Chatbot-delivered Screening and Brief Intervention for Alcohol Reduction in Working-age Adults

Effectiveness and Usability of a Chatbot-delivered Screening and Brief Intervention for Alcohol Reduction in Working-age Adults: A Proof-of-concept Randomised Controlled Trial

Alcohol abuse led to 5.3% of all deaths and 5.1% of all disability-adjusted life years globally in 2016, representing a heavier public health burden than diabetes, tuberculosis or HIV/AIDS (as documented in the World Health Organization (WHO) Global Status Report on Alcohol and Health). The increasing consumption of alcohol for a few decades has led to a higher risk of cirrhosis, cancers, hypertension, and cardiovascular and cerebrovascular diseases. Strengthening of the prevention and treatment of alcohol abuse has been incorporated in the Sustainable Development Goals (SDG3) by the United Nations.

Strong evidence from a meta-analysis demonstrated the efficacy of screening and brief intervention (SBI) in reducing weekly alcohol consumption. Although SBI is known to be effective in reducing alcohol consumption in at-risk drinkers, barriers to implementing SBI have been an issue. A systematic review identified that common barriers to the routine delivery of SBI by doctors and nurses included a lack of alcohol-related knowledge, time, confidence, ability, and incentive to intervene; worrying about offending patients; and SBI being an uncomfortable and frustrating task.

To scale up behavioural change interventions in primary care for expanding the scalability and reachability, artificial intelligence (AI) and AI-chatbots have been increasingly used in recent years. A systematic review showed that chatbots for mental health counselling were effective and safe. Other reviews also reported that chatbots might improve physical activity, diet, and weight management and oncology care. However, having searched PubMed and the Cochrane Library, there was no a randomised controlled trial on the use of an AI-chatbot for alcohol reduction.

Study Overview

• Status: Active, not recruiting
• Conditions: Excessive Alcohol Consumption
• Intervention / Treatment: Behavioral: chatbot-delivered screening and brief intervention

Detailed Description

Aim:

To adapt a self-developed SBI chatbot and conduct a proof-of-concept evaluation on its preliminary effectiveness and usability in reducing alcohol consumption after 4 weeks for at-risk working-age adults by using a randomised, open label, two-arm, parallel-group controlled trial.

Objectives of this project are:

1. To evaluate the SBI chatbot for its short-term effectiveness in reducing alcohol consumption in at-risk working-age adults over a 4-week period (primary outcome).
2. To evaluate the SBI chatbot in reducing alcohol-related harm risk measured by AUDIT scores over 4 week in at-risk working-age adults.
3. To assess the usability of the SBI chatbot by at-risk working-age adults.
4. To explore which factors are associated with, moderate, or mediate the effects of the SBI chatbot.

Hypotheses Hypothesis 1 (Primary outcome): Participants receiving chatbot-delivered SBI (intervention group) will have a higher reduction in weekly alcohol consumption (grams/week) than those in the waitlist control group at 4-week follow-up.

Hypothesis 2 (Secondary outcome): The intervention group will have a lower AUDIT score than the control group at 4-week follow-up.

• Study Type: Interventional
• Enrollment (Estimated): 300
• Phase: Not Applicable

Contacts and Locations

Study Locations

• Hong Kong
  • Hong Kong Island
    • Hong Kong, Hong Kong Island, Hong Kong, 000
      • School of Nursing

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult
• Accepts Healthy Volunteers: No

Description

Inclusion Criteria:

1. aged 18-59 years
2. being employed in the past 12 months
3. a modified Alcohol Use Disorders Identification Test (AUDIT-C) score ≥3
4. having a smartphone with internet access
5. able to communicate in Chinese

Exclusion Criteria:

1. mentally and/or
2. physically unable to complete the intervention and/or the two follow-ups
3. undergoing any alcohol-related intervention

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Screening
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: Single

