We conducted a 2-arm single-blind randomized controlled trial with refugees from Afghanistan living in Indonesia. Indonesia hosts approximately 12,000 refugees, with approximately half of these living in urban settings in the greater Jakarta area [21]. While all refugees ostensibly live in Indonesia temporarily as they await resettlement to other countries, in actuality, this waiting period is often indefinite due to limited resettlement places [22]. During this protracted displacement, refugees in Indonesia do not have the right to work or to reunite with their families. They have little access to education and government-sponsored services, with the limited support available to refugees being provided by nongovernment and refugee-led community-based organizations [23]. Approximately half of the refugees in Indonesia originate from Afghanistan [18]. Refugees from Afghanistan have experienced decades of armed conflict, political instability, and human rights abuses, with a renewed wave of mass displacement occurring following the Taliban’s return to power in 2021, which sharply curtailed civil liberties, education, and employment [24]. The humanitarian crisis in Afghanistan is considered one of the most complex crises of our time, given the lack of resolution after more than 2 decades of international intervention. Many Afghan refugees have also been exposed to prolonged premigratory trauma and perilous migration journeys, including violence, loss, and extended periods of insecurity in transit or displacement settings, contributing to high levels of psychological distress on arrival in host countries [25].

We elected to undertake this study in a digital format (ie, recruitment via social media and assessment and delivery of DWM materials via Zoom [Zoom Video Communications] calls) to increase access to the intervention in a population that was geographically dispersed around the greater Jakarta area. This methodology was considered feasible given (1) widespread internet access in Indonesia [26], (2) the successful implementation of digital procedures in our previous longitudinal study of over 1300 refugees living in Indonesia [19], and (3) reports from our study partners that smartphone use was ubiquitous among refugees in Indonesia, consistent with evidence that refugees commonly use digital methods to stay in contact with family in the country of origin or displaced in other settings [26,27]. In recognition of the cost of accessing reliable internet in Indonesia, participants in both conditions were provided with an internet allowance.

The trial protocol (Multimedia Appendix 1) was prospectively registered at the Australian and New Zealand Clinical Trials Registry (ACTRN12624000609550) and approved by the UNSW Sydney Human Research Ethics Committee in Australia (iRECS4937) and the Badan Riset dan Inovasi Nasional in Indonesia (25032024000003 and 07052025008596). This superiority trial was conducted in collaboration with the Universitas Gadjah Mada, the Indonesian Civil Society Association for Refugee Rights Protection (SUAKA), and Refugee Learning Nest. The CONSORT-eHealth (Consolidated Standards of Reporting Trials of Electronic and Mobile Health Applications and Online Telehealth) checklist is available in Checklist 1. Participants provided informed consent and were able to opt out of the study at any time without penalty. After each assessment, participants received a US $6 digital grocery voucher as compensation and an internet allowance. Participants in the RAC group were offered access to the self-help version of the DWM intervention after completing the assessments, ensuring that no participants were denied the opportunity to receive the intervention. This study complied with informed consent guidelines and adhered to local, national, regional, and international law and regulations regarding protection of personal information, privacy, and human rights

Participants were recruited online through social media, advertisements on the study website, and flyers distributed to organizations supporting refugees in the Greater Jakarta area in Indonesia (Jakarta, Bogor, Depok, Tangerang, and Bekasi). Interested participants completed an online registration form that included study information, consent to a screening call with a trained assessor, the Kessler Psychological Distress Scale (K10) [28], an initial eligibility question confirming refugee status in Indonesia, and demographic questions. Those who consented and met the initial eligibility criteria were contacted by a trained assessor for an online screening call. During this call, information was collected on access to a digital device and internet connection, current treatment, suicide risk, and severe cognitive impairment. The K10 was also readministered to verify self-reported responses. Participants who met the full eligibility criteria proceeded to the baseline assessment prior to randomization. All assessments were conducted online by trained assessors.

The inclusion criteria were (1) being aged 18 years or older, (2) being from a refugee background (ie, living in Indonesia and having registered with the United Nations High Commissioner for Refugees (UNHCR) or intending to register with the UNHCR), (3) residing in Greater Jakarta area (Jakarta, Bogor, Depok, Tangerang, or Bekasi), (4) ability to speak and read Farsi, (5) elevated psychological distress (≥20 on K10 [28]), and (6) providing written or verbal informed consent before entering the study. The Farsi language was selected for this study, as it represented a common language for Afghan refugees from Dari, Hazaragi, and Pashto-speaking backgrounds. Exclusion criteria included (1) acute medical condition, (2) imminent suicide risk or expressed acute needs or protection risk (ie, being at acute risk of domestic and/or sexual violence), (3) severe mental disorder (psychotic disorder or substance dependence), (4) severe cognitive impairment (eg, severe intellectual disability or dementia), (5) concurrent psychological treatment (eg, concurrently receiving psychological services from a psychologist, counselor, or mental health professional), (6) no access to smartphone or internet connection, and (7) having another household member who had already applied to take part in the study. Those excluded based on suicide risk, severe mental disorder, or severe cognitive impairment were referred to the available general or specialist services.

