Study design and intervention
A quasi-experimental multiple-baseline design will compare repeated measurements of the same individual during an active control period to those during the intervention, eliminating the need for a control group [48]. During the active control component, participants will be provided with freely available MHPSS focused resources, collated in partnership with our stakeholders. During the intervention component, MHPSS practitioners in Syria, Türkiye, and Bangladesh will participate in fortnightly 90-minute supervision sessions facilitated by two co-supervisors; one international and one local psychologist, counseling social worker, or counselor.
MHPSS practitioners will be asked to complete online surveys during the active control component and intervention component to examine the effect of the supervision program on the well-being, burnout level, and counseling self-efficacy of MHPSS practitioners (Objectives 1 & 2). MHPSS service users will be interviewed by trained research assistants to rate service satisfaction and quality (Objective 3).
Supervision intervention
The intervention was designed by an Australian clinical psychologist and clinical supervisor with experience providing supervision in both Australia and in humanitarian contexts (SW). The program aims to develop key competencies in mental health practice based on professional practice standards outlined in Supplementary material 1. As the supervision process draws heavily on Western, English-language approaches to supervision, the intervention was adapted in collaboration with local psychologists and psychiatrists (AB, SL, MKM, SJ, OF) following a pilot program [see [49]].
The program uses Australian mental health professionals with specialized training in supervision, along with local supervisors who have professional experience and cultural insight. Supervision will be in groups of 4–6 supervisees, 2 co-supervisors (one Australian and one local supervisor), and a research assistant. Sessions will run for 90 min on Zoom, fortnightly, for 16 months, split into two terms as per postgraduate mental health programs. Groups will be closed once they start, to allow for cohesion and safety. At the end of the first term, participants will be reassigned to a different group with new supervisors and co-supervisees. Supervisor dyads are then placed into a Whatsapp group with a Research Assistant/translator two weeks prior to the supervision groups commencing and will be encouraged to begin a dialogue. The supervisors will be provided with the written guidance and asked to discuss and share ideas with each other.
The supervision models built on reflective and supportive supervision. Reflective supervision helps supervisors guide supervisees to better understand clinical issues. It involves two-way communication and draws on the supervisee’s expertise. This approach is suitable for cross-cultural programs where Australian supervisors may not understand cultural and contextual dynamics. The program also draws on the Integrated Model for Supervision by the International Federation of Red Cross and Red Crescent Societies (IFRC) [12]. Supervisors attend two preparatory workshops and regular reflective group supervision sessions to support problem-solving in a transcultural context. Sessions are designed to focus on case presentations, which are a common format for supervision [50,51,52]. The program takes a flexible, needs-based approach to supervision, given the varied backgrounds of supervisees and supervisors, as well as the unique cross-cultural, cross-discipline, online, and co-supervision factors. Given that there approximately 52 models of clinical supervision, many with limited research support [53], supervisors are encouraged to apply their preferred models based on the needs of supervisees in each session. While not prescribing a specific model of supervision, the program offers readings and training on various supervision models, as well as a handbook with contextual information and suggestions for structure and process.
Participants and recruitment
Study population
The total sample size for the project is 2,300 comprised of the following samples from each of the participant groups:
-
1.
Participant Group 1 MHPSS Clinicians: 100 (2300 within-subject measurements; 23 monthly per clinician). Participant Group 1 will be equally split between the two data collection sites i.e. 50 practitioners sampled from Türkiye and Northwest Syria, and 50 practitioners sampled from Bangladesh.
-
2.
Participant Group 2 Beneficiaries: 2,200 (between-subjects; 22 monthly per clinician).
This sample size is sufficient to meet the research aims and answer the research questions because in longitudinal growth modelling, sample size is calculated based on the number of assessment occasions and does not require large numbers of participants to achieve sufficient statistical power. We have previously conducted Monte Carlo simulations using the same primary outcome to determine that a sample size of 80 is sufficient to achieve a power of 80% with a similar multiple baseline design with 10 measurement occasions [54]. We have oversampled by 20% given the power calculation of 80 because we expect at least 20% attrition in the unpredictable study locations.
