Background
At the end of 2021, almost 90 million people were forcibly displaced; 27 million refugees, and more than 40% children [
1]. Throughout 2022, this figure has increased to more than 100 million [
1]. Armed conflict is a major driver of displacement and has a huge impact on mental health and wellbeing of populations. It is estimated that one in five individuals in conflict-affected areas will have a diagnosable mental disorder, with the burden of disease of common mental disorders more than five times the global average [
2]. Children and young people frequently experience additional stressors and barriers to healthy development before, during and after conflict, including displacement, poverty, education interruptions, increased family and community violence, and child protection risks [
3], placing them at greater risk of mental health problems [
4].
The majority (83%) of refugees are hosted in Low and Middle Income Countries (LMICs) [
1], where host communities face similar social determinants of poor mental health, and health systems are drastically under-resourced to deal with the increased demands [
5]. Therefore, in these settings, a vast treatment gap exists, whereby the majority of individuals needing mental health treatment do not receive minimally adequate care [
6,
7]. This is especially the case for young people where there are significant system-level barriers to providing necessary support, including scarcity of mental health professionals [
8] and inadequate funding for mental health, especially for children and adolescents [
9,
10].
To overcome this treatment gap, effort has shifted towards scalable psychological interventions that can be delivered by trained and supervised non-specialist providers, following a task-sharing approach [
6]. International guidelines [
11] recommend delivering focused, non-specialist interventions for individuals experiencing distress, reserving specialist care for more severe and complex cases. Several such interventions for adults have been developed and evaluated across different settings with good effectiveness and feasibility, including the World Health Organization’s (WHO) (Group-) Problem Management Plus (PM +) [
12‐
15]. Yet similar evidence-based brief non-specialist interventions specifically for children and adolescents remain scarce.
Although WHO guidelines recommend psychological interventions such as cognitive behavioural therapy for adolescents with emotional disorders [
16], available evidence for effectiveness of interventions with children and adolescents in humanitarian settings is mixed, with a strong focus on treatments for posttraumatic stress disorder (PTSD) [
17]. A recent individual patient data meta-analysis found significant effects of focused psychosocial interventions in reducing symptoms of PTSD, and improving functioning and strengths including coping, hope, and social support [
18]. However, no significant effects were seen for reduction of depression and anxiety, and effects on other outcomes were stronger in older adolescents (15–18 years old). Similarly, an umbrella review of systematic reviews found only suggestive evidence of effectiveness for psychosocial interventions for children in humanitarian settings in reducing PTSD, and for general child populations in reducing disruptive behavior, with weak or no evidence for other outcomes [
19].
WHO recently developed the Early Adolescent Skills for Emotions (EASE) intervention to reduce psychological distress for adolescents aged 10–14 years who are living in adversity and experiencing significant internalizing symptoms (e.g. anxiety or depression) [
20]. EASE is designed to be applicable across common mental disorders, brief, transdiagnostic, and deliverable by non-specialists, reducing the need for diagnostic specialist services which may not be available. It consists of manualized evidence-based techniques found to be most commonly included in effective psychological interventions for this age group and deemed to be safe and feasible for delivery by trained non-specialists.
We previously conducted a cultural adaptation of the EASE intervention for the context of Lebanon [
21]. Following recommendations for developing complex interventions [
22], we first conducted this randomized feasibility trial to inform necessary adaptations prior to the implementation of a fully powered RCT to assess effectiveness. Specifically, the aims were to:
-
1) Determine recruitment, screening, completion, and retention rates for the EASE programme and follow up assessments
-
2) Evaluate the feasibility and acceptability of intervention delivery of EASE by trained non-specialists
-
3) Assess feasibility of outcome measures and their psychometric properties, and explore trends in changes from baseline to endline within each group, and between groups
-
4) Assess feasibility and safety of trial procedures such as randomization, blinding of assessors, contamination, and occurrence and monitoring of adverse events
This feasibility trial and a simultaneous study in Jordan [
23] were conducted in preparation for the first full-scale evaluations of EASE in Lebanon and Jordan [
24,
25]. Additional studies have been conducted in Tanzania [
26] and Pakistan [
27].