Eligible participants were randomly assigned to DWM or the RAC condition at a 2:1 ratio. This ratio was implemented to allow for the investigation of mechanisms of action within the DWM condition (which will be detailed in future reports). Randomization was conducted by a researcher who was not involved in the delivery of the intervention. Randomization was performed using a computerized software (REDCap [Research Electronic Data Capture]; Vanderbilt University) [29,30]. Assessors were masked to treatment condition allocation. Assessors were managed separately from other members of the research team and did not interact with the DWM facilitators. Participants were instructed not to share their condition allocation with the assessors to ensure objectivity of the assessments.

Based on previous studies with Self-Help Plus (the group-based version of DWM) [13,15], we hypothesized a 0.40 between-groups effect size on our primary outcome (K10) and calculated that a minimum of 73 participants per condition would be required to achieve 80% power for α=.05. Given that we were also interested in investigating mechanisms within the DWM condition, we required a larger sample size in this group (n=146). Estimating at 350% attrition rate due to the mobility of the refugee population in Indonesia, a final sample size of 303 participants (DWM: n=202 and RAC: n=101) was planned.

For our primary analysis, we investigated the impact of DWM on the total K10 score at each time point in the intention-to-treat sample. We used a linear mixed models approach to estimate the between-groups effect of DWM versus RAC at mid-treatment, posttreatment, and follow-up. This model had time as a fixed effect, covaried for baseline K10 score, and included participant as a random effect. We constrained the fixed effect of the intervention to be 0 and created dummy variables for time at mid-treatment, posttreatment, and follow-up (with pretreatment being the reference group). We then included dummy-coded time×intervention interaction terms in the model for each time point. This yielded estimates of the average between-group differences in K10 score for DWM versus RAC at each time point, controlling for baseline scores on K10. Robust inference for fixed effects was obtained using cluster-robust (sandwich) SEs with the small-sample correction, implemented via the clubSandwich package in R (R Foundation for Statistical Computing), with clustering at the participant level. Statistical tests and P values were based on Satterthwaite-type degrees of freedom derived from the CR2 variance estimator. We next tested whether the same pattern of results was found when adjusting for covariates (age, gender, count of traumatic experiences, and time spent in Indonesia), in the per-protocol population, among follow-up completers, and for secondary outcomes using the same approach described earlier. We calculated effect sizes (Cohen d) by dividing the between-groups effect by the pooled baseline SD, adjusting for baseline scores on the outcome variable. We calculated Reliable Change Index scores for the K10 following the procedure outlined by Jacobson and Traux [47] to examine whether changes from baseline to mid-treatment, posttreatment, and 1-month follow-up assessments were reliable and clinically meaningful in addition to being statistically significant. All statistical analyses were conducted in R Studio (version 4.4.3; Posit Software, PBC).

We found that DWM was effective in reducing psychological distress and PTSD symptoms, as well as improving well-being and overall daily functioning, in refugees living in a transit country. To our knowledge, this is the first fully powered randomized controlled trial to investigate the efficacy of DWM with refugees, demonstrating both the efficacy and feasibility of an individually guided self-help approach in improving psychological well-being in a context characterized by significant ongoing stressors, prolonged uncertainty, and limited resources. The clinical significance of these results is reflected in the finding that, compared to the repeated assessment group, participants who received DWM showed significantly greater rates of recovery (45/202, 26% vs 9/101, 10%) and improvement (50/202, 28% vs 18/101, 20%) at the follow-up assessment as measured by the Reliable Change Index. A similar pattern of findings was also observed at posttreatment. This study builds on previous research, demonstrating that DWM’s group-based alternative (Self-Help Plus) improved mental health outcomes in Sudanese refugees in Uganda [15], and among refugees without a psychiatric diagnosis in Western Europe [14], and in Türkiye [13]. It also extends on preliminary results, indicating that DWM reduced PTSD symptoms in Syrian refugees [12].

