REmote SUpport for Low-Carbohydrate Treatment of Type 2 Diabetes

This is an individually randomised controlled trial. Participants, recruited via GP practices, will be randomised 1:1 with simple randomisation and no stratification to either the intervention group or control (usual care). They will be enrolled in the study for 1 year, of which the intervention (dietary support via the app) will run for 3 months. Most study procedures will be conducted online or remotely (by telephone) including: screening, eligibility assessment, informed consent, randomisation, engagement with the app-based intervention. Participants will be asked to attend 3 face to face appointments for the research study at their local GP practice for measurements to be taken (weight, blood pressure) and blood tests, at baseline, 3 months and 1 year, and will have 1 additional telephone contact with their GP or practice nurse before starting the intervention. Face to face appointments will be incorporated with routine clinical appointments when possible.

The investigators aim to recruit 100 participants via their GP practices. Based on a conservative response rate of 10% from similar previous studies, the investigators will approach approximately 1000 eligible individuals from approximately 10 GP practices in England. The primary care provider will search their electronic registers for eligible individuals and GPs will screen out those to whom it would be inappropriate to send a letter inviting them to enrol and referring them to a website to give informed consent.

Once consent has been given, participants will click on a link to the electronic data capture system database which will be activated following their consent, to complete the baseline assessment. All participants will complete a 2 part baseline assessment, firstly online and then at their GP practice.

1. Online: Participants will be asked to complete a baseline questionnaire requesting demographic information and their current medication, and the standard questionnaires that will be used in the study (the Problem Areas in Diabetes (PAID) score, the EQ5D (quality of life assessment), and a 24 hour dietary recall questionnaire ("Intake24 dietary assessment tool", where participants are asked to record all food and drink intake in the last 24 hours). This will also further assess whether people are sufficiently able to complete online trial procedures, and only those who complete the questionnaire will be randomised. Participants will be informed in the Participant Information Sheet that they will only be able to enter the study after completing the questionnaire fully in one attempt as the system will not allow them to go back into the questionnaire to answer the remaining questions later. This should take less than 30 minutes to complete.
2. At the GP practice (visit 1) Having completed the online assessment, participants will be asked (via message on their screen on completing the assessment, and followed up via email) to book the in-person portions of their baseline assessment at their GP practice for measures which cannot be conducted online. At this appointment, their height, weight and blood pressure will be measured and recorded, and a blood sample taken for baseline values (HbA1c, lipid profile, liver function tests).

The clinician will randomise the participant at the baseline appointment. This is an individually randomised 1:1 to 2 arms: Intervention (12 week, app-based dietary support), or control (usual care; no active additional intervention). The randomisation will be conducted using the in-built randomisation software. This ensures full allocation concealment as information on future allocations is not accessible to the person randomising. Due to the nature of the intervention it will not be possible to blind members of the study team or participants to allocation.

Participants randomised to the intervention group will receive an email giving 12 weeks of access to the app/website. The intervention involves diet, activity and behaviour change components. It comprises a three-month remote behavioural change programme with mentoring from a registered dietitian or nutritionist (health coach), peer group support, structured education articles and activity tracking technology. These elements are accessed via a smartphone or web-based application. Each participant will also receive a hard copy of an instructional handbook and a recipe book.

Participants randomised to the intervention group will consult their GP to adjust medication as appropriate to avoid hypotension or hypoglycaemia following a guideline.

Participants randomised to the control group will receive no additional intervention, and will continue to receive their usual NHS diabetes care from their general practice.

All Participants will be followed up after 3 months (visit 2) and at 12 months (visit 3). At this appointment they will have the same procedures as at their baseline assessment repeated: their weight and blood pressure will be recorded, and blood tests taken (3 sample tubes - for HbA1c, lipid profile and liver function). Participants will be asked to complete the same online questionnaires (PAID score, EQ5D, and 24 hour dietary recall) as they completed at baseline, and serious adverse events and events of concern (serious hypoglycaemia) will be recorded.

A purposively sampled proportion of participants from the intervention group (aiming to achieve diversity of age, race, and success and engagement with the programme measured by weight/HbA1c outcomes and engagement metrics respectively) will be contacted via telephone for a semi-structured interview to discuss their experience of the intervention, after they complete the 3 month intervention period. Interviews will last approximately 30 minutes. With participants consent this will be audio-recorded.

The duration of the SAE recording period for each participant lasts from their enrolment on to the study, to their completion of the study.

The study results will be reported in accordance with the CONSORT (Consolidated Standards of Reporting Trials) 2010 statements. A full analysis plan will be prepared and finalised before any data analysis.

The primary statistical analysis of efficacy outcomes will be carried out on the basis of intention-to-treat (ITT). This is, after randomisation, participants will be analysed according to their allocated intervention group irrespective of what intervention they actually receive. The investigators will endeavour to obtain full follow-up data on every participant to allow full ITT analysis, but the investigators will inevitably experience the problem of missing data due to withdrawal, loss to follow up, or non-response to questionnaire items.

The investigators will analyse the primary continuous outcome with a linear model adjusting for baseline value of the relevant variable where relevant. For exploratory analyses, the investigators will use analogous models, but interpret the data in an exploratory manner rather than confirmatory. For binary outcomes (e.g. remission), the investigators will test for a difference in proportions using logistic regression.

Based on previous studies of this intervention(5) and studies in similar populations(3), an 8-10mmol/mol change in HbA1c at 12 months in the intervention group in those who engage with the programme (SD 8.6mmol/mol), with a 1-2mmol/mol (SD 6mmol/mol) change in the control group is estimated. With α and β set at 0.05 this would require approximately 48 participants; however to power for a co-primary outcome would require 72 to 88 participants in total (allowing for variability in estimated effect size). In similar trials of behavioural interventions an expected retention to follow up of >80%; allowing for the online nature of this trial, and the retention rates of observational studies of this intervention the investigators will aim to recruit 100 participants to allow for cumulative attrition at the intervention sign-up, engagement, and follow up stages. The results from the trial will be prepared as comparative summary statistics (difference in proportions or means) with 95% confidence intervals. All the tests will be done at a 5% two-sided significance level.

This study is funded by the Wellcome Trust (220551/Z/20/Z) and the NIHR Oxford Biomedical Research Centre.