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: Intervention group; Participants in the intervention group will engage with a chatbot-delivered screening and brief intervention (SBI) over a 30-minute period on the chatbot webpage. The SBI chatbot involves screening individuals for alcohol consumption by AUDIT before recruitment and then providing a brief intervention comprising personalised feedback. Its delivery is fully automated via a smartphone app.	Behavioral: chatbot-delivered screening and brief intervention; Screening and personalised advice based on AUDIT-C scores, personal reasons for drinking, and specific barriers: For a score of ≥ 3, the participant will be reminded of the level of risks and dangers of alcohol consumption and the benefits of consumption reduction/cessation and will be assessed for the personal reasons for drinking and specific barriers.; Warning of alcohol and alcohol-related harms and reminding of benefits of reduced consumption: The Department of Health's health warning booklet, which has been designed by two of our Co-Is, will be used as the educational content.; Realistic and personal goals setting: A specific and measurable goal will be set for the upcoming week. For example, "I will reduce to 1 bottle of beer in the upcoming week.; Motivational enhancement and practical coping strategies: Tangible advice will be given for coping with situational barriers according to their personal reasons for drinking.
Experimental: Waitlist control group; All participants in the waitlist control will receive the SBI chatbot 4 weeks later. The intervention has the same content for both intervention and waitlist control groups.	Behavioral: chatbot-delivered screening and brief intervention; Screening and personalised advice based on AUDIT-C scores, personal reasons for drinking, and specific barriers: For a score of ≥ 3, the participant will be reminded of the level of risks and dangers of alcohol consumption and the benefits of consumption reduction/cessation and will be assessed for the personal reasons for drinking and specific barriers.; Warning of alcohol and alcohol-related harms and reminding of benefits of reduced consumption: The Department of Health's health warning booklet, which has been designed by two of our Co-Is, will be used as the educational content.; Realistic and personal goals setting: A specific and measurable goal will be set for the upcoming week. For example, "I will reduce to 1 bottle of beer in the upcoming week.; Motivational enhancement and practical coping strategies: Tangible advice will be given for coping with situational barriers according to their personal reasons for drinking.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Alcohol consumption in grams per week	The primary outcome is the amount of alcohol consumed per week (grams/week) at the 4-week follow-up, which is recommended as the gold standard for determining effectiveness in many alcohol trials. The data collected for computing the alcohol consumption comprise type of alcohol (e.g. wine, beer, liquor, etc.) and units (e.g. 1 glass of 250ml, 1 can of 330 ml, etc.) consumed in the past week, which will be computed according to a formula with standard alcohol (unit per week) x 10 gram.	4-week follow-up

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
AUDIT scores	AUDIT has 10 items and the total scores range from 0 to 40. Scores of 1 to 7 suggest low-risk alcohol consumption, 8 to 15 suggest hazardous or harmful alcohol.	4-week follow-up
The 15-item Bot Usability Scale	It was designed to evaluate chatbot usability in 5 domains: perceived accessibility to the chatbot function, perceived quality of chatbot functions, perceived quality of conversation and information provided, perceived privacy and security, and time response. Participants will rate each item on a 5-point Likert scale (1=strongly disagree to 5=strongly agree).	4-week follow-up
Economic evaluation	An EQ-5D-5L utility score will be estimated using the Hong Kong EQ-5D-5L value set. The EQ-5D-5L is a 25-item self-reported health questionnaire plus a visual analogue scale to describe how good or bad your health is today.	4-week follow-up

Collaborators and Investigators

• Sponsor: The University of Hong Kong

Study record dates

Study Major Dates

• Study Start (Actual): December 1, 2023
• Primary Completion (Estimated): June 30, 2025
• Study Completion (Estimated): June 30, 2025

Study Registration Dates

• First Submitted: September 16, 2024
• First Submitted That Met QC Criteria: September 16, 2024
• First Posted (Estimated): September 19, 2024

Study Record Updates

• Last Update Posted (Estimated): September 19, 2024
• Last Update Submitted That Met QC Criteria: September 16, 2024
• Last Verified: September 1, 2024

More Information

Terms related to this study

• Keywords: randomised controlled trial; adults; Chatbot; screening and brief intervention; excessive alcohol consumption
• Additional Relevant MeSH Terms: Drinking Behavior; Alcohol Drinking
• Other Study ID Numbers: Alcohol chatbot for adults

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No