Recruitment strategy
MHPSS practitioners will be recruited via the network of the project partners (Hope Revival Organization (HRO) in Türkiye/Syria; Suicide Prevention Sub-Group (SPSG) of the MHPSS Working Group in Bangladesh). MHPSS organisations will be invited to participate in the study. Upon approval to participate in the study, those organisations will be asked to provide a list of consenting MHPSS practitioners in their organisations who have indicated interest in the study. The research team will oversee the recruitment of the practitioners. Recruitment will be open to new participants during the 6-month baseline period and cease once the first term of the supervision program starts. If appropriate (others have dropped out and new practitioners have joined the organisation), new practitioners may join in the break between the two supervision terms when new groups are formed.
Inclusion criteria
Inclusion criteria for the MHPSS practitioners are: (1) 18 years or over; (2) self-identify as Syrian or Bangladeshi; (3) working as an MHPSS practitioner (psychosocial worker, psychologist, psychiatrist, case worker or psychological counsellor) with displaced Syrian (in Northwest Syria or Türkiye) or Rohingya community (in Bangladesh). Rohingya MHPSS practitioners cannot be included in the study due to Bangladesh Telecommunication Regulatory Commission restrictions on internet access for Rohingya living in Cox’s Bazaar refugee camps since 2019 [55].
Inclusion criteria for MHPSS service users are: (1) 18 years or over; and (2) receiving MHPSS services from an MHPSS practitioner recruited in the study. MHPSS service users will be recruited among the beneficiaries of the practitioners involved in the study.
Inclusion criteria for Australian supervisors are: (1) 18 years or older; (2) psychologists, clinical psychologists, social workers or counsellors; (3) completed tertiary training in clinical psychology, social work or counselling or Registered Psychologists.
Inclusion criteria for local supervisors are: (1) Completed a university degree in psychological counselling, psychology or psychiatry; and, (2) to have participated in the pilot supervision program since the beginning of 2020 or have other supervision experience.
Remuneration
Syrian and Bangladeshi supervisors and clinicians will be offered two free online short courses to support their participation in the program. Upon completion, they will receive two accredited certificates and digital badges stating that they have completed two short courses at the University of New South Wales Faculty of Medicine and Health, Sydney, Australia. These two certificates confirm participation in a 16-month supervision program.
For practitioners participating in the supervision program, the five people who answer the highest number of questionnaires closest to the date they are sent out will be awarded $50 for each supervision term.
Measures
Practitioner online surveys
The Kessler-6 [56] ), a 6-item measure of general distress which is sensitive to change during treatment; The Copenhagen Burnout Inventory (CBI) [57] 19 item self-report measure with personal, work-related and client-related burnout sub-scales; The Professional Quality of Life (Stamm, 2005), 30 items assessing clinician compassion satisfaction, compassion fatigue and secondary traumatic stress (symptoms of posttraumatic stress disorder associated with helping populations that have experienced trauma). Counseling Activity Self-Efficacy Scales (CASES) [36] a self-rating scale for counselling clinicians to rate their confidence in providing effective counselling; The PTSD-8 [58], a brief measure of PTS symptoms which has been derived from the Harvard Trauma Questionnaire, along with a list of Traumatic Events (HTQ-TEs) Inventory [59]; modified version of Post-migration Living Difficulties (PMLD-17) Questionnaire [60, 61]. Subjective experiences of supervision were measured with the six-item Perceived Supervision Scale (PSS) [62]. A shortened, six-item version of the Turnover Intention Scale (TIS-6) [63] will measure MHPSS practitioners’ intention to leave their current employment. Nine questions adapted from the Demographic and Health Survey Service Provision Assessment [64], will capture organizational and workforce characteristics.
Beneficiary interviews
Service satisfaction and quality among MHPSS service users will be measured by the Client Satisfaction Questionnaire (CSQ-8) [65] and an 18-item measure developed in this project to evaluate displacement context-specific MHPSS service use experiences.
The data collection plan with an overview of measures for each participant group is given in Table 1 for the active control period and Table 2 for the intervention period.
Procedures
Practitioner online surveys
Each month, field research teams for each site contact all currently eligible practitioners (WhatsApp and email) and provide them with an online survey link containing all planned measures for that cross-section of the overall program. Surveys are delivered using the online KoBoToolbox platform [66]. KoBoToolbox was selected over alternatives (for e.g., REDCap) due to its offline data collection and multilingual support.
