Smoking Cessation Coaching in High-Risk Situations: A Virtual Reality Study

Examining the Effect of Personalized Advice of a Conversational Agent for Smoking Cessation in High Risk Situations: An Experimental Study Using Virtual Reality

Smoking tobacco is an important preventable risk factor for chronic illnesses and premature death and is most prevalent among groups with a lower socio-economic position (SEP). High relapse rates show that smoking cessation interventions are often not sufficiently effective on the long-term. Potential reasons for this limited effectiveness are that these interventions are not tailored to lower-SEP smokers and do not provide sufficient support in situations when the (re)lapse risk is high; that is, high-risk situations (HRSs). A mobile phone application using an automated conversational agent could be a useful approach to promote long-term smoking cessation, as it can be tailored to lower-SEP smokers and provide support at any time of the day (also in HRSs). However, evidence on the effectiveness of this kind of applications is scarce and it is still unclear how automated conversational agents can effectively promote lapse prevention. Therefore, it is important to explore what type of lapse prevention strategies these conversational agents should use in HRSs and how these different types of support are experienced by smokers.

This virtual reality (VR) experiment will examine the preliminary effectiveness and usability of a conversational agent that supports smokers in personal HRSs. More specifically, the investigators primarily aim to examine whether the three different lapse prevention dialogs increase abstinence self-efficacy in adult smokers from different SEP groups during simulated HRSs, compared to a neutral dialog (i.e., control condition). In addition, the investigators examine the effect of the lapse prevention dialogs, compared to the neutral dialog, in simulated HRSs on subjective craving and affect. Finally, the investigators examine how adult smokers from different SEP groups experience the personalized support of a simulated conversational agent in simulated HRSs.

VR will be used to expose smokers to their personal HRSs and let them interact with a conversational agent via a simulated mobile phone. Using computer-based VR technology, three-dimensional environments can be created based on environments that smokers encounter in their daily lives (e.g., their living room or the train station from where they travel to work). This way, controlled but at the same time natural-looking environments can be used to expose smokers to their personal HRSs and measure their responses in this situation.

Study Overview

• Status: Completed
• Conditions: Smoking Cessation
• Intervention / Treatment: Behavioral: Lapse prevention dialog: Boost motivation and self-efficacy; Behavioral: Lapse prevention dialog: Future-selves and implementation intentions; Behavioral: Lapse prevention dialog: Identity-related positive self-talk; Behavioral: Attention control dialog: neutral topic

• Study Type: Interventional
• Enrollment (Actual): 25
• Phase: Not Applicable

Contacts and Locations

Study Locations

• Netherlands
  • Zuid-Holland
    • Leiden, Zuid-Holland, Netherlands, 2333 RC
      • Leiden University Medical Center (LUMC)

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 14 years and older (Adult, Older Adult)
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria:

• 18 years or older
• Being able to read and understand Dutch
• Smoking ≥ 10 cigarettes a day
• Intention to quit smoking sometime in the future
• Being able to wear a VR helmet for approximately 30 minutes, with breaks in between
• Willingness to take and send videos and audio of three locations where the participant is most likely to smoke (i.e., highest chance of smoking when in the environment)

Exclusion Criteria:

• Visual problems (e.g., limited visibility without glasses) that affect viewing VR environments (based on self-report)
• Currently involved in smoking cessation activity or therapy (based on self-report)
• When patients are deemed unfit to participate (due to, for example, psychological problems or medication). This decision is left to the discretion of the responsible researcher.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Prevention
• Allocation: N/A
• Interventional Model: Single Group Assignment
• Masking: None (Open Label)

Number of Arms

1

Arms and Interventions

Participant Group / Arm

Intervention / Treatment

Other: Chatbot coaching; Due to the single-group design there will only be one arm. All participants will be exposed to four different chatbot relapse prevention coaching dialogs, which are presented in a random order.