Discussion
The overall aim of this randomized feasibility study of EASE in Lebanon was to inform adaptations needed to intervention and research protocols for a fully-powered RCT to assess effectiveness. We noted challenges with outreach and saw significant attrition between registration of interest and screening time-points, and between screening and baseline for those eligible. There were challenges among the EASE participants in terms of regular attendance at sessions, with only 63% of adolescents completing five or more EASE sessions; common reasons cited were school and work commitments, or transport issues. Nonetheless we found good retention rates among adolescents and caregivers in completing endline and follow-up assessments. Randomisation and blinding procedures were effective and feasible, and we noted minimal contamination. Fidelity and competency of EASE implementation was satisfactory, though time management and behaviour management were issues in some sessions. Outcome measures showed good psychometric properties on total scale scores, with low item-level missing data for most questionnaires. Our screening measure resulted in a positive screening rate of 49%, while baseline levels of distress on outcome measures were relatively low. We did not identify substantial patterns of change overtime on our outcome measures, (between or within groups) despite qualitative reports of positive impacts of EASE. We responded to a high number of adverse events throughout the study, necessitating referrals to other services, though none were deemed to be caused by or linked to participation in the study.
Findings of this feasibility trial resulted in several improvements to study procedures to inform a forthcoming larger definitive trial [
24]. Firstly, recommendations from this feasibility study (along with concurrent studies conducted in Jordan [
23] and Tanzania [
26]), lead WHO developers to make the following modifications to the EASE intervention: (i) more interactive activities were added to adolescent sessions, (ii) the storybook used in adolescent sessions was re-written by a creative writer to enhance engagement, (iii) caregiver sessions were shortened and revised to reduce the need for literacy. For the Lebanon context specifically, guidance was added to hold the first caregiver session prior to the first adolescent session, to allow a clearer explanation to caregivers about the content of the adolescent sessions including confidentiality expectations, and the extent to which ‘taboo’ topics such as suicide and traumatic events will be discussed. Furthermore, additional facilitator training was provided on behaviour management, engaging the group, and showing empathy for challenges raised, and avoiding giving advice. Supervision sessions were adapted to be more active and skills based.
RCT procedures were adapted to overcome transportation challenges: (i) group sessions were held as close as possible to homes; (ii) caregivers were provided with the choice between provision of transportation via bus, or reimbursement of travel costs; (iii) outreach teams improved communication with caregivers before and during adolescent transportation; (iv) improvements were made to the contamination tool and assessor training in order to improve reliability of this measurement. We added an explicit check at screening that both adolescents and caregivers were willing to commit to attend to all EASE sessions should they be allocated to that condition. Recommendations were made to enable flexible scheduling of sessions to include evenings and weekends, and have facilitators remain in contact with caregivers over WhatsApp. Given the intense outreach efforts required to reach our pilot sample, we employed additional outreach staff for the RCT, and provided substantive training on methods for explaining the intervention, responding to concerns, and generating interest. To respond to the high rates of child protection and acute mental health needs experienced in the feasibility study, we hired additional staff members to support on these aspects during the full trial to ensure safety and adequate referral and follow up. The suitability of these adjustments will be monitored in the full trial.
