Exposure to natural disasters increases the risk of mental health problems in children and adolescents, including post-traumatic stress disorder (PTSD), depression, and anxiety (Wang et al., 2013). While not all individuals exposed to a natural disaster will experience PTSD or mental health concerns, addressing symptoms is important to reduce related distress and impairments such as disruptions to education and daily functioning (Hiller et al., 2016).
Schools are key to supporting the health and wellbeing of students, their families and the wider school community following natural disasters. As concrete (figuratively and often literally) facilities with existing infrastructure, schools can serve as evacuation or medical sites and be places to disseminate information. To keep students physically safe, schools are increasingly expected to have disaster preparedness plans, especially in countries that are particularly vulnerable to climate change. However, evidence-based guidance on how schools can support students’ mental health following a natural disaster is less established.
To address this, Laksmita and colleagues (2025) conducted a systematic review and meta-analysis synthesising research on the efficacy of randomised controlled trials (RCTs) of school-based interventions for children and adolescents in post-natural disaster settings globally.
Schools can be evacuation or medical sites following a natural disaster, but high-quality evidence is needed for how best to support students’ mental health.
Methods
The authors searched multiple electronic databases and reference lists of included studies to identify English language articles of school-based RCTs with a control group of any kind (i.e., no intervention, usual care, waitlist, or alternative intervention), with participants aged 3-18 years.
The primary outcome was PTSD symptoms; secondary outcomes were depression and anxiety. Effect sizes for pooled estimates were calculated using Hedges’ g (0.2 = small, 0.5 = medium, 0.8 = large effect) with 95% confidence intervals (CIs; Cohen, 1988).
Risk of bias (RoB) was assessed by two researchers, with strong agreement between raters (κ = 0.88). Most studies were rated as high RoB, with moderation analyses finding that studies with a high RoB yielded a large, statistically significant effect, whereas those with fewer concerns yielded a smaller, non-significant effect. Publication bias and heterogeneity were also assessed.
Results
Study characteristics
Fifteen studies were included in the systematic review and 13 in the meta-analysis. These studies, published between 1999 and 2024, involved 2,418 participants aged 6-16, and used active (n = 7), passive (n = 4), and waitlist (n = 2) control groups.
Most school-based interventions were non-Cognitive Behavioural Therapy (CBT) approaches such as drama or music therapy, Eye Movement Desensitization and Reprocessing (EMDR), and psychosocial support. Interventions were delivered by teachers, mental health professionals (e.g., psychologists), and other trained therapists (e.g., music therapists).
Primary outcome
Compared with control conditions, school-based interventions demonstrated statistically significant reductions in PTSD symptoms at all three timepoints (see Table 1). Effect sizes were large immediately post-intervention (g = -1.203, 95% CI [-2.202 to -0.203], p < .001), small-to-medium at short-term follow-up (g = -0.252, 95% CI [-0.479 to -0.026], p = .029), and medium at long-term follow-up (g = -0.450, 95% CI [-0.816 to -0.084], p = .016), indicating that the effectiveness of the interventions on PTSD symptoms persisted over time.
Secondary outcomes
Compared with control conditions, school-based interventions demonstrated statistically significant reductions in depressive symptoms immediately post-intervention (g = -0.243, 95% CI = [-0.433 to -0.054], p = .049) and at short-term follow-up (g = -0.234, 95% CI = [-0.417 to -0.050], p = .013), but not long-term follow-up (g = -0.331, 95% CI = [-0.784 to 0.122], p = .152), with small effects, suggesting that the effectiveness of the interventions on depression diminished over time.
Statistically significant reductions in anxiety symptoms were also found immediately post-intervention with a large effect (g = -4.602, 95% CI = [-8.807 to -0.396], p = .032; see Table 1).
Heterogeneity
Heterogeneity was variable across outcomes and timepoints. Generally, where effect sizes were large, heterogeneity was high.
Moderators
Passive control groups (e.g., waitlist, no treatment) and per-protocol analyses (i.e., data analysed only for those who completed the interventions) were associated with large, statistically significant effects immediately post-intervention for PTSD symptoms, whereas active control groups and intention-to-treat analysis (i.e., data analysed for everyone allocated to an intervention) showed small, non-significant effects. A similar pattern was seen for depression.
For PTSD symptoms immediately post-intervention: CBT-based interventions produced small-to-moderate statistically significant effects, whereas non-CBT interventions produced larger but non-significant effects; and interventions delivered by non-health professionals (large effect) and health professionals (small effect) yielded statistically significant effects. Shorter intervention sessions (≤60 minutes) delivered more frequently (≥6 sessions) and over a longer period (≥6 weeks) were more effective than longer sessions, delivered less frequently over a shorter time period. For anxiety symptoms, non-health professionals yielded a very large, significant effect whereas health professionals did not.