Behavioral: Lapse prevention dialog: Boost motivation and self-efficacy; In the boost motivation and self-efficacy dialog, the conversational agent will tell the participant that they can successfully resist smoking, try to take away their self-doubts and assert that the participant can and will succeed. For example, the conversational agent will send text messages such as: "Even though the temptation can be high, I have no doubts that you can resist smoking in this situation. If you want something, you can do it. Believe in yourself!".; Behavioral: Lapse prevention dialog: Future-selves and implementation intentions; In the future-selves and implementation intentions dialog, the conversational agent will first tell the participant to imagine themselves in a future in which they successfully quit smoking (i.e., desired future self). The conversational agent can send text messages such as: "Please think about yourself in the future. Imagine that you have quit smoking successfully. Think about the person you will be. What do you look like? What does your life look like? Consider this future image as well as you can.". This procedure will be repeated for a future in which the participant continued smoking (i.e., the undesired future self). Finally, the conversational agent will explain to the participant that it is important to think about how to resist smoking in HRSs (i.e., implementation intentions) to ensure that the participant comes closer to becoming their desired future self as non-smoker and to avoid their undesired future self as a smoker.; Behavioral: Lapse prevention dialog: Identity-related positive self-talk; In the identity-related positive self-talk dialog, the conversational agent will tell the participant to use positive self-talk focused on their identity to motivate themselves to successfully resist smoking. For example, the conversational agent will send text messages such as: "When you feel tempted to smoke, it can help to say positive things to yourself and motivate yourself to resist the temptation. What positive things can you say about yourself to motivate yourself to resist smoking? For instance, 'I am a strong person who can resist smoking' or 'I am a person who is persistent and has control over the urge to smoke'. Now say the sentence(s) you find motivating or your own motivating words to yourself, by thinking it, and repeating it in your head or out loud.".; Behavioral: Attention control dialog: neutral topic; The neutral dialog will act as an attention control condition. In this dialog, the conversational agent will not provide support using relapse prevention strategies. Instead, the conversational agent will start with a short introduction (e.g., "Hi, how are you doing?"), ask questions to show interest (e.g., "How are you feeling today?") and will talk about a neutral topic (e.g., animals; "What is your favorite animal?"), and then closes the conversation (e.g., "It was nice speaking with you. I hope you have a good rest of the day!").

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Abstinence self-efficacy	Smoking abstinence self-efficacy will be assessed with a single question derived from the 'Smoking Abstinence Self-efficacy Questionnaire' (Spek et al., 2013). The six-item questionnaire measures self-efficacy regarding smoking abstinence in six different situations. However, in this experiment the situation is shown using VR (e.g., a personalized VR environment of a participant's favorite bar). Therefore, participants will only answer the question "Are you confident that you will not smoke in this situation?" while being exposed to the VR environment. Total scores range from 0 to 4, with a higher score indicating higher abstinence self-efficacy.	Four times during the experiment which has a duration of 1.5 hours. Assessed directly after each chatbot dialog (T1-T4).

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Phasic (state) tobacco craving	To identify the level of subjective smoking craving experienced (in the VR environments), a single-item will be used (Shadel & Cervone, 2006; Vinci et al., 2020). Participants are asked to indicate how strong their craving for a cigarette is right now. Answers are given on a scale ranging from 1 (no craving at all) to 10 (extremely strong craving), with a higher score indicating stronger subjective tobacco craving.	Four times during the experiment which has a duration of 1.5 hours. Assessed directly after each chatbot dialog (T1-T4).
Positive and negative affect	The level of positive and negative affect is assessed by having participants indicate to what extent they experienced the four basic emotions on a scale from 1 (not at all) to 10 (very much; Versluis et al., 2018). The scores on anger, anxiety and sadness will be averaged to represent negative affect. The average score on negative affect ranges from 3 to 30, with a higher score indicating stronger feelings of negative affect. The happiness-rating will represent positive affect. The average score on positive affect ranges from 1 to 10, with a higher score indicating stronger feelings of positive affect.	At the start of the experiment (baseline) and four times during the experiment, which has a duration of 1.5 hours. Assessed directly after each chatbot dialog (T1-T4)
Acceptance of the conversational agent	Acceptance of the conversational agent will be assessed using a set of six items. Two example items are: "How satisfied were you with the virtual coach?" and "How easy was it talking to the virtual coach?". Answers are given on a 7-point scale. The questionnaire was originally used in an intervention study (Provoost et al., 2020). Three of the items of the questionnaire had to be adjusted to make them applicable to the experimental set-up of this study. For example, the item "How much would you like to continue working with the virtual coach if the course continued?" was changed into "How much would you like to continue working with the virtual coach if it was available on your smartphone?". Total scores range from 6 to 42, with a higher score indicating higher acceptance of the conversational agent.	At the end of the experiment (post-intervention); experiment has a duration of 1.5 hours
User-experience / usability	The usability of the conversational agent will be measured with a number of items of the System Usability Scale (Brooke, 1996). The total questionnaire consists of ten items and items are rated on a 5-point Likert scale ranging from 0 (strongly disagree) to 4 (strongly agree). Four questions of the original questionnaire are used in this study. In addition, to get a better in-depth understanding of the participants' user experience with the virtual coach, the following open-ended question will be asked: "Would it be useful to have a virtual coach available on your phone to assist with smoking cessation? Please explain.".	At the end of the experiment (post-intervention); experiment has a duration of 1.5 hours

