Trial Review

Registering a new trial?

To achieve prospective registration, we recommend submitting your trial for registration at the same time as ethics submission.

The safety and scientific validity of this study is the responsibility of the study sponsor and investigators. Listing a study does not mean it has been endorsed by the ANZCTR. Before participating in a study, talk to your health care provider and refer to this information for consumers
Trial registered on ANZCTR


Registration number
ACTRN12621000174886
Ethics application status
Approved
Date submitted
25/11/2020
Date registered
18/02/2021
Date last updated
20/02/2023
Date data sharing statement initially provided
18/02/2021
Type of registration
Retrospectively registered

Titles & IDs
Public title
Coordinating Healthcare with Artificial Intelligence-supported Technology in Atrial Fibrillation: CHAT-AF
Scientific title
Effect of Coordinating Healthcare with Artificial Intelligence-supported Technology on Atrial Fibrillation related Quality of Life – a randomised controlled trial [CHAT-AF]
Secondary ID [1] 302814 0
None
Universal Trial Number (UTN)
U1111-1261-3254
Trial acronym
CHAT-AF
Linked study record

Health condition
Health condition(s) or problem(s) studied:
Atrial Fibrillation 319777 0
Condition category
Condition code
Cardiovascular 317709 317709 0 0
Other cardiovascular diseases

Intervention/exposure
Study type
Interventional
Description of intervention(s) / exposure
The aim of the intervention is to support atrial fibrillation (AF) self-management and care coordination with artificial intelligence (AI)-supported technology. The intervention consists of a technology platform that enables the programming of a complex algorithm of automated voice calls (interactive voice response (IVR) technology) and text-message (SMS) and/or email, as well as includes an educational website.

OUTREACHES

The intervention will consist of a total of seven patient outreaches, delivered over the course of 6 months. The outreaches will occur at the following timepoints: 24-48 hours, 14 days, 1-, 2-, 3-, 4-, 5- months post-discharge from the hospital. All these outreaches will be automated and follow a complex algorithm.

The format of each outreach consists of:
1. Automated phone call (IVR): The first point of contact in each outreach is via an automated and pre-recorded (real) voice phone call with interactive voice response technology (IVR) capabilities. Using IVR, patients can interact with a structured series of recorded messages and respond to queries using their voice (e.g. by saying “Yes”, “No” over the phone). Based on their responses to these queries, patients will receive recorded messages tailored to their individual needs.

2. SMS or email: If the patient cannot be reached via phone after three attempts, an SMS/email will be sent containing a link to a secure webpage. This webpage will have all the content addressed in the IVR call, in a written format, with options for patient’s to self-report using multiple choice answers.

Each outreach will have specific goals and will vary slightly in the information delivered and querying items. Through these outreaches we are able to collect self-reported patient information, for example, does the patient have a regular GP/medical centre, have they booked/attended their GP appointment, do they have transportation barriers to accessing appointments, along with assessment of overall health status, AF symptoms and their impact on daily life, medication confidence and adherence. During these outreaches, we will also provide information on stroke, sleep apnoea, weight management, AF medical procedures and how to reduce alcohol intake. Patients will also receive health tips that can be incorporated into daily activities, these will comprise of broad and specific AF themes, which include, tips on maintaining a healthy lifestyle, ways to improve medication adherence, avoiding caffeine and energy drinks, reducing salt intake, healthy eating, exercise and managing AF.

MANAGING ALERTS

Based on patient responses to the questions asked in the outreaches, certain answers will trigger an alert, these include; poor overall health status (options ranging from; ‘excellent’ to ‘poor’, where ‘poor’ triggers an alert), impact of AF symptoms on daily life (options ranging from; ‘not at all’ to ‘extremely’, where ‘extremely’ triggers an alert), medication confidence is 1 or 2 out of a scale of 7 (where 1 is ‘not confident at all’ and 7 is ‘very confident’), non-adherence to medication, and not having booked or attended a GP appointment within 1 month of hospital discharge. These alerts will be sent directly to the Westmead Research team and escalated to clinicians or other members of the research team, as appropriate. These alerts will be addressed by calling the patient within 24-72 hours and triaging for priority.

