Web-based self-management support for people with type 2 diabetes (HeLPDiabetes): Randomised controlled trial in English primary care
Murray E., Sweeting M., Dack C., Pal K., Modrow K., Hudda M., Li J., Ross J., Alkhaldi G., Barnard M., Farmer A., Michie S., Yardley L., May C., Parrott S., Stevenson F., Knox M., Patterson D.
© Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2017. Objective To determine the effectiveness of a web-based self-management programme for people with type 2 diabetes in improving glycaemic control and reducing diabetes-related distress. Methods and design Individually randomised two-arm controlled trial. Setting 21 general practices in England. Participants Adults aged 18 or over with a diagnosis of type 2 diabetes registered with participating general practices. Intervention and comparator Usual care plus either Healthy Living for People with Diabetes (HeLP-Diabetes), an interactive, theoretically informed, web-based selfmanagement programme or a simple, text-based website containing basic information only. Outcomes and data collection Joint primary outcomes were glycated haemoglobin (HbA1c) and diabetesrelated distress, measured by the Problem Areas in Diabetes (PAID) scale, collected at 3 and 12 months after randomisation, with 12 months the primary outcome point. Research nurses, blind to allocation collected clinical data; participants completed self-report questionnaires online. Analysis The analysis compared groups as randomised (intention to treat) using a linear mixed effects model, adjusted for baseline data with multiple imputation of missing values. Results Of the 374 participants randomised between September 2013 and December 2014, 185 were allocated to the intervention and 189 to the control. Final (12 month) follow-up data for HbA1c were available for 318 (85%) and for PAID 337 (90%) of participants. Of these, 291 (78%) and 321 (86%) responses were recorded within the predefined window of 10-14 months. Participants in the intervention group had lower HbA1c than those in the control (mean difference -0.24%; 95% CI -0.44 to -0.049; p=0.014). There was no significant overall difference between groups in the mean PAID score (p=0.21), but prespecified subgroup analysis of participants who had been more recently diagnosed with diabetes showed a beneficial impact of the intervention in this group (p = 0.004). There were no reported harms. Conclusions Access to HeLP-Diabetes improved glycaemic control over 12 months.