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
ACTRN12616000579493
Ethics application status
Approved
Date submitted
28/04/2016
Date registered
4/05/2016
Date last updated
2/08/2023
Date data sharing statement initially provided
27/03/2019
Date results provided
27/04/2023
Type of registration
Prospectively registered

Titles & IDs
Public title
Imagery-enhanced versus verbally-based group cognitive behavioural therapy for social anxiety disorder.
Scientific title
Imagery-enhanced versus verbally-based group cognitive behavioural therapy for social anxiety disorder: a randomised controlled trial.
Secondary ID [1] 289059 0
None
Universal Trial Number (UTN)
Trial acronym
Linked study record

Health condition
Health condition(s) or problem(s) studied:
Social anxiety disorder 298502 0
Condition category
Condition code
Mental Health 298588 298588 0 0
Anxiety

Intervention/exposure
Study type
Interventional
Description of intervention(s) / exposure
This project aims to examine the efficacy, mechanisms of change, and cost-effectiveness of a novel group cognitive behavioural therapy, imagery-enhanced cognitive behavioural group therapy (IE-CBGT), compared to gold standard verbally-based cognitive behavioural group therapy (VB-CBGT), in patients with social anxiety disorder (SAD).

Both protocols target the same maintaining factors (negative cognitions, avoidance, safety behaviours, self-focused attention, negative beliefs about how others’ perceive them, and core beliefs), but differ critically in the mode within which they are modified. The imagery-enhanced protocol encourages patients to focus on multi-sensory negative social images as a key maintaining factor and imagery monitoring, rather than focusing on verbal thoughts and thought monitoring. Cognitive challenging occurs within imagery rather than only using traditional verbal-linguistic methods. Behavioural experiment predictions are guided by negative imagery rather than thoughts, new information is incorporated into imagery, and coping imagery is constructed. Imagery exercises are included in all within-session behavioural experiments. Negative core beliefs are identified and challenged in IE-GCT via past imagery rescripting techniques, rather than more traditional verbal cognitive methods.

Both protocols involve 12, 2-hour group therapy sessions (once/week) plus a one month group-based follow-up session of 2 hours duration. Groups comprise of between 6 and 12 patients. All treatment sessions will be facilitated by a registered psychologist with a masters or doctoral degree in clinical psychology. Clinicians will maintain an attendance register. Protocol adherence and homework completion will be monitored throughout the trial through coding of in-session audio recordings and use of the Homework Rating Scale (HRS-II), respectively. Additionally, the Working Alliance Inventory (WAI) and Gross Cohesion Scale will be used to measure therapist-client alliance and cohesiveness/bond amongst group members.
Intervention code [1] 294566 0
Treatment: Other
Intervention code [2] 294630 0
Behaviour
Comparator / control treatment
The verbally-based protocol uses verbal-linguistic techniques with no reference to imagery, except within the context of challenging negative self-images via video-feedback. Patients are encouraged to challenge their beliefs regarding negative feelings associated with social situations using a range of structured cognitive behavioural exercises, which includes identifying perpetuating factors (e.g., safety behaviours), psychoeducation, thought monitoring, gathering contrary evidence, and developing new core beliefs.
Control group
Active

Outcomes
Primary outcome [1] 298084 0
Social Interaction Anxiety Scale (SIAS) - a 20-item measure of interaction/performance anxiety with high internal and test-retest reliability,
Timepoint [1] 298084 0
Baseline, immediately prior to 4th and 8th treatment sessions, and immediately after the 12th treatment session and 1- and 6-month follow-ups.
Primary outcome [2] 298085 0
Diagnosis of social anxiety disorder on structured clinical interview (Structured Clinical Interview for DSM-5 [SCID-5]).
Timepoint [2] 298085 0
Baseline, 1-month follow-up, and 6-month follow-up.
Primary outcome [3] 298086 0
Clinician-rated anxiety severity. Clinical severity of each disorder will be rated on a 0–8 scale reflecting severity relative to other people with the disorder and life impact.
Timepoint [3] 298086 0
Baseline, 1-month follow-up, and 6-month follow-up.
Secondary outcome [1] 323145 0
Duration of observable safety (anxiety-mediated) behaviours assessed during a social speech task, the Trier Social Stress Test (TSST).