While most outcome measures were deemed understandable and appropriate, we adjusted the wording of some items that were found to be challenging and removed non-endorsed traumatic events from the trauma inventory. Notably, the CPDS screening tool lead to a sample with relatively low levels of distress overall on outcome measures, despite a prior validation study (Brown F, Steen F, Taha K, Koppenol-Gonzalez GV, Aoun M, Bryant RA, et al: Validation of Arabic versions of the Child Psychosocial Distress Screener and Pediatric Symptom Checklist for young adolescents living in vulnerable communities in Lebanon, Under review). One possibility for the lack of change over time in our study is a floor effect, possibly due to the fact that the CPDS assesses presence of psychosocial risk factors, rather than psychological distress. Based on this, we changed our screening tool for the RCT to the PSC-17 tool. While the increased length adds challenges for population-level screening, it has shown good internal consistency, test–retest reliability, and concurrent validity against a gold-standard psychiatrist assessment in Lebanon (Brown F, Steen F, Taha K, Koppenol-Gonzalez GV, Aoun M, Bryant RA, et al: Validation of Arabic versions of the Child Psychosocial Distress Screener and Pediatric Symptom Checklist for young adolescents living in vulnerable communities in Lebanon, Under review). Given the poor psychometric properties of PSC and CRIES subscales in this sample, our primary hypotheses for the full trial will focus on total scale scores.
We expect that these amendments will lead to improved implementation of the EASE intervention and trial procedures for the RCT. A simultaneous feasibility trial of EASE in Jordan found similar lack of movement on outcomes, and similar possible floor effects [
23]. This sample was similar in terms of past traumatic events experienced, although there were no adverse events reported throughout this study, more adolescents were currently attending school, and less were working to provide income for the family. Better attendance was observed, which might be explained by less disruptive circumstances. Another study of a non-specialist psychological intervention in Lebanon conducted during the same period demonstrated similar difficulties with engagement [
42], suggesting that pervasive stressors and community dynamics in Lebanon may influence demand-side factors in delivering psychological care.
Although we have made significant adjustments to our implementation plan for the RCT to promote smooth implementation, these adjustments predominantly involve mobilising additional resources in terms of outreach staff and activities, support for following up with significant child protection and mental health needs, and enhanced transport options to sessions. While these resources may be needed in order to ensure adequate and comprehensive care is available for those who need it, it is important to note for future implementation of such interventions in protracted crises that these interventions are unlikely to be viable as standalone programmes, and must be integrated within existing services and support structures in order for them to be sustainable and effective. EASE may be optimally positioned within a multi-level and multi-sectoral care system whereby adolescents may be engaged in community level psychosocial support activities and educational activities, through which identification and referral to EASE can be made, and through which direct referrals can be available to more intensive case management for protection concerns, or specialised services for more severe mental health concerns, as required [
43]. In addition, it will be important that attention also turns to community engagement activities, including mechanisms to sensitise communities to mental health activities, and reduce stigma around accessing such interventions.
While the group format of EASE has advantages in terms of increasing numbers of adolescents reached, and promoting social support, the scheduling of group session times and coordinating common locations reachable by adolescents across a fairly broad geographical area presented significant challenges for attendance, and necessitated significant transport costs. Future research should compare the effectiveness, overall cost-effectiveness, and sustainability of individual versus group treatment models of EASE.
Limitations
The small sample size, though intentional, prevents us from conducting meaningful subgroup analyses to understand whether EASE may show trends in change overtime for certain demographic groups, or in situations of higher adolescent and caregiver participation in the intervention. This will be explored in a forthcoming fully-powered RCT. This feasibility trial was registered retrospectively which limits the ability to guarantee non-selective outcome reporting. The full RCT has been registered prospectively. Furthermore, the applicability and impact of the EASE intervention may have been under-estimated due to a sample that overall was not scoring high on distress at baseline. Intervention costs and health economic analyses are important in definitive trials to estimate cost-effectiveness of interventions, and feasibility of collecting relevant data could have been piloted in this study. Lastly, while we conducted follow up interviews with drop-outs from the intervention to understand the reasons, relatively little is known about the perceptions and barriers for engagement of those community members who we approached but declined to participate. Understanding these experiences would be helpful to inform outreach and engagement activities and inform future service delivery.