EDUCATION WEBSITE

In addition to the outreaches, patients will have access to an education website, which will be tailored to each individual (based on whether or not they are smokers, drink alcohol, have hypertension, are on warfarin medication). The information will be organized into modules, these include; general AF information, AF procedures, medications, alcohol, blood pressure, smoking, weight management and general resources. Content will be in the form of links to online resources, videos and webpages (e.g. Heart Foundation, NPS). Majority of the content on the education website is readily available online on websites of reputable sources, such as The Heart Foundation Australia, NPS and HealthDirect. However, the website also contains a video series about atrial fibrillation that was developed by cardiologists from Westmead Hospital for their patients and not for the specific purpose of this research study. We will send out weekly emails/SMS (depending on their communication preference) containing a link to their personalised website to encourage patients to visit. We will not specify the frequency or duration of site usage, but will be able to record site access during the study for each participant.

PROGRAM ENGAGEMENT STATISTICS

Program engagement statistics will be collected using web analytics and system engagement metrics (e.g. number of successful phone calls, number of completed phone calls, comprehensiveness of data collected during IVR calls, number of visits on online education webpage and completion of education modules). We will also be able to determine the completion status of each outreach [querying items (e.g. overall health status, AF symptoms, impact on daily life, GP access, medication confidence and adherence) and knowledge assessment (stroke, sleep apnoea, weight management, AF medical procedures and how to reduce alcohol intake]. This will assist in determining the aspects of the program that were most engaging and those that participants found less appealing.
Intervention code [1] 319092 0
Lifestyle
Intervention code [2] 319093 0
Behaviour
Intervention code [3] 319104 0
Treatment: Other
Comparator / control treatment
Control group will receive usual care from their nominated health professionals. In general, usual care of AF patients consists of post-discharge instructions from the Cardiologist regarding medications and lifestyle modifications, recommended General Practitioner follow-up to be organised by the patient, and additional Cardiologist appointments as needed.
Control group
Active

Outcomes
Primary outcome [1] 325751 0
The primary outcome is the difference in AF-related quality of life at baseline and 6-months in the intervention group compared to the control group and will be adjusted for baseline co-variance. The primary outcome will be measured by the Atrial Fibrillation Effect on QualiTy-of-life (AFEQT) questionnaire.
Timepoint [1] 325751 0
Baseline and 6 months post-randomisation
Secondary outcome [1] 388991 0
AF knowledge (measured by the Atrial Fibrillation Knowledge Scale questionnaire) in the intervention group compared to the control group at 6-months post-randomisation.
Timepoint [1] 388991 0
6 months from randomisation
Secondary outcome [2] 388992 0
Patient-reported experience measures (assessed via the Patient Assessment of Chronic Illness Care (PACIC) questionnaire) in the intervention group compared to the control group at 6-months post-randomisation.
Timepoint [2] 388992 0
6 months from randomisation
Secondary outcome [3] 388993 0
Health outcomes (e.g. stroke, myocardial infarction, mortality rates; assessed using medical records and a study-specific questionnaire for information not available on the medical records) in the intervention group compared to the control group at 6-months post-randomisation.
Timepoint [3] 388993 0
6 months from randomisation
Secondary outcome [4] 388994 0
Feasibility with respect to potential for wider scale implementation including acceptability, perceived utility (demand), implementation and integration into IT systems (barriers and enablers) through the conduct of participant surveys and qualitative data capture (semi-structured interviews) from patients and healthcare providers.
Timepoint [4] 388994 0
6 months from randomisation
Secondary outcome [5] 388995 0
Feasibility of Interactive Voice Response to capture patient-reported risk assessments (e.g. overall health status, AF symptom score information) and knowledge assessments (e.g. stroke), via system engagement metrics and semi-structured interviews.
Timepoint [5] 388995 0
6 months post-randomisation
Secondary outcome [6] 389115 0
Lifestyle behavioural outcomes (e.g. number of cigarettes daily, alcohol drinks per week, exercise time, servings of fruit and vegetables; assessed using a study-specific survey) in the intervention group compared to the control group at 6-months post-randomisation.
Timepoint [6] 389115 0
6 months post-randomisation
Secondary outcome [7] 389116 0
Patient-reported outcome measures—patient activation (measured by the Patient Activation Measure (PAM)-13 questionnaire) in the intervention group compared to the control group at 6-months post-randomisation.
Timepoint [7] 389116 0
6 months post-randomisation
Secondary outcome [8] 389117 0
Mean difference in AF-related quality of life at baseline and 3 months in the intervention group compared to the control group, measured by the Atrial Fibrillation Effect on QualiTy-of-life (AFEQT) questionnaire.
Timepoint [8] 389117 0
Baseline and 3 months post-randomisation
Secondary outcome [9] 389238 0
Self-reported medication adherence (assessed by the question "In the last 7 days, on how many days did you miss a dose of any of your prescribed medications?") in the intervention group compared to the control group at 6-months post-randomisation.
Timepoint [9] 389238 0
6 months post-randomisation
Secondary outcome [10] 390277 0
Healthcare service utilisation (e.g. General Practice visits or contacts, Cardiology consultations, presentation to the ED and/or unplanned hospitalisation; assessed using hospital medical records and self-reported study-specific survey) in the intervention group compared to the control group at 6-months post-randomisation.
Timepoint [10] 390277 0
6 months post-randomisation