In the TSST participants are asked to perform an impromptu speech about a topic of their own choosing to the experimenter and a video camera. Participants are allowed to talk on any topic that is of personal relevance to them aside from their participation in the research. Safety behaviours will be rated by trained coders who are blind to group allocation.
Timepoint [1] 323145 0
Baseline and 1-month and 6-month follow-ups.
Secondary outcome [2] 323146 0
Psychophysiological responding (heart rate variability and skin conductance level) will also be obtained throughout the TSST.

The procedure begins with a 5 min baseline rest during which physiological activity is recorded. They are then given instructions for the subsequent task and 3 min of preparation time. After this they are instructed to talk for 5 min. The task concludes with a further 3 min baseline to assess recovery to baseline responding.

Skin conductance level (SCL) will be recorded with two disposable electrodes attached to the thenar and hypothenar prominences of their non-dominant hand.

Heart rate variability will be measured via electrocardiogram (ECG).

Respiration will be recorded using a chest strap transducer attached to the participant’s lower chest. It is used as a control measure for the main measures SCL and ECG.
Timepoint [2] 323146 0
Baseline and 1-month and 6-month follow-ups.
Secondary outcome [3] 323147 0
Depression severity using PROMIS scale of depressive symptomatology. The PROMIS scale is an 8-item scale of depression severity.
Timepoint [3] 323147 0
Baseline, prior to all treatment sessions (x 12), and immediately after 1- and 6-month follow-ups.
Secondary outcome [4] 323330 0
Negative Self Portrayal Scale (NSPS) - measures concern about perceived negative self-attributes (e.g., lacking personality, stuttering) exposed for public scrutiny in social situations.
Timepoint [4] 323330 0
Baseline, immediately prior to 4th and 8th treatment sessions, and immediately after the 12th treatment session and 1- and 6-month follow-ups.
Secondary outcome [5] 323331 0
Cognitive Avoidance Questionnaire (CAQ) - a key imagery-enhanced group cognitive therapy-specific mediator measuring imagery suppression; tendency to actively transform distressing images into less distressing verbal thoughts.
Timepoint [5] 323331 0
Baseline, immediately prior to 4th and 8th treatment sessions, and after the 12th treatment session and 1- and 6-month follow-ups.
Secondary outcome [6] 323332 0
Vividness of Visual Imagery Questionnaire (VVIQ) - provides a valid and highly reliable psychometric measure of self-reported vividness of visual imagery.
Timepoint [6] 323332 0
Baseline, immediately prior to 4th and 8th treatment sessions, and immediately after the 12th treatment session and 1- and 6-month follow-ups.
Secondary outcome [7] 323333 0
Brief Fear of Negative Evaluation (BFNE) - a self-report measure of fear or worry about being negatively evaluated.
Timepoint [7] 323333 0
Baseline, immediately prior to all treatment sessions (x 12), and after 1- and 6-month follow-ups.
Secondary outcome [8] 323334 0
Fear of Positive Evaluation (FPE) - measures fear and distress related to positive evaluation from others, a core feature of social anxiety.
Timepoint [8] 323334 0
Baseline, immediately prior to all treatment sessions (x 12), and after 1- and 6-month follow-ups.
Secondary outcome [9] 323335 0
Repetitive Thinking Questionnaire (RTQ-10) - a transdiagnostic measure of repetitive negative thinking with high internal reliability and sensitivity to change.
Timepoint [9] 323335 0
Baseline, and immediately after 1- and 6-month follow-ups.
Secondary outcome [10] 323336 0
Self Consciousness Scale (SCS) - a measure of self-focussed attention, including items assessing tendency to take an external, observer’s perspective during interactions.