Our findings provide evidence that DWM may be well-suited for refugees subject to prolonged displacement in LMICs. Importantly, guided self-help interventions require relatively few resources to implement, as participants are afforded an illustrated workbook that is complemented by audio files and facilitator support to aid the learning and practice of the intervention strategies [10]. While this approach has previously been found to be efficacious when delivered in a group-based setting [13-15], in this study, participants were provided with individual-level support via five 30-minute Zoom calls from lay facilitators from the refugee community. There is evidence that individually supported approaches may be more effective than group-based interventions [16]. In addition, this one-to-one support mechanism overcomes key barriers to implementation and engagement for refugees in transit settings such as mental health stigma and reluctance to engage in group-based interventions, especially among survivors of persecution and human rights violations [48]. Individual assistance can also overcome logistical challenges, including geographical dispersal of refugee communities, scheduling difficulties, and challenges with assembling and retaining groups of adequate size. In contrast to group-based approaches, individualized support is conducted at a time that is convenient for both the facilitator and the individual, is readily rescheduled, and can be tailored to address the unique psychological needs and practical circumstances of the individual [17]. While the intervention was delivered online in this study (which can increase privacy, reduce stigma [49], and may be well-suited to refugee populations who regularly use smartphones to remain connected to family and friends in other locations [26,27]), it could also be readily delivered in a face-to-face modality, enhancing its generalizability across settings.

The content of the DWM intervention may also be especially appropriate for refugees living with high levels of insecurity in transit settings. In the context of a myriad of uncontrollable stressors relating to refugee status determination, financial insecurity, lack of access to services, and insecure housing in transit countries [50], concrete problem-solving approaches may be difficult to implement. In contrast, the acceptance and commitment-based strategies implemented in DWM, such as “unhooking” from negative thoughts and behaving in a way that is consistent with one’s values in the face of stress [10,31], may be more actionable and provide refugees with tools to cope in contexts where external stressors are intractable and enduring.

It is notable that, in this study, participants in the control condition also showed significant improvements in most study outcomes, with 30% (27/101) reporting clinically significant improvement or recovery at the 1-month follow-up assessment (compared to 95/202, 50% of the DWM condition). This group received matched assessments (4 in total) to the DWM condition from assessors who were trained in empathic listening and basic help-seeking skills. This supportive contact, combined with the expectancy of receiving the DWM materials after the follow-up assessment, may have contributed to symptom improvement in the control group [33]. This finding highlights the potential importance of connection and emotional support in fostering refugees’ well-being. Further, it is interesting to note that the majority of participants who took part in this study were men. While this represents the broader gender balance of refugees in Indonesia, it is contrary to general trends of lower help-seeking among male refugees [51-53]. One possibility is that the online modality may have facilitated greater accessibility and engagement of the intervention among men; alternatively, the economic benefits arising from the vouchers in this study may have encouraged men to participate. Future research should examine patterns of engagement and the benefits of DWM with subgroups of refugees.

Strengths of this study included excellent treatment adherence (99%) and retention to follow-up (88%), nonspecialist community members acting as study facilitators and providing high-quality intervention delivery, and blind and standardized assessments by trained assessors.

We note a number of limitations. First, the study relied on a relatively short follow-up assessment period (1 month); future research testing the efficacy of DWM should evaluate its efficacy in the longer term. Second, the study lacked an active control condition, and as a result, we cannot exclude the possible role of nonspecific factors such as facilitator support, contributing to the effects of DWM. Third, we note that a number of the assessment measures have not been validated with the cultural groups who participated in the study. Finally, digital access requirements may have introduced selection bias.

Amid global calls to better integrate refugee care into the health systems of host transit countries [3], DWM and other guided self-help approaches may form an effective component of stepped-care approaches that combine evidence-based scalable interventions with specialized psychological treatments that are embedded in broader systems of care [11]. Low-intensity approaches that can be effectively delivered by nonprofessionals may be especially beneficial in transit countries like Indonesia, where the gap in access to formal support services is often bridged by refugee-led organizations [23]. Empowering these groups to support community members through guided self-help interventions like DWM has the potential to reduce the substantial mental health burden of refugees living in transit settings. Alongside advocacy and policy-focused efforts to facilitate resettlement opportunities and address structural barriers to well-being in transit contexts, such community-supported mental health initiatives represent scalable solutions that can be sustainably implemented by refugee communities in the context of limited resources.

This study provides evidence for DWM as an effective approach to reducing psychological distress that can be feasibly supported by lay individuals from refugee communities. The scaling-up of this and other low-cost guided self-help interventions represents an important pathway to increasing access to mental health care for the millions of refugees worldwide.