How and why was a digital type 2 diabetes self-management intervention changed during national roll-out? A mixed methods study of fidelity

Background:
‘HeLP-Diabetes’ is a theory-based digital self-management intervention for people with type 2 diabetes mellitus (T2DM), which encouraged behaviour change using behaviour change techniques (BCTs; the ‘active ingredients’ of behaviour change interventions) and promoted self-management. HeLP-Diabetes was effective in reducing HbA1c in a randomised controlled trial (RCT). NHS England have since commissioned a national roll-out of HeLP-Diabetes into routine care (now called ‘Healthy Living’). Healthy Living presents a unique opportunity to examine the fidelity of the national roll-out of an intervention originally tested in an RCT.

Objectives:
Objectives of this research were to: (1) describe the Healthy Living BCT and self-management content, and features of intervention delivery; (2) compare the fidelity of Healthy Living with the original HeLP-Diabetes intervention; and (3) explain the reasons for any fidelity drift during national roll-out through qualitative interviews.

Methods:
A content analysis of Healthy Living was conducted using three coding frameworks (Objective 1): BCT Taxonomy v1; a new coding framework for assessing self-management tasks; and Template for Intervention Description and Replication (TIDieR). The extent to which BCTs and self-management tasks were included in Healthy Living was compared with published descriptions of HeLP-Diabetes (Objective 2). Semi-structured interviews were conducted with nine stakeholders involved in the development of HeLP-Diabetes or Healthy Living to understand the reasons for any changes during national roll-out (Objective 3). Qualitative data were thematically analysed using a modified framework approach.

Results:
The content analysis identified 43 BCTs in Healthy Living. Healthy Living included all but one of the self-regulatory BCTs (‘commitment’) in the original HeLP-Diabetes intervention. Healthy Living was found to address all areas of self-management (medical, emotional and role), in line with the original HeLP-Diabetes intervention. However, two important changes were identified. First, facilitated access from a healthcare professional was not implemented; interviews revealed this was because general practices had fewer resources in comparison with the RCT. Second, Healthy Living included an additional structured online learning curriculum that was developed by the HeLP-Diabetes team, but was not included in the original RCT; interviews revealed this was because of changes in NHS policy that encouraged referral to structured education. Interviewees described how the service provider had to reformat the content of the original HeLP-Diabetes website to make it more usable and accessible to meet multiple digital standards required for implementation in the NHS.

Conclusions:
The national roll-out of Healthy Living had good fidelity to the BCT and self-management content of HeLP-Diabetes. Important changes were attributable to the challenges of scaling up a digital intervention from an RCT into a nationally implemented intervention, mainly due to fewer resources available in practice and the length of time since the RCT. This study highlights the importance of considering implementation throughout all phases of intervention development.