Timepoint [10] 323336 0
Baseline, immediately prior to 4th and 8th treatment sessions, and immediately after the 12th treatment session and 1- and 6-month follow-ups.
Secondary outcome [11] 323337 0
Social Phobia Scale-6 (SPS) - a 6-item measure of performance anxiety.
Timepoint [11] 323337 0
Baseline, immediately prior to 4th and 8th treatment sessions, and immediately after the 12th treatment session and 1- and 6-month follow-ups.
Secondary outcome [12] 323338 0
Subtle Avoidance Frequency Examination (SAFE) - measures safety behaviours and subtle avoidance.
Timepoint [12] 323338 0
Baseline, immediately prior to 4th and 8th treatment sessions, and immediately after the 12th treatment session and 1- and 6-month follow-ups.
Secondary outcome [13] 323339 0
Self-Beliefs Related to Social Anxiety Scale (SBSA) - assesses three types of beliefs characteristic of social phobia (high standards, and conditional and unconditional beliefs).
Timepoint [13] 323339 0
Baseline, immediately prior to 4th and 8th treatment sessions, and immediately after the 12th treatment session and 1- and 6-month follow-ups.
Secondary outcome [14] 323340 0
For health economic analysis the primary outcome measure will be the EQ-5D-5L, a standardised self-report measure of health-related quality of life providing a single value for health status.
Timepoint [14] 323340 0
Baseline, and immediately after 1- and 6-month follow-ups.
Secondary outcome [15] 323341 0
TiC-P Adults Questionnaire on Medical Consumption and Productivity Losses associated with Psychiatric Illness - measures health services used by patients and costs arising as a result of productivity losses.
Timepoint [15] 323341 0
Baseline and immediately after 6-month follow-up.
Secondary outcome [16] 323342 0
Homework Rating Scale II - Client Version (HRS-II) - consists of 12 questions regarding homework completion over the past week.
Timepoint [16] 323342 0
Prior to treatment sessions 2, 5, 8, and 11.
Secondary outcome [17] 323343 0
Working Alliance Inventory (WAI) - a 36-item measure of therapist-client alliance.
Timepoint [17] 323343 0
Prior to treatment sessions 3, 6, and 9.
Secondary outcome [18] 323344 0
Group Cohesion Scale (GCS) - a brief 9-item measure designed to assess the degree of perceived cohesiveness and bond among members of a group.
Timepoint [18] 323344 0
Prior to treatment sessions 4, 7, and 10.
Secondary outcome [19] 325371 0
Probability and cost tracking measure. Instruction: You are asked to give a 3-minute speech on a topic of your choice. Your speech will be video-taped and watched by two people who will rate various aspects of your performance. What do you most fear happening in this situation? [participant writes a feared outcome]. Participant is then asked, if you were going to enter this situation today, How likely is it that this would happen? (0 = not at all likely, 4 = moderately likely, 8 = extremely likely) and How bad would it be for you if this happened? (0 = not at all bad, 4 = moderately bad, 8 = extremely bad.).
Timepoint [19] 325371 0
Participants rate the probability (likelihood) and cost (how bad would it be) of the feared social consequence of a 3-minute speech task before each session (sessions 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, plus 1-month and 6-month follow-up).

Eligibility
Key inclusion criteria
At least 18 years of age, principle diagnosis of social anxiety disorder (as assessed by the SCID-5), Stable medications for at least 1 month, and willingness to be randomised.
Minimum age
18 Years
Maximum age
No limit
Sex
Both males and females
Can healthy volunteers participate?
No
Key exclusion criteria
Bipolar disorder/psychosis or substance use disorder (all as judged by the SCID-5), currently receiving CBT elsewhere, high suicidal or self-harm risk (i.e., plans and/or intent).

