Development and evaluation by a cluster randomised trial of a psychosocial intervention in children and teenagers experiencing diabetes: the DEPICTED study.

OBJECTIVE: To develop and evaluate a health-care communication training programme to help diabetes health-care professionals (HCPs) counsel their patients more skilfully, particularly in relation to behaviour change.
DESIGN: The HCP training was assessed using a pragmatic, cluster randomised controlled trial. The primary and secondary analyses were intention-to-treat comparisons of outcomes using multilevel modelling to allow for cluster (service) and individual effects, and involved two-level linear models.
SETTING: Twenty-six UK paediatric diabetes services. PARTICIPANTS: The training was delivered to HCPs (doctors, nurses, dietitians and psychologists) working in paediatric diabetes services and the effectiveness of this training was measured in 693 children aged 4-15 years and families after 1 year (95.3% follow-up).
INTERVENTIONS: A blended learning programme was informed by a systematic review of the literature, telephone and questionnaire surveys of professional practice, focus groups with children and parents, experimental consultations and three developmental workshops involving a stakeholder group. The programme focused on agenda-setting, flexible styles of communication (particularly guiding) and a menu of strategies using web-based training and practical workshops. MAIN OUTCOME MEASURES: The primary trial outcome was a change in glycosylated haemoglobin (HbA1c) levels between the start and finish of a 12-month study period. Secondary trial outcomes included change in quality of life, other clinical [including body mass index (BMI)] and psychosocial measures (assessed at participant level as listed above) and cost (assessed at service level). In addition, patient details (HbA1c levels, height, weight, BMI, insulin regimen), health service contacts and patient-borne costs were recorded at each clinic visit, along with details of who patients consulted with, for how long, and whether or not patients consulted on their own at each visit. Patients and carers were also asked to complete an interim questionnaire assessing patient enablement (or feelings towards clinic visit for younger patients aged 7-10 years) at their first clinic visit following the start of the trial. The cost of the intervention included the cost of training intervention teams.
RESULTS: Trained staff showed better skills than control subjects in agenda-setting and consultation strategies, which waned from 4 to 12 months. There was no effect on HbA1c levels (p = 0.5). Patients in intervention clinics experienced a loss of confidence in their ability to manage diabetes, whereas controls showed surprisingly reduced barriers (p = 0.03) and improved adherence (p = 0.05). Patients in intervention clinics reported short-term increased ability (p = 0.04) to cope with diabetes. Parents in the intervention arm experienced greater excitement (p = 0.03) about clinic visits and improved continuity of care (p = 0.01) without the adverse effects seen in their offspring. The mean cost of training was £13,145 per site or £2163 per trainee. There was no significant difference in total NHS costs (including training) between groups (p = 0.1).
CONCLUSIONS: Diabetes HCPs can be trained to improve consultation skills, but these skills need reinforcing. Over 1 year, no benefits were seen in children, unlike parents, who may be better placed to support their offspring. Further modification of this training is required to improve outcomes that may need to be measured over a longer time to see effects. TRIAL REGISTRATION: Current Controlled Trials ISRCTN61568050. FUNDING: This project was funded by the NIHR Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 15, No. 29. See the HTA programme website for further project information.