Eligibility
Key inclusion criteria
Patients will be eligible to participate if they (1) are >18 years old, (2) have a documented diagnosis of AF, (3) have a mobile phone that is able to receive calls, (4) are able to receive SMS or emails and open weblinks embedded in them, and (5) are competent with English, as ascertained by the study researcher.
Minimum age
18 Years
Maximum age
No limit
Sex
Both males and females
Can healthy volunteers participate?
No
Key exclusion criteria
Participants will be excluded from the study if they (1) are participating in a concurrent clinical trial focusing on atrial fibrillation (2) have a concomitant illness, physical impairment or mental condition which in the opinion of the study team/primary care physician could interfere with the conduct of the study including outcome assessment (e.g. hearing impairment not corrected by a digital hearing aid), (3) are pregnant, (4) have a medical illness with anticipated life expectancy of < 3 years, and (5) are unable or unwilling to provide written consent.

Study design
Purpose of the study
Treatment
Allocation to intervention
Randomised controlled trial
Procedure for enrolling a subject and allocating the treatment (allocation concealment procedures)
• Allocation concealment will be ensured utilizing systems in our data management system (RedCap; Research Electronic Data Capture). Thus, we will create separate Data Access Groups to ensure that blinded researchers (e.g. outcome assessors and data analysts) will not be able to see randomization lists or access the randomisation form in RedCap. A non-blinded researcher will manage participant randomization within RedCap as they are recruited.
Methods used to generate the sequence in which subjects will be randomised (sequence generation)
Randomisation will be managed through REDCap (data management system). This software will automatically allocate participants to the intervention or control group, according to the randomisation sequence generated in R (using the randomiseR package). Randomisation will 4:1 (intervention: control) and will be stratified based on gender.
Masking / blinding
Blinded (masking used)
Who is / are masked / blinded?


The people assessing the outcomes
The people analysing the results/data
Intervention assignment
Parallel
Other design features
N/A
Phase
Not Applicable
Type of endpoint/s
Efficacy
Statistical methods / analysis
Our primary analysis will be follow the principles of intention-to-treat to compare participants in the intervention group to that in the control group. We will also report results from analysis per-protocol. T-tests using analysis of covariance will be used to compare mean changes in the primary outcome. For secondary analyses, regression models will be used to compare the treatment groups using the appropriate model for the type of outcome variables, adjusting for baseline variables if relevant. Log binomial models will be used for dichotomous variables and regression models for continuous variables.

A sample size of 350 in a 4:1 ratio (280 in the intervention group and 70 in the control group) is determined to have 80% power to detect a difference of 7 in the total score of the AFEQT questionnaire (alfa set at 0.05; Standard deviation 192). Accounting for a dropout rate of 10%, the total sample size to be recruited is 385.

Recruitment
Recruitment status
Stopped early
Data analysis
Data collected is being analysed
Reason for early stopping/withdrawal
Participant recruitment difficulties
Date of first participant enrolment
Anticipated
Actual
19/01/2021
Date of last participant enrolment
Anticipated
1/02/2023
Actual
13/07/2021
Date of last data collection
Anticipated
1/03/2023
Actual
3/06/2022
Sample size
Target
385
Accrual to date
Final
103
Recruitment in Australia
Recruitment state(s)
NSW
Recruitment hospital [1] 18048 0
Westmead Hospital - Westmead
Recruitment hospital [2] 21776 0
Blacktown Hospital - Blacktown
Recruitment postcode(s) [1] 32026 0
2145 - Westmead
Recruitment postcode(s) [2] 36833 0
2148 - Blacktown