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 involves contacting the holder of an allocation schedule (study statistician: Co-Investigator Dr Robert Kane) who is located at the School of Psychology and Speech Pathology at Curtin University, and is not on the same site as the clinician who determines study eligibility.
Methods used to generate the sequence in which subjects will be randomised (sequence generation)
Randomisation will be generated at www.random.org by the study statistician (AI Kane) prior to recruitment.
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
Phase
Not Applicable
Type of endpoint/s
Efficacy
Statistical methods / analysis
Analyzing continuous treatment outcomes:
Non-linear change across the duration of the study is expected. In these circumstances, the most powerful approach to data analysis (regardless of the significance of the Time x Group interaction) is to treat time as an ordered categorical (i.e., ordinal) variable and focus on between-group comparisons at each post-treatment assessment (e.g., post-treatment, 1-month follow-up, 6-month follow-up) after controlling for between-group differences at baseline. The 1-month follow-up is nominated as the primary endpoint, because historical data suggest that differences between treatments will be optimal at that point. If the outcome measure does not have a normal distribution, but is nevertheless distributed according to a well-understood function (e.g., gamma, inverse Gaussian), then between-group comparisons will be tested via a Generalized Linear Mixed Model (GLMM) that specifies this function. Two types of GLMM will be tested. The potential clustering of outcome data within treatment groups will be acknowledged in one model but not the other. The better fitting model will be selected. The method of analysis will depend on whether the slope of the treatment trajectories is non-linear (in which case time would once again be treated as an ordered categorical variable and analyzed as before) or linear (indicating a piecewise growth curve analysis). Standardized rates of reliable and clinically significant change will be calculated and compared at 1- and 6-month follow-up.

Analyzing the binary diagnostic outcome:
A GLMM will be tested to determine whether the imagery-enhanced group contributes to a better diagnostic outcome for SAD than the verbally-based group. The GLMM will include one nominal random effect (participant), one nominal fixed effect (group: imagery-enhanced, verbally-based), one ordinal fixed effect (time: 1-month follow-up, 6-month follow-up), and the Group x Time interaction. The GLMM will use a binomial probability distribution for the outcome (SAD: yes, no) and link it to the fixed effects with a logit function. Once again, two types of GLMM will be tested. The potential clustering of outcome data within treatment groups will be acknowledged in one model but not the other. The better fitting model will be selected. Simple main effects tests will be used to test for between group differences in caseness at 1- and 6-month follow ups.

Testing moderated mediation:
Structural Equation Modeling (SEM), with single indicator latent variables to control for measurement error, will be used to determine i) whether the relationship between time (Session 4, Session 8, post-treatment, 1-month follow-up, 6-month follow-up) and SIAS is mediated by changes in negative self-beliefs, repetitive negative thinking, self-focused attention, and avoidance and safety behaviors, and ii) whether group (IE-CBGT, VB-CBGT) moderates the strength of the time-mediator pathway and the mediator-symptomatology pathway. The approach to the moderated mediation analysis follows that of Hofmann and colleagues. Bootstrapping procedures will be used to estimate the significance of the mediation pathways; the error inherent in the psychometric measures will be included in the estimation of the SEMs; any clustering of outcome data within treatment groups will be accommodated by using a sandwich estimator for the path coefficients’ standard errors.

Balancing Type I and Type II error rates:
Conducting multiple statistical tests on the secondary outcomes will inflate the familywise error rate (FWER; the probability of at least one false positive). Inflation of the familywise error rate is normally controlled by applying a Bonferroni (or a modified Bonferroni) adjustment to the per-test alpha-level. Controlling FWER guards against false positives (Type I errors) at the cost of missing real effects (Type II errors). Instead of controlling the FWER, however, we could control the false discovery rate (FDR; the expected proportion of false positives). In contrast to FWER controlling procedures, FDR controlling procedures guard against missing real effects (Type II errors) at the cost of an increased number of false positives (Type I errors). In order to optimize statistical power in this study, the FDR rather than the FWER will be controlled. Analyses of the primary outcomes (SIAS, SAD diagnosis, and severity) are hypothesis-driven and will therefore be evaluated at the conventional per-test alpha level of .05.

