further written and verbal information. Patients were stratified by age and sex and randomly selected so as to form six courses with eight patients per group. Signed consent was obtained 2 weeks before the course started. Children with learning, behavioural or language difficulties were excluded since the course involves group interaction and communication. The study flowchart is shown in fig 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Flow of patients through the pilot study. MDI, multiple daily injections.

Separate courses were held for 11–13 and 14–16 year olds between November 2004 and March 2005 in local secondary schools (Sheffield and Derby) or an education centre (Manchester). Courses were held during school time and took place over five consecutive school-length days. Information letters were sent to head teachers and parents’ employers (requesting time off school or work). Each course was taught by three educators (a KICk-OFF paediatric diabetes specialist nurse (PDSN) and dietician and the PDSN from each centre). All educators attended a 3-day training course to improve teaching skills and gain experience in a secondary school.

Biomedical measures were assessed at baseline, 3 and 6 months after intervention and psychosocial measures at baseline, 2 weeks, 3 and 6 months after intervention. On completion of the course, children returned to routine clinic care.

Evaluation

Evaluation of educational content: quality assurance

An educationalist (EM), who had no other role in the project, was employed to review each course and determine the age-appropriateness of the curriculum and consistency of teaching. This included observation and informal discussions with children and their parents.

Acceptability to children and families

Semi-structured interviews were conducted (by HW) separately with all participants and parents before and after the courses (within 2 weeks of completion). Interviews focused on expectations and worries before the course and perceived advantages and disadvantages after the course.

Biomedical evaluation

Clinical and demographic data were obtained from medical records. Socio-economic status (SES) was determined using the ACORN classification.18 This yields five categories based on postcode (1 = wealthy achievers, 2 = urban prosperity, 3 = comfortably off, 4 = moderate means, 5 = hard-pressed).

HbA1c levels (glycaemic control) were measured in a central laboratory (Sheffield). Episodes of mild (self-treated), moderate (requiring third party help) and severe (coma/convulsions) hypoglycaemia and ketosis were recorded by diary (completed by children and parents) during the study period. Height and weight were measured at each time point to calculate BMI.

Psychosocial evaluation

Children and parents completed parallel versions of the following questionnaires: the Pediatric Quality of Life Inventory Version 4.0 (PedsQL 4.0),19 the PedsQL Diabetes Module (PedsQL DM),20 the Diabetes Treatment Satisfaction Questionnaire (DTSQ),21 the Diabetes Family Responsibility Questionnaire (DFRQ),22 Self-Efficacy for Diabetes (SED)23 and the Diabetes Family Conflict Scale (DFCS).24

Statistical analysis

Analyses were conducted using SPSS v 10.0. Descriptive statistics were calculated for all outcomes at each time point. Baseline differences between completers and non-completers on clinical and demographic data (SES, sex, age group, pre-course regimen, course centre) were explored using a series of independent t tests and one-way between-measures ANOVAs. The impact of the intervention was assessed using one-way repeated-measures ANOVAs for each biomedical and psychosocial outcome across all time points.

RESULTS

Baseline characteristics

Baseline demographic and clinical characteristics are shown in table 1. There were no differences between those who completed questionnaires at 6-month follow-up (n = 41) and those who did not (n = 7) in chronological age, duration of diabetes, HbA1c, SES, sex or pre-course insulin regimen.

Evaluation of educational content

Discussion notes were coded using content analysis,25 and recommendations for improving the curriculum are shown in box 1. Conclusions were that “the observed courses were well planned and created a good balance between individual and group learning”. Sessions were deemed to be age appropriate and participants “expressed high levels of satisfaction with both the activities and the way they were involved in learning”. The evaluation also found high consistency between courses through use of “well written [support materials] and the curriculum stages were easy to follow”.

Patient and parent views

Interviews were transcribed and analysed using thematic analysis.26 Ten transcripts were second-coded (CE) to establish inter-rater reliability. Key themes in relation to four main areas were described: pre-course expectations, pre-course worries, positive experiences of the course and negative experiences of the course (table 2). Overall, the courses appeared to be well received and helpful suggestions were made for improving the course.

Outcomes

Biomedical data

There were no significant changes in mean HbA1c, BMI (standard deviation scores) or episodes of hypoglycaemia over the three time points (table 3).

Psychosocial data

Both patients and parents reported significantly improved QoL (generic and diabetes-specific) and satisfaction with treatment from baseline to 6 months after the course. Children reported improved self-efficacy and both children and parents reported greater child responsibility for a range of management tasks. No significant changes were seen in either child or parent reported family conflict (table 4).

DISCUSSION

The course was carefully evaluated by an independent assessor who identified a number of issues for improvement. These include simplified worksheets for more difficult topics (eg, estimating the carbohydrate content of foods without packets and nutritional information, especially when eating out), provision of individual support where necessary and comparable education for parents to enable them to support their child. Some assessment of knowledge is important in the future.

Our results suggest that families were interested in the programme, and with supportive letters from the hospital, head teachers were agreeable to children attending during school time. All children completed the course (although not all completed the final questionnaires).

Only 31% of eligible families expressed initial interest in participating in the project. Course timing and need for time off work, managing other children and everyday commitments were all reasons for refusal. While this suggests a need for flexibility in timing of courses, it is important to acknowledge the demands on modern families which may prove an obstacle to participation for some.

Analysis of the pre-course interviews provided insight into children’s and parents’ reasons for taking part, and complemented those described previously.18 Reasons for participating included anticipated improved glycaemic control and self-management skills, and improved QoL through dietary freedom and increased independence. Pre-course worries included an increased number of injections and tests, diabetes management during school time, and anticipated difficulties learning to count carbohydrates and make insulin adjustments. After the course, increased self-management skills and QoL through greater dietary freedom, independence and increased child responsibility for diabetes management were reported.

Due to the uncontrolled nature of the study, the results of biomedical and psychosocial outcome measures should be treated with caution. Glycaemic control did not improve following the course. Our failure to observe a fall in HbA1c levels was disappointing (although this was not anticipated given the sample size). HbA1c levels often rise over time in this age group2 4 and therefore preventing deterioration of HbA1c during adolescence may in itself confer some long-term benefits. The adult DAFNE trial and similar interventions have consistently reported falls in HbA1c, but mean baseline HbA1c was 9.4% (only poorly controlled patients were recruited), which is considerably higher than that in the present study (8.58%). We did not measure how much children altered self-management behaviour following the course and it is possible that more ongoing support was needed to put their theoretical knowledge into practice. “Top-up” sessions will be incorporated into future work. The post-course interviews identified some “teething problems” in adjusting to the new insulin regimen and carbohydrate counting. Assimilation of these skills into everyday diabetes self-management might also influence the rate of change of HbA1c and any future evaluation of the course will require a longer follow-up period. Interview data and the external assessor showed that parents did not receive the same level of training as their children. This will be addressed in the revised curriculum, given the need for parents to support their children in this new approach.

It was not appropriate to undertake a detailed health economics analysis in a pilot study, although this would be included in any future RCT. However, we estimate that the cost of providing skills based training (educator salaries and resources) in this study to around 50 children was approximately £600 per child.

The improvement in both general and diabetes-specific QoL appears to be clinically significant (a 5–10-point improvement on 0–100-point QoL scale26). This provides evidence that adolescents are able to adapt to MDI and the demands of self-care during school time without detriment to their QoL. Our results provide the necessary evidence for further evaluation of KICk-OFF by RCT and detailed health economic analysis. However, we would suggest some modifications to ensure that parents and children are able to work together after training.