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Sense of presence	To gain insight into whether the participant had the impression of being in the virtual environment. The Single-Item Measure of Presence will be used (Bouchard et al., 2004). Participants will indicate on a scale ranging from 0 to 10 to what extent they experience presence, with higher scores representing a higher sense of presence.	Four times during the experiment, which has a duration of 1.5 hours. Assessed directly after each chatbot dialog (T1-T4)
Demographic characteristics	General information about the participants will be obtained, namely gender (i.e., male, female, gender neutral, I would rather not say), age (birth year) and level of education (i.e., primary education, vmbo, havo, vwo, mbo, hbo, wo; as measure for SEP).	At the start of the experiment (baseline); experiment has a duration of 1.5 hours
Previous experience with VR	Participants familiarity with VR was assessed using the following question: "Have you ever used virtual reality before". If participants answers this question with yes, they will be asked how frequently they have used VR.	At the start of the experiment (baseline); experiment has a duration of 1.5 hours
Intention to quit smoking	Participants indicate whether they intend to quit in the future. Answer categories are based on the stages of change of the Transtheoretical model (Prochaska & DiClemente, 2005) and include: within the next month (1), between 1 to 6 months from now (2), sometime in the future, beyond 6 months (3) or not planning to quit (4; Hummel et al., 2018). Participants that choose answer category four will be categorized as 'non-intenders'. This variable will be recoded so that higher scores indicate stronger intention to quit.	At the start of the experiment (baseline); experiment has a duration of 1.5 hours
Tonic (trait) tobacco craving	Trait tobacco craving will be assessed using the craving subscale of the Wisconsin Smoking Withdrawal Scale (Welsch et al., 1999). The subscale consists of four items that are scored on a 5-point scale ranging from 0 (strongly disagree) to 4 (strongly agree). An example item is "I have had frequent urges to smoke".	At the start of the experiment (baseline); experiment has a duration of 1.5 hours
Physical nicotine dependence.	Physical nicotine dependence will be assessed at baseline using the Dutch version of the Fagerström Test of Nicotine Dependence (Heatherton et al., 1991; Penfornis et al., 2021).The questionnaire consists of six items, of which four items are scored on a dichotomous scale (yes / no), one item is scored using four answer categories, and one item has an open answer format (asking participants about the number of cigarettes smoked per day). An example item is "Do you smoke more in the morning than during the rest of the day?". Total scores can range from 5 to 12, with a higher score indicating higher nicotine dependence.	At the start of the experiment (baseline); experiment has a duration of 1.5 hours
Latency since last cigarette smoked	To assess the time since last exposure to nicotine, participants were asked how long ago they had their last cigarette. Participants will be able to specify both the number of hours and/or minutes.	At the start of the experiment (baseline); experiment has a duration of 1.5 hours

Collaborators and Investigators

• Sponsor: Leiden University Medical Center
• Investigators: Principal Investigator: Anke Versluis, Dr., Leiden University Medical Center (LUMC); Principal Investigator: Eline Meijer, Dr., Leiden University Medical Center (LUMC)

Study record dates

Study Major Dates

• Study Start (Actual): November 4, 2022
• Primary Completion (Actual): April 19, 2023
• Study Completion (Actual): April 19, 2023

Study Registration Dates

• First Submitted: November 28, 2022
• First Submitted That Met QC Criteria: December 8, 2022
• First Posted (Actual): December 16, 2022

Study Record Updates

• Last Update Posted (Actual): May 25, 2023
• Last Update Submitted That Met QC Criteria: May 24, 2023
• Last Verified: May 1, 2023

More Information

Terms related to this study

• Keywords: Smoking; Relapse prevention; Virtual Reality; eHealth; Smoking Cessation; Coaching; Chatbot; High-Risk situations
• Other Study ID Numbers: N22.013 METC-LDD; 628.011.211 (Other Grant/Funding Number: NWO program Commit2Data - Big Data & Health)

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: YES
• IPD Plan Description: Pseudonymized data will be made available for reuse at the end of the project. Certain demographic characteristics might have to be removed from the data set, to prevent traceability of individual participants. It has yet to be decided whether data can best be made available through a repository or through reasonable request.

Drug and device information, study documents

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