Sample size estimation/Power analysis:
We are planning for a minimum of 12 groups with an average of 8/group (N = 96) to ensure feasibility to detect a small to medium effect size, as indicated by our pilot work. However, we will continue to recruit for the duration of the funding period (from July 2016 until December 2019) to derive a larger sample size so that smaller effect sizes can be detected. Larger numbers will enable greater precision in effect size estimates and will also allow for multiple comparisons and additional analyses of multiple moderators/mediators. Assuming negligible variance is accounted for by dependencies on the outcome measures within the 12 treatment groups, ninety-six participants (48 in each group) are required for an 80% chance of detecting a ‘moderate’ (d = .615) difference between the two group means at a Bonferroni-adjusted alpha-level of .017 (= 3 congeneric outcomes). In order to increase the probability of capturing smaller effects, however, we will accommodate multiple testing by controlling the FDR.

There are many procedures for estimating an adequate sample size for the binary logistic regression - each procedure based on its own prescriptive set of assumptions. We estimated the sample size required for capturing a ‘moderate’ association between group (imagery-enhanced, verbally-based) and recovery (yes, no) at each post-treatment assessment. Assuming once again that negligible variance is accounted for by intra-group dependencies on the diagnostic measure, 96 participants (48 in each group) will provide sufficient power for an 80% chance of detecting a ‘moderate’ (w = .29) association between group and recovery.

Each of the four moderated mediator models (one for each of the four mediators) includes 15 free parameters (i.e., parameters that must be estimated from the data). In order to reliably test an SEM, Our planned sample size of 96 exceeds the minimum sample size (N = 75) required for testing our model (a minimum of 5 participants per parameter is recommended), hence contributing to the appropriateness of implementing SEM.

Handling Missing Data:
Missing outcomes can potentially bias the results of clinical trials. We will limit the frequency of missing data by attempting to collect follow up measures from all participants, including those who do not complete treatment. Participants will receive a series of phone, text, and email contacts to schedule, and reschedule (where necessary), assessment appointments. If participants refuse to return to the clinic to complete a full assessment, permission will be sought to collect diagnostic interview data via telephone as a minimum. Contact will be made within three days of each assessment via a phone call and one text message reminder. All missed treatment sessions will be followed up by the treating clinician by phone call and/or text message, plus an email with the weekly questionnaires to be completed and returned at the next session. Individual sessions will be allowed if clients are in crisis or specifically request this, which will be noted for consideration in subsequent analyses. Analyses will utilize techniques (such as GLMM and multiple imputation) that make full use of all observed data and help correct for potential biases caused by missing observations. We will conduct sensitivity analyses to evaluate the potential impact of missing data on the robustness of our findings.

Recruitment
Recruitment status
Completed
Date of first participant enrolment
Anticipated
15/07/2016
Actual
14/07/2016
Date of last participant enrolment
Anticipated
30/06/2019
Actual
4/02/2019
Date of last data collection
Anticipated
26/03/2020
Actual
5/11/2019
Sample size
Target
96
Accrual to date
Final
107
Recruitment in Australia
Recruitment state(s)
WA
Recruitment hospital [1] 5716 0
Centre for Clinical Interventions - Northbridge
Recruitment postcode(s) [1] 13167 0
6003 - Northbridge

Funding & Sponsors
Funding source category [1] 293440 0
Government body
Name [1] 293440 0
National Health and Medical Research Council of Australia (NHMRC)
Country [1] 293440 0
Australia
Funding source category [2] 293441 0
University
Name [2] 293441 0
Curtin University
Country [2] 293441 0
Australia
Primary sponsor type
Individual
Name
Associate Professor Peter McEvoy
Address
School of Psychology and Speech Pathology

Curtin University

Kent Street

Bentley WA 6102
Country
Australia
Secondary sponsor category [1] 292264 0
None
Name [1] 292264 0
Address [1] 292264 0
Country [1] 292264 0

