A Bot-based Self-help Program (WELL) (WELL)

A Randomized Controlled Trial of a Bot-based Self-help Program (WELL)

In the current project, the effectiveness of a new bot-based universal prevention program aimed at improving well-being will be developed and tested. This prevention program includes a variety of modules that target common areas related to well-being including interpersonal relationships, emotional regulation, beliefs and behaviors, problem-solving, sleep hygiene, goal setting, personal growth, and healthy habits.

Study Overview

• Status: Completed
• Conditions: Well-Being, Psychological
• Intervention / Treatment: Behavioral: Conversational Agent Zenny; Behavioral: Web-based Wellness Resources

Detailed Description

The purpose of the data collection is to evaluate a new intervention as part of a research study.

In the current project, a new bot-based universal prevention program aimed at improving well-being will be developed and tested. This prevention program includes a variety of modules that target common areas related to well-being including interpersonal relationships, emotional regulation, beliefs and behaviors, problem-solving, sleep hygiene, goal setting, personal growth, and healthy habits. The intervention modules are based on existing empirically supported intervention techniques from traditional cognitive-behavioral therapies and other third generation therapies such as mindfulness training and dialectical behavioral therapy. The focus of the program is to introduce psychological concepts that have been shown to be effective for managing emotions and problems to a broader public audience. Another focus is on helping individuals connect with resources that are available related to subjective well-being. Often individuals who may be interested in self-help interventions do not know where to seek help and interventions that they may find (e.g., apps), may not be scientifically valid. By providing recommendations on resources that are specific to an individual's needs, this may reduce one barrier to care.

The focus of this project is not only on testing the effectiveness of the program on well-being compared to an online control condition in a randomized controlled trial, but on improving REACH to individuals who may benefit from such an intervention. Thus, the focus is on brief, low burden assessments, rather than a full battery typical in psychological studies, to improve acceptability, recruitment rates, and retention rates. The implementation science model, RE-AIM, will be used as a framework for this study (http://www.re-aim.org/).

H1: The intervention group will evidence significant improvements in well-being at the post-assessment and follow-up assessments (1-month, 6- month, 12-month) compared to the control group.

H2: The effectiveness of the intervention will be similar across gender and age after controlling for participation/engagement and user satisfaction.

H3: The intervention group will have significantly higher psychological growth compared to the control group at post-assessment and follow-ups assessments (1-month, 6- month, 12-month).

H4: The intervention group will have higher response rates, higher retention rates, and more self-reported active participation/engagement than the control group.

H5: Higher user satisfaction and more active participation/engagement will be associated with greater improvements in well-being.

H6: Higher user satisfaction will lead to subsequent increases in participation/engagement over time.

• Study Type: Interventional
• Enrollment (Actual): 1595
• Phase: Not Applicable

Contacts and Locations

Study Locations

• United States
  • California
    • Menlo Park, California, United States, 94025
      • Meta Platforms Inc.

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria:

• 18 years or older
• proficient in spoken and written English
• residents of the United States
• provided consent to be eligible to participate in the study

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Prevention
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: Double

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: Conversational Agent Zenny; Participants received access to a self-help program delivered via an automated conversational agent on WhatsApp.	Behavioral: Conversational Agent Zenny; The intervention group involves a newly developed conversational agent self-care program. This program consists of 40 separate modules based on empirically supported techniques. Modules include the following core sections: cognitive behavioral skills, interpersonal relationships, positive psychological growth, relaxation, goal setting, and emotional regulation skills. The modules focus on cognitive-behavioral skills for reducing emotional distress, improving behavioral activation and healthy behaviors, and problem-solving skills. The intervention was delivered via a decision-tree based conversational agent in WhatsApp messenger.
Active Comparator: Web-based Wellness Resources; Participants received a list of three freely accessible wellness resources available on the web.	Behavioral: Web-based Wellness Resources; The control group involved referral to a menu of freely-available evidence-based wellness resources linked from the study website, "Your Healthiest Self: Wellness Toolkits", "Doing What Matters in Times of Stress", and "How Right Now". These website choices involved self-selection of relevant modules for improving mood, reducing stress and anxiety, addressing relationship problems, and other wellness topics. These programs required only access to the websites and could be completed in a self-help format by participants at their own pace.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Subjective well-being	The WHO-5 Well-Being Scale (WHO-5) is a 5-item short self-reported measure of current subjective wellbeing. Participants are asked to rate the frequency of well-being past month (e.g., "My daily life has been filled with things that interest me") based on a Likert scale from 0 to 5 ("At no time" to "All of the time"). Scores range from 0 to 25 with higher scores indicating higher subjective well-being.	At pre-assessment and 1, 6 and 12 months after pre-assessment
Psychosocial flourishing	The Flourishing Scale (FS) is an 8-item scale of positive human functioning. Items assess features of human flourishing such as positive relationships, feelings of competence, meaning and purpose in life, and engagement with daily activities. Statements are rated on a 7-point scale, ranging from 1 (strongly disagree) to 7 (strongly agree). Scores can range from 8 to 56, with higher scores reflecting higher well-being.	At pre-assessment and 1, 6 and 12 months after pre-assessment
Psychological well-being	The Mental Health Continuum Short Form (MHC-SF) measures positive psychological health and comprises 14 items, representing various feelings of well-being. Three items were chosen (happy, interested in life, and satisfied) to represent emotional well-being, six items (one item from each of the 6 dimensions) were chosen to represent psychological well-being, and five items (one item from each of the 5 dimensions) were chosen to represent social well-being. Respondents rate the frequency of every feeling in the past month on a 6-point Likert scale, ranging from 1 (never) to 6 (every day).	At pre-assessment and 1, 6 and 12 months after pre-assessment