Acknowledgements
We firstly express our gratitude to all adolescents and caregivers who dedicated their time to this research and shared their experiences with us. We would also like to thank the EASE and ETAU facilitators (Rana Wehbe, Fatima Ali Anka, Imane Adnan El Tabaa, Samia Khodor Seif, Moemen Mohamad Cherif Shanan, Wisal Mohamad Yaser, Alaa El Jarrach, Hiba Hindawi, Moujahed Ahmad El Mouzawek, and Ranine Wajih El Omar) who worked in challenging conditions to support adolescents and their caregivers and provided invaluable feedback to improve implementation. Thank you to Reem Masri and Hasan Mortada who co-lead the facilitator training and provided supervision and conducted observations; outreach workers Walid Najib, Bassel Meksassi, and Nour El Cheikh who supported the recruitment and coordination of the sessions; Abeer Hudery and the rest of the child protection team for following up on cases needing support and referral; and the research assistant team who conducted all assessments with families (Ahmad Al-Hajj, Fadi Matar, Grace Harb, Mohammad Safsouf, Hala Rashid Dalal, Asma Ghassan Kammoun, Mohamad Khalaf, Nadine El Homsi, Reve Assaad Abou Rjeily, Mathiew Harran, Bilal El Asaad, Sasha Hilani, Hazar Khodor, Khaled Chkhaidem, Mireille Chamoun, Hana Chrayteh, Hilane El Zarif, Lamia Allaoui, Mostafa Amin Al Sabin). We appreciate the important and careful oversight of the safety of the study by our Data Safety Management Committee: Nayla Nahas, Samar Zebian, and Frank Elbers. Lastly, we acknowledge the support from the full team at War Child Holland in Lebanon for making this study possible.
EASE Intervention development & training team
The EASE Intervention Development and Training Team consists of Katie S Dawson7, Sarah Watts6, Maha Ghatasheh12, May Aoun3, Aiysha Malik6, and Felicity L Brown1,2
STRENGTHS consortium
The STRENGTHS Consortium members in alphabetical order are: Ceren Acarturk13, Aemal Akhtar7, Ibrahim Akinçi14, Ahmed Bawaneh15, Martha Bird16, Felicity L Brown1,2, Richard Bryant7, Sebastian Burchert17, Pim Cuijpers8, Anne de Graaff8, Annelieke Drogendijk18, Daniela Fuhr19, Jonas Maria Hessling17, Zeynep Ilkkursun13, Mark J.D. Jordans1,2, Christine Knaevelsrud17, Gülşah Kurt13, David McDaid20, Saara Martinmäki18, Cansu Mirzanlı14, Trudy Mooren18, Naser Morina21, A-La Park20, Monique Pfaltz22, Bayard Roberts19, Matthis Schick21, Ulrich Schnyder21, Marit Sijbrandij8, Egbert Sondorp23, Julia Spaaij21, Frederik Steen1, Karine Taha3, Peter Ventevogel24, Claire Whitney15, Nana Wiedemann16, Aniek Woodward23
1Research and Development Department, War Child Holland, Amsterdam, The Netherlands
2Amsterdam Institute of Social Science Research, University of Amsterdam, Amsterdam, the Netherlands
3War Child Lebanon, Beirut, Lebanon
6Department of Mental Health and Substance Abuse, World Health Organization, Geneva, Switzerland
7School of Psychology, University of New South Wales, Sydney, NSW, Australia
8Clinical, Neuro and Developmental Psychology, VU University, Amsterdam, The Netherlands
12Institute for Family Health, Amman, Jordan
13Koc University, Istanbul, Turkey
14Refugees and Asylum Seekers Assistance and Solidarity Association, Istanbul, Turkey
15International Medical Corps, Amman, Jordan
16International Federation of Red Cross and Red Crescent Societies, Copenhagen, Denmark
17Freie Universität Berlin, Berlin, Germany
18ARQ International, Amsterdam, The Netherlands
19London School of Hygiene and Tropical Medicine, London, UK
20London School of Economics and Political Science, London, UK
21University of Zurich, Zurich, Switzerland
22Mid Sweden University, Sundsvall, Sweden
23KIT Royal Tropical Institute, Amsterdam, The Netherlands
24United Nations High Commissioner for Refugees, Geneva, Switzerland