Ethics approval
Ethics application status
Approved
Ethics committee name [1] 294886 0
WA Health North Metropolitan Health Service Mental Health Human Research Ethics Committee
Ethics committee address [1] 294886 0
Ethics committee country [1] 294886 0
Australia
Date submitted for ethics approval [1] 294886 0
21/03/2016
Approval date [1] 294886 0
28/04/2016
Ethics approval number [1] 294886 0
04_2016
Ethics committee name [2] 294917 0
Curtin University Human Research Ethics Committee
Ethics committee address [2] 294917 0
Ethics committee country [2] 294917 0
Australia
Date submitted for ethics approval [2] 294917 0
29/04/2016
Approval date [2] 294917 0
Ethics approval number [2] 294917 0

Summary
Brief summary
Trial website
Trial related presentations / publications
Public notes

Contacts
Principal investigator
Name 65334 0
Prof Peter McEvoy
Address 65334 0
Curtin University, School of Psychology, Kent Street Bentley WA 6102
Country 65334 0
Australia
Phone 65334 0
+61 8 9266 5110
Fax 65334 0
+61 8 9266 2464
Email 65334 0
[email protected]
Contact person for public queries
Name 65335 0
Peter McEvoy
Address 65335 0
Curtin University, School of Psychology, Kent Street Bentley WA 6102
Country 65335 0
Australia
Phone 65335 0
+61 8 9266 5110
Fax 65335 0
+61 8 9266 2464
Email 65335 0
[email protected]
Contact person for scientific queries
Name 65336 0
Peter McEvoy
Address 65336 0
Curtin University, School of Psychology, Kent Street Bentley WA 6102
Country 65336 0
Australia
Phone 65336 0
+61 8 9266 5110
Fax 65336 0
+61 8 9266 2464
Email 65336 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
With respect to data retention: Participants were informed that "You will be assigned a unique ID, which is stored on a password-protected database containing your personal details, accessible only to select study staff to maintain confidentiality." Participants provided the following consent: "I agree for my data to be retained in a restricted use data repository for 25 years at Curtin University, per the Western Australian University Sector Disposal Authority policy." Given that explicit written consent was not sought to publish or share data outside of the research team, this will require an ethics amendment to ensure that sharing occurs within ethical guidelines. An amendment will be sought within one year of the study completion. If this amendment is approved, this sharing item within the registry will be updated. The researchers support open science and will ensure consent is sort for sharing/publishing de-identified data in future studies.


What supporting documents are/will be available?

Doc. No.TypeCitationLinkEmailOther DetailsAttachment
7412Study protocolMcEvoy, P. M., Mould, M. L., Grisham, J. R., Holmes, E. A., Moscovitch, D. A., Hendrie, D., Saulsman, L. M., Lipp, O. V., Kane, R. T., Rapee, R. M., Hyett, M. P., & Erceg-Hurn, D. M. (2017). Assessing the efficacy of imagery-enhanced cognitive behavioral group therapy for social anxiety disorder: Study protocol for a randomized controlled trial. Contemporary Clinical Trials, 60, 34-41.https://www.sciencedirect.com/science/article/pii/S1551714417301088 
19886Statistical analysis plan    Also available at https://osf.io/msq9w/ 370569-(Uploaded-13-06-2023-14-59-00)-Study-related document.pdf
19887Data dictionary    Data Dictionary 370569-(Uploaded-13-06-2023-14-59-45)-Study-related document.pdf



Results publications and other study-related documents

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

Documents added automatically
SourceTitleYear of PublicationDOI
EmbaseAssessing the efficacy of imagery-enhanced cognitive behavioral group therapy for social anxiety disorder: Study protocol for a randomized controlled trial.2017https://dx.doi.org/10.1016/j.cct.2017.06.010
N.B. These documents automatically identified may not have been verified by the study sponsor.