Diabetes Companion App for Adults With Diabetes and Their Carers

Diabetes self-management support, education, and training are increasingly being delivered through digital technology such as mobile phones. This protocol aims to evaluate the effectiveness of GLOW, a diabetes companion app with a conversational agent.

Adopting a healthier lifestyle and disease self-management skills is critical to reducing the risk of diabetes-related complications. Internationally, and in Singapore, the majority of people with diabetes report not having received structured diabetes education, while others have the knowledge and fail to apply it in their daily lives. Key self-management behaviours recommended by the American Association of Diabetes Educators include physical activity, healthy diet, blood glucose monitoring, and medication adherence, but equally important is healthy coping to maintain a positive attitude toward diabetes management. Life-long patient education and continuous empowerment are paramount to successful self-management in people with diabetes, as recommended by evidence-based diabetes management guidelines. In addition to patients, their informal carers often play an important role in the management of diabetes. As such, diabetes education for informal carers is essential to achieve the best possible diabetes outcomes in the patient. There are hundreds of diabetes apps available for the public to download and use, but few are rigorously evaluated for clinically meaningful outcomes. We designed GLOW, a diabetes companion app with a conversational agent, to support people with diabetes and their carers with self-management and knowledge.

Conversational Agent, Chatbot, Diabetes Literacy, Diabetes Knowledge, Self Management, Digital Therapeutics, Digital Health

Participants in the intervention group will use GLOW the conversational agent/app for 6 months in addition to their usual diabetes care offered by their attending doctor or endocrinologist.

The control group will continue with their usual care, i.e., scheduled consultations with their diabetes team at TTSH and any consultations with healthcare professionals and diabetes education support received either during the consultations or during dedicated times if there are any. They will not be asked to use GLOW during the trial.

GLOW is an app for type 2 diabetes support and education that uses a conversational agent to help users develop self-management skills and learn about diabetes and healthy behaviours. This is achieved through a range of interactive strategies.

Changes in diabetes knowledge among patients and carers using Diabetes Knowledge Questionnaire at baseline, interim (3 months) and end-of-trial (6 months).

The changes in diabetes knowledge will be measured using the scores of a validated questionnaire, Diabetes Knowledge Questionnaire (DKQ).

Changes in diabetes knowledge among patients and carers in the intervention group using GLOW Quiz at baseline, interim (3 months) and end-of-trial (6 months).

The changes in diabetes knowledge among patients and carers in the intervention group will be measured using the scores of a new quiz developed based on GLOW modules, GLOW Quiz.

Changes in waist circumference among patients at baseline, interim (3 months) and end-of-trial (6 months).

Changes in glycemic controls measured by HbA1c among patients at baseline, interim (3 months) and end-of-trial (6 months).

Changes in fasting blood glucose levels among patients at baseline, interim (3 months) and end-of-trial (6 months).

Changes in systolic and diastolic blood pressure among patients at baseline, interim (3 months), end-of-trial (6 months).

Systolic and diastolic blood pressure will be retrieved from patient's clinic records and recorded in mmHg.

Changes in serum high density lipoprotein (HDL) levels among patients at baseline, interim (3 months), end-of-trial (6 months).

High density lipoprotein (HDL) will be retrieved from patient's clinic records and recorded in mmol/L.

Changes in serum low density lipoprotein (LDL) levels among patients at baseline, interim (3 months), end-of-trial (6 months).

Low density lipoprotein (LDL) will be retrieved from patient's clinic records and recorded in mmol/L.

Changes in total cholesterol levels among patients at baseline, interim (3 months), end-of-trial (6 months).

Changes in number of diabetes related hospitalization among patient at baseline, interim (3 months) and end-of-trial (6 months).

Data on hospitalization in term of frequency of hospitalization will be retrieved from patient's clinic records.

Changes in medical cost of diabetes related hospitalization among patient at baseline, interim (3 months) and end-of-trial (6 months).

Data on direct medical cost, direct non-medical cost and indirect cost will be collected in Singaporean Dollar (SGD) terms from patient's clinic records.

Changes in patients empowerment levels score using, Diabetes Empowerment Scale, at baseline, interim (3 months) and end-of-trial (6 months).

Score from a validated questionnaire, Diabetes Empowerment Scale, will be used to measure patient's empowerment levels.

Changes in patients attitude and behaviour score using Diabetes Intention, Attitude, and Behaviour Questionnaire (DIAB-Q), at baseline, interim (3 months) and end-of-trial (6 months).

Score from a validated questionnaire, Diabetes Intention, Attitude, and Behaviour Questionnaire (DIAB-Q), will be used to measure patient's attitude and behaviour towards diabetes.

Changes in patients and carers quality of life will be measured using EQ-5D-5L Questionnaire, at baseline, interim (3 months) and end-of-trial (6 months).

Score from a validated questionnaire, EQ-5D-5L Questionnaire, will be used to evaluate patients and carers quality of life.

Changes in caregiving levels among carers using Diabetes Caregiver Activity and Support Scale at baseline, interim (3 months) and end-of-trial (6 months).

Score from a validated questionnaire, Diabetes Caregiver Activity and Support Scale will be used to evaluate the carer's caregiving levels

Changes in caregiving burden among carers using Brief Assessment Scale for Caregivers at baseline, interim (3 months) and end-of-trial (6 months).

Scores from a validated questionnaire, Brief Assessment Scale for Caregivers, will be used to evaluate carer's perception and sense of burden.

GLOW App login frequency among patients and carers in the intervention group at will be measured the end-of-trial (6 months).

Data on app login frequency will be retrieved from the backend database of the app at the end-of trial.

GLOW App login period among patients and carers in the intervention group will be measured at the end-of-trial (6 months).

Data on app login period in term of "minutes spent in the app" will be retrieved from the backend database of the app at the end-of trial.

GLOW App module completion among patients and carers in the intervention group will be measured at the end-of-trial (6 months).

Data on "number of module started" and "number of module completed" will be retrieved from the backend database of the app at the end-of trial.

GLOW App usability perception among patients and carers in the intervention group will be measured at the end-of-trial (6 months).

Score from a validated questionnaire, MHealth App Usability Questionnaire will be used to evaluate usability of the app. This questionnaire will be built in the app and user will be requested to complete it at the end of the trial period. Data on the questionnaire scores will be retrieved from the backend database of the app at the end-of trial.

Inclusion Criteria: Diagnosed with type 2 diabetes and attending Tan Tock Seng Hospital (TTSH) Diabetes Clinic OR caring for a person with type 2 diabetes 21 years old or above Able to speak and read English Singapore nationality or permanent resident or foreign domestic worker (for carers) Own a personal mobile device which can download study mobile app Access to the internet Exclusion Criteria: Are unable to consent; Are pregnant (for patients) Received formal training in medicine or allied health services

Nanyang Technological University, Center for Population Health Sciences, Lee Kong Chian School of Medicine