Cerina-A Pilot Randomized Controlled Trial

The study is a pilot Randomised Controlled Trial (RCT) with two conditions: Treatment and a waitlist control group. Included eligible participants who have completed the baseline questionnaire will be randomized either into the treatment group in which they have access to the Cerina application for six weeks or into the wait-list control group in which they have access to the campus-based wellbeing services (i.e. routine care) until the intervention group completes Cerina. They will then receive access to Cerina. The randomization scheme will be derived using random allocation software by the Cerina technical lead who is not involved in the study.

All participants will be asked to fill in online self-report measures of anxiety, worrying, depression, and general functioning at baseline (t0), at 3 weeks (t1), and 6 weeks (t2). Additionally, semi-structured interviews will be conducted with the participants in both groups at 6 weeks (t2). The topic guide for the interview was developed by the Cerina team based on the Consolidated Framework for Implementation Research (CFIR), and it is finalized in consultation with the Principal Investigator (PI). The core topics included participants' overall experience and perceptions on the usability of the user interface of Cerina, existing pathways and barriers to using the app in daily life, the clinical utility of the application, the contexts in which the implementation of the clinical content take place and the processes of intervention delivery. The data collection will last 40 minutes, will be remote (online via the MS teams account of Ulster University), and will be facilitated by the PI and the researchers from the Ulster University. Sample size will be determined by data saturation, and sampling will be iterative to ensure theoretical saturation.

Intervention The intervention consists of 7 sessions of CBT for the treatment of anxiety and worry symptoms as part of GAD. Each session contains a range of information and tasks/exercises to help the user understand the condition of GAD, the treatment approach, and how it will apply to them. Once all sessions have been completed, the user can go back over any of the sessions. There are anxiety management exercises, which the user can go to whenever they wish. There are also a therapy reflection journal and self-care resources including further anxiety management techniques, resources, and podcasts that the user will have access to whenever they want. The names of the sessions are: Learning about GAD, dealing with worry, managing worry and anxiety, beliefs about worry, managing uncertainty, testing beliefs, review, and therapy blueprint.

Wait-list Control Condition Participants in the waiting-list control condition will have access to the campus-based well-being services offered by the Student Wellbeing team. The Wellbeing team is based on Belfast, Coleraine, and Magee campuses and provides free and confidential support and guidance to students with a broad range of issues, concerns, and challenges, to help them successfully engage in their studies. Student Wellbeing Assistants provide an initial assessment to determine a student's primary need and then Wellbeing Advisers, Student Mental Health Advisers, and Accessibility Advisers are available to provide a variety of solution-focused interventions based on that need. The Accessibility Advisers also help students with disabilities including diagnosed mental health conditions to get additional support and access funded disability support through Disabled Students' Allowance. Each campus team is led by an experienced Student Wellbeing Manager to support the advisers in the management of risk and lead to an efficient and robust response if a Clinical Incident occurs. Additionally, therapeutic counselling support is also available free to students through the external counselling provider via a dedicated 24/7 counselling helpline. The Student Wellbeing Team works closely with the counselling provider to monitor student engagement and promote the service as appropriate.

1. Statistical techniques

Sample size We seek to recruit 90 participants (N=45 in treatment, N=45 in Wait-list control) in total. The audit of the sample sizes for the pilot and feasibility RCTs indicated that the median sample size per arm across all the types of studies was 30 (Bilingham, Whitehead \& Julious, 2013). Browne also recommended that 30 participants per condition are needed to estimate a parameter (Browne, 1995). We take a possible dropout rate into account and aim to include 50% more participants than initially required in both conditions.

Outcome measures:

The outcome measures are online (Qualtrics) self-report questionnaires. The primary outcome measure is anxiety symptoms measured by the Generalised Anxiety Disorder Scale-7 (GAD-7). The secondary outcome measures are: worrying measured by the Penn State Worry Questionnaire (PSWQ-PW), depression measured by the Patient Health Questionnaire (PHQ-9), general functioning measured by the Work and Social Adjustment Scale (WSAS), and usability of the intervention measured by the System Usability Scale (SUS) and semi-structured interviews.

Data analysis plan:

The recruitment and consent rates will be carefully monitored according to the CONSORT guidelines. The group allocation procedure will be monitored by the technical lead of Cerina (independent researcher). The response rates to the study questionnaires and the adherence rates will be monitored by Dr. Özlem Eylem-van Bergeijk.

Participants' views on the quality and usability of the User-Interface Design will be investigated through the SUS questionnaire and semi-structured interviews. These interviews will be audio-recorded and transcribed. The transcripts will be analyzed thematically and independently by two researchers in NVivo software.

The preliminary effects of Cerina in reducing elevated levels of anxiety and worry among students will be tested through the linear mixed models in RStudio version 3.6.1. We will compare reductions in primary/secondary outcomes between and within groups across two time points in the intention-to-treat sample, we will use linear mixed models in RStudio version 3.6.1. This method allows the number of observations to vary between participants and handles missing outcome data. The mixed model uses a longitudinal data structure that includes both fixed and random effects. Time (categorical), group (treatment v. wait-list control), and interactions between group and time will be included as fixed effects in mixed models together with a random intercept and random time effect. Differences in least-squares mean (intervention effects) at each time point with 95% confidence intervals will be derived. Cohen's d for the effect of the intervention will be estimated by calculating the difference between estimated means (corrected for baseline) divided by raw pooled standard deviation. A two-sided p < 0.05 indicates statistical significance. Missing data will be handled through multiple imputations.

We will also monitor people's help-seeking behavior throughout the study, and we will look at the differences between those who used Cerina alone compared to those who used Student Wellbeing and/or counselling services in addition to Cerina.

The reliable change index will also be used to evaluate whether participants have reliable and clinically significant change scores from baseline to post-assessment.