Usage and Health Effects of Embodied Conversational Agents Among Older Adults

Rationale: Embodied Conversational Agents (ECAs), could be a highly effective medium to address health behaviour change among older adults. As compliance to health advice is important for positive health outcomes, successful design of persuasive ECAs can have huge health benefits. However, insights in the mechanisms underlying usage and health behaviour change via ECAs are lacking.

Objective: The objective is to unravel the mechanisms behind the use of an ECA intervention, and understand the mechanism behind the observed behaviour change

Study design: A randomized staggered-entry waitlist-controlled trial will be carried out.

Study population: The study population consist of Dutch-speaking older adults, who live independently, are without partner, are retired, 65+ and possess basic computer skills.

Intervention: The application PACO has been created for (and with) older adults with the goal to motivate them to improve their eating behaviour and decrease their feelings of loneliness.

Main study parameters/endpoints: The main study parameters are use, loneliness, and eating behaviour.

Nature and extent of the burden and risks associated with participation, benefit and group relatedness: Subjects are not exposed to any risks, nor have they any costs. They do have to fill in questionnaires and use the application. The duration and data collection moments are needed to gain a fine-grained understanding of the use, relationship development and health change process. For subjects, the main benefit is to gain insight in their health behaviour via the PACO-application. Although this might be experienced as confronting by some. The technology was developed based on sound theories, with input from the target group, so the investigators expect positive experiences and an improvement in participants loneliness and eating behaviour. However, this can only be proven after the study.

Study Overview

• Status: Completed
• Conditions: Loneliness; Eating Behavior; Technology Use
• Intervention / Treatment: Behavioral: PACO; Behavioral: Waiting list + PACO

• Study Type: Interventional
• Enrollment (Actual): 52
• Phase: Not Applicable

Contacts and Locations

Study Locations

• Netherlands
  • Gelderland
    • Wageningen, Gelderland, Netherlands, 6706 KN
      • Wageningen University & Research

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 63 years and older (Older Adult)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

Description

Inclusion Criteria:

• Living independently at home
• Living without a partner
• No paid profession and at least 65 years of age
• Dutch speaking
• Able to use a tablet or computer
• Internet connection at home

Exclusion Criteria:

• Not willing to provide inform consent

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Supportive Care
• Allocation: Randomized
• Interventional Model: Sequential Assignment
• Masking: Single

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: PACO; Participants are asked to use the intervention for eight weeks.	Behavioral: PACO; The intervention is a fully-automated web-based eHealth application in which two embodied conversational agents engage in dialogue with older adults in order to motivate them towards dietary behavior change and decrease loneliness. The application consists of five different modules, each one applying different behavior change techniques.
Other: Waiting list + PACO; Participants receive the same intervention, but only after a waiting period of four weeks.	Behavioral: Waiting list + PACO; Waiting list + PACO

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Logdata on use application	Amount of time spend on the application during the 8 weeks intervention period	8 weeks
Change in De Jong Gierveld Loneliness Scale	Score from 1-5, higher score means a better outcome.	Four weeks before intervention to eight weeks
Change in eating behaviour based on 24 hours recall		Four weeks before intervention to eight weeks

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Enjoyment via affect scale	Pleasure in using the application. Score from 1-5, higher score means a better outcome.	Eight weeks after intervention
Classic aesthetics Scale	The design of the application. Score from 1-7, higher score means a better outcome.	Eight weeks after intervention
Concern for privacy scale	Concerns about sharing personal data, e.g.name and date of birth. Score from 1-7, higher score means a better outcome.	Four and eight weeks after intervention
Active control scale	Control over the system, as perceived by the user. Score from 1-7, higher score means a better outcome.	Eight weeks after intervention
System Usability Scale	Whether PACO is easy to use. Score from 1-5, higher score means a better outcome.	Eight weeks after intervention
Perceived usefulness scale	Score from 1-5, higher score means a better outcome.	Eight weeks after intervention
Rapport scale	Relationship with the embodied conversational agent. Score from 1-5, higher score means a better outcome.	Four weeks before intervention to eight weeks
Brief Older People's Quality of Life Scake	Score from 1-5, higher score means a better outcome.	Four weeks before intervention to eight weeks
Change in Basic Psychological Need Satisfaction and Frustration Scales	Score from 1-5, higher score means a better outcome.	Four weeks before intervention to eight weeks

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Willingness to pay single questions	Willing to pay for PACO (y/n)	Eight weeks after intervention
User experience via self-formulated open question		Four and eight weeks after intervention

Collaborators and Investigators

• Sponsor: Wageningen University and Research
• Collaborators: ZonMw: The Netherlands Organisation for Health Research and Development; Roessingh Research and Development; WAAG; National Foundation for the Elderly
• Investigators: Principal Investigator: Emely de Vet, PhD, Wageningen University & Research

Publications and helpful links

General Publications

• Kramer LL, van Velsen L, Clark JL, Mulder BC, de Vet E. Use and Effect of Embodied Conversational Agents for Improving Eating Behavior and Decreasing Loneliness Among Community-Dwelling Older Adults: Randomized Controlled Trial. JMIR Form Res. 2022 Apr 11;6(4):e33974. doi: 10.2196/33974.
• Kramer LL, Mulder BC, van Velsen L, de Vet E. Use and Effect of Web-Based Embodied Conversational Agents for Improving Eating Behavior and Decreasing Loneliness Among Community-Dwelling Older Adults: Protocol for a Randomized Controlled Trial. JMIR Res Protoc. 2021 Jan 6;10(1):e22186. doi: 10.2196/22186.

Study record dates

Study Major Dates

• Study Start (Actual): July 13, 2020
• Primary Completion (Actual): December 30, 2020
• Study Completion (Actual): December 30, 2020

Study Registration Dates

• First Submitted: July 13, 2020
• First Submitted That Met QC Criteria: August 11, 2020
• First Posted (Actual): August 12, 2020

Study Record Updates

• Last Update Posted (Actual): April 8, 2021
• Last Update Submitted That Met QC Criteria: April 6, 2021
• Last Verified: April 1, 2021

More Information

Terms related to this study

• Keywords: eHealth; Embodied Conversational Agent; Technology use
• Other Study ID Numbers: 40-44300-98-110

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: Yes
• IPD Plan Description: In accordance with the Data Management Plan, part of the data will be anonymized and made available in DANS-EASY with restricted access.
• IPD Sharing Time Frame: The data will become available within 12 months after the end of the study.
• IPD Sharing Supporting Information Type: Study Protocol; Statistical Analysis Plan (SAP); Informed Consent Form (ICF)

Drug and device information, study documents

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