Beneficiary interviews
Practitioners who deliver MHPSS services directly to beneficiaries are eligible for enrolment in our beneficiary interview data collection program. Note, not all practitioners enrolled in our supervision program are eligible for beneficiary interviews for a range of circumstances. Agreements with the MHPSS service organisations are required to contact beneficiaries, with some organisations not able to agree to this process, other reasons for not collecting beneficiary data include: practitioner has changed jobs into a non-service delivery role (i.e. line manager/supervisor); fractional unemployment; as well as illness or holiday. For each eligible practitioner, field researchers at each site will attempt to conduct an interview with one of the beneficiaries of their MHPSS services, on a one-to-one basis per interview cycle (see Table Beneficiary data collection for description of interview cycles). Beneficiaries are eligible for an interview up to 21 days after their session with their MHPSS practitioner.
Practitioners are blinded from knowing which of their beneficiaries receive an interview; except in such cases where only a single beneficiary is available for interview per interview cycle. Beneficiaries were randomly selected from among all beneficiaries seen by the practitioner in a given week based on the time at which the session with the practitioner occurred to minimize day-of-week and time-of-day sampling biases [67, 68]. A novel sampling procedure was developed to counter-balance across available time windows (see Supplementary Material 2).
Supervision program participation
The proposed online supervision program for practitioners in will be conducted over 16 months, divided into two 8-month terms, and co-facilitated by an international and a local supervisor. To ensure feasibility and cost-effectiveness in displacement contexts [5], group supervision sessions for 4 to 6 practitioners will be held fortnightly for 90 min on the Zoom platform. In-country research assistants (referred to as “field researchers”) will coordinate meeting invitations and hosting and attend each session. They will also seek consent to record the sessions and remind practitioners two days before their scheduled sessions while monitoring their attendance.
Analytic design
Quantitative data
Hierarchical models will be used to compare practitioner rate of change in reported outcomes between the Active control period and each of the Supervision Terms 1 and 2.
As such, prospective models will consider the Active Control data as a within-subjects control condition that can be jointly estimated across levels of the model (i.e., practitioner, supervision group, site). Cross-sectional data drawn from the beneficiary interviews will be nested within practitioner from the longitudinal data drawn from the online practitioner surveys, cross-classified across time. Supervision program participation data may also be incorporated following qualitative analysis.
To address Objective 1, data from the active control period will be used to model the relationships between psychological hazards and outcomes, using the hierarchical model structure described above. This model will take into account sociodemographic characteristics, organizational factors, group allocation, exposure to the intervention, and other contextual factors that may be identified over the course of the intervention program. To address Objective 2, the same model will be applied to data from the supervision terms to determine whether the introduction of supervision moderates the relationships between psychological hazards and outcomes identified in Objective 1.
Data may be transformed and/or combined in order to achieve appropriate variance partitioning (for e.g., factor analytic techniques, clustering), informed by gold-standard approaches [69]. All candidate variables will be visualized and modelled at the bivariate level prior to final analysis in order to mitigate multicollinearity during model fitting; as such, not all planned variables may be suitable for inclusion in finalized models. This iterative exploratory process means that models cannot be specified in advance of data collection; analyses will therefore be pre-registered, where possible, to ensure best practise [70].
Planned statistical analyses will primarily be carried out in the R language ecosystem [71] within the RStudio IDE [72], however, MPlus [73], STATA [74], and SPSS [75] may also be utilized. Data collection will be conducted primarily using platforms such as KoBoToolbox [66] and Qualtrics [76].
Planned reporting will be performed in accordance with Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [77].
Qualitative data
Video recordings of the supervision sessions will be analysed using content analysis [78] to identify both process (how do things happen in the session) and content (what is being discussed in the session) codes. Content codes will be iteratively devised in collaboration with researchers across all sites to promote the cultural and contextual relevance to codes. Qualitative analysis of the videos will help us to gain insights into the supervision process and examine the proposed mechanisms. Information on the number of supervision sessions attended, logistical barriers (e.g., connection issues), and session structure will also be extracted from the video recordings.
Thematic analysis [79, 80] will be conducted on a subset of supervision videos from the beginning, middle, and end of the supervision program to elucidate proposed mechanisms (supervisor practices; group processes; barriers and facilitators to participation). Further, at the start and end of each supervision term, semi-structured interviews will be conducted with the practitioners to gain insights into the impact, acceptability, and appropriateness of the supervision program. NVivo 12 software will be used to aid qualitative data analysis [81].