Funding & Sponsors
Funding source category [1] 307179 0
Other Collaborative groups
Name [1] 307179 0
Digital Health Cooperative Research Centre (DHCRC)
Country [1] 307179 0
Australia
Primary sponsor type
Government body
Name
Western Sydney Local Health District
Address
Research and Education Network
Darcy Rd, Westmead Hopsital
Westmead NSW 2145
Country
Australia
Secondary sponsor category [1] 307846 0
University
Name [1] 307846 0
The University of Sydney
Address [1] 307846 0
The University of Sydney
Camperdown NSW 2006
Country [1] 307846 0
Australia
Other collaborator category [1] 281542 0
Commercial sector/Industry
Name [1] 281542 0
Address [1] 281542 0
Country [1] 281542 0
Other collaborator category [2] 281543 0
Other Collaborative groups
Name [2] 281543 0
Digital Health Cooperative Research Centre (DHCRC)
Address [2] 281543 0
Sydney Knowledge Hub, Level 2, Merewether Building H04, The University of Sydney NSW 2006
Country [2] 281543 0
Australia

Ethics approval
Ethics application status
Approved
Ethics committee name [1] 307292 0
Western Sydney Local Health District Human Research Ethics Committee
Ethics committee address [1] 307292 0
Research and Education Network
Darcy Road, Westmead Hospital
Westmead 2145 NSW
Ethics committee country [1] 307292 0
Australia
Date submitted for ethics approval [1] 307292 0
23/09/2020
Approval date [1] 307292 0
04/11/2020
Ethics approval number [1] 307292 0
6661 - 2020/ETH02546

Summary
Brief summary
The aim of this study is to determine the feasibility and efficacy of an AI-supported intervention that comprises automated phone calls (interactive voice response), and emails or text-message (SMS) delivered according to an algorithm, as well as an educational website. Participants interact with the automated phone calls via voice and the system interprets participant responses using natural language processing (AI component). The intervention incorporates periodic assessments of overall health status, AF symptoms and impact on daily life, General Practice (GP) access, and medication confidence and adherence, along with providing education content related to AF.
We hypothesize that a AI supported technology platform for patients with AF can help identify higher-risk patients needing closer follow-up from the healthcare team, provide behaviour support for AF self-management, and identify and address barriers to AF management (attendance at regular GP and specialist appointments), leading to improved outcomes and care experience.
We aim to evaluate the effect of the intervention on quality of life, behavioural and self-management outcomes (e.g. medication adherence), patient knowledge, health outcomes (stroke, myocardial infarction, mortality), healthcare service utilisation, feasibility outcomes (acceptability, usability, program engagement) and patient care experience.
Trial website
Trial related presentations / publications
Public notes

Contacts
Principal investigator
Name 106714 0
Prof Clara Chow
Address 106714 0
Westmead Applied Research Centre, Faculty of Medicine and Health, University of Sydney, Level 6, Block K, Entrance 10, Westmead Hospital, Hawkesbury Road, Westmead, NSW, 2145
Country 106714 0
Australia
Phone 106714 0
+61 414488325
Fax 106714 0
Email 106714 0
[email protected]
Contact person for public queries
Name 106715 0
Ritu Trivedi
Address 106715 0
Westmead Applied Research Centre, Faculty of Medicine and Health, University of Sydney, Level 6, Block K, Entrance 10, Westmead Hospital, Hawkesbury Road, Westmead, NSW, 2145
Country 106715 0
Australia
Phone 106715 0
+61 481 395 165
Fax 106715 0
Email 106715 0
[email protected]
Contact person for scientific queries
Name 106716 0
Ritu Trivedi
Address 106716 0
Westmead Applied Research Centre, Faculty of Medicine and Health, University of Sydney, Level 6, Block K, Entrance 10, Westmead Hospital, Hawkesbury Road, Westmead, NSW, 2145
Country 106716 0
Australia
Phone 106716 0
+61 481 395 165
Fax 106716 0
Email 106716 0
[email protected]

Data sharing statement
Will individual participant data (IPD) for this trial be available (including data dictionaries)?
No
No/undecided IPD sharing reason/comment


What supporting documents are/will be available?

No Supporting Document Provided



Results publications and other study-related documents

Documents added manually
No documents have been uploaded by study researchers.

Documents added automatically
No additional documents have been identified.