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
User satisfaction	This is assessed with four items during the post-assessment for both conditions. Satisfaction with the program is formed by individual items on satisfaction, usefulness, relevance, and helpfulness. Participants rate these dimensions on 4-point Likert scales between 0 (strong disagreement) to 4 (strong agreement). Only the item on usefulness is rated on a 5-point Likert scale. A sum score is calculated by summarizing all single item scores, resulting in a range from 0 to 17. Higher scores indicate greater satisfaction with the content.	1 month after pre-assessment
Participation/Engagement	This is assessed by the reported frequency using the intervention or control condition resources at the post-assessment. For the first item, participants rate on a 6-point Likert scale how often they participated in the last month, ranged from "not at all" (1) to "daily" (6). On a second item, participants are asked to rate their perceived level of activity on a 4-point Likert scale, ranging from "not at all active" (1) to "very active" (4). The scores for both items are summarized into a sum score, with higher scores indicating greater participation/engagement.	1 month after pre-assessment

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Module rating of user satisfaction	Within the intervention group, user satisfaction is assessed following each module. Users are able to rate satisfaction with each completed module from 1 "not at all" (1) to "extremely" (5). An average satisfaction rating across modules over a thirty-day period is computed.	1 month after pre-assessment
Days chatted bot	Engagement with the conversational agent is defined as the count of days that a user engaged in conversations within the last 30 days from 0 days to 30 days.	1 month after pre-assessment
Total messages	The number of messages sent by the user to the conversational agent is quantified as a further objective measure. This refers to any situation where the user is starting an interaction with the conversational agent or is responding to a message from the conversational agent within the last 30 days from 0 days to 30 days.	1 month after pre-assessment
Modules started	This is the total number of modules that is initiated by the participant within the first 30 days, whether they were completed or not (i.e., sum of finished and unfinished modules).	1 month after pre-assessment
Modules completed	This is the number of modules that is completed within the first 30 days (i.e., finished modules).	1 month after pre-assessment
Modules started not completed	This is the number of modules that is started but not completed within the first 30 days (i.e., unfinished modules).	1 month after pre-assessment

Collaborators and Investigators

• Sponsor: University of Klagenfurt
• Collaborators: Meta Platforms, Inc.
• Investigators: Principal Investigator: Heather M Heather, Ph.D., University of Klagenfurt; Principal Investigator: Y-Lan Boureau, Ph.D., Meta Platforms

Publications and helpful links

Helpful Links

• Open Science Framework Preregistration (June 3, 2022)
• Study Website

Study record dates

Study Major Dates

• Study Start (Actual): July 4, 2022
• Primary Completion (Actual): March 20, 2024
• Study Completion (Actual): March 20, 2024

Study Registration Dates

• First Submitted: December 13, 2023
• First Submitted That Met QC Criteria: January 5, 2024
• First Posted (Actual): January 17, 2024

Study Record Updates

• Last Update Posted (Actual): May 16, 2024
• Last Update Submitted That Met QC Criteria: May 14, 2024
• Last Verified: May 1, 2024

More Information

Terms related to this study

• Keywords: randomized trial; well-being; digital intervention; conversational agent
• Other Study ID Numbers: 101095530

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