Studies from developing countries showed larger, statistically significant effects on PTSD and anxiety symptoms immediately post-intervention, whereas those from developed countries showed small, non-significant effects.
Publication bias
No publication bias was detected. Studies published after 2011 reported large, statistically significant effects for PTSD and anxiety symptoms immediately post-intervention while earlier studies showed small, non-significant effects, potentially reflective of improved interventions or research methods.
Table 1. Summary of primary and secondary outcome results
| Outcome | Number of effect sizes for meta-analysis | Statistically significant | Effect size | Heterogeneity |
| Primary outcome – PTSD symptoms | ||||
| Immediate effect | 13 | Yes | Large | High |
| Short-term effect | 9 | Yes | Small-to-medium | Low-to-moderate |
| Long-term effect | 2 | Yes | Medium | Low |
| Secondary outcome – Depressive symptoms | ||||
| Immediate effect | 8 | Yes | Small | Moderate |
| Short-term effect | 6 | Yes | Small | Low |
| Long-term effect | 2 | No | Small | Low-to-moderate |
| Secondary outcome – Anxiety symptoms | ||||
| Immediate effect | 4 | Yes | Large | High |
| Short-term effect | 1 | N/A* | N/A* | N/A* |
| Long-term effect | 0 | N/A* | N/A* | N/A* |
| *Insufficient data available | ||||
School-based interventions were delivered by a range of professionals such as teachers, psychologists, and social workers, and resulted in reductions to PTSD, depressive and anxiety symptoms immediately post-intervention.
Conclusions
The authors concluded that the,
study provides robust evidence that school-based interventions are effective in reducing PTSD symptoms and, to a lesser extent, depression and anxiety in children and adolescents following natural disasters.
However, there was relatively high heterogeneity among immediate effects, and there is limited data on anxiety outcomes and long-term effects (≥6 months). Further factors influencing intervention outcomes included methodological (e.g., how the data is analysed), intervention (e.g., who delivers the intervention), and contextual characteristics (e.g., country). These findings suggest the need for rigorously designed RCTs that evaluate implementation as well as effectiveness across high-, middle-, and low-income countries.
Identifying effective systems-level interventions in addition to individual-level interventions is an important next step in protecting students’ mental and physical health following natural disasters.
Strengths and limitations
This study was a well-conducted review of evidence from RCTs on an important and timely topic. The review was comprehensive, including several relevant databases, grey literature, and no date restrictions on the search strategy which increases the reliability of the findings, as relevant studies were likely caught and included. A key strength was the inclusion of three post-intervention timepoints which allowed for an examination of effects over time; this provides helpful practical information when considering implementation. In addition, several moderator analyses enabled an examination of how effects on PTSD symptoms, depression and anxiety differ according to context and different methodological and intervention approaches, providing us with more information about the conditions surrounding these effects.
However, as the authors noted, there was high heterogeneity between the studies (e.g., type and delivery of interventions), limiting the reliability and generalisability of findings. Further, reported effects might be inflated given that the type of control group (i.e., passive control only/absence of a competing treatment) and analyses (i.e., individuals who did not complete the interventions were excluded from analyses) moderated intervention effects.
Laksmita et al.’s review has many strengths, such as including data to examine whether intervention effects persisted or diminish over time. However, high heterogeneity limits the reliability of these comparisons.
Implications for practice
While the findings from this review are consistent with global polices which conceptualise schools as places that can support education and health simultaneously, they do not support changes to practice or policy yet, as more high-quality evidence is required.
This review would have been strengthened by considering whether interventions were universal or targeted, and delivery timing following the natural disaster. For example, for children and adolescents exposed to trauma, the Australian Guidelines (Phoenix Australia, 2021) recommend universal interventions delivered with the first three months after exposure, provide information, emotional support and practical assistance, rather than individual psychological debriefing. In contrast, for those experiencing symptoms of PTSD three months after trauma exposure, there are strong recommendations for trauma-focused CBT for children and adolescents (and their caregivers) and conditional recommendations for EMDR “where trauma-focussed CBT is unavailable or unacceptable” (Phoenix Australia, 2021). It is perhaps unsurprising, therefore, that the review found school-based CBT interventions to be more effective reducing PTSD symptoms than non-CBT interventions. Understanding how clinical guidelines align with evidence for school-based interventions for PTSD and other mental health outcomes will be important for designing stepped-care approaches within school settings, ensuring interventions are delivered to the right people at the right time, and critically, do not cause inadvertent harm.
The findings suggest that school-based interventions delivered in “developing” countries might be more effective immediately post-intervention due to “higher baseline need or greater responsiveness” (Phoenix Australia, 2021). Given that targeted CBT interventions are typically resource intensive (e.g., delivered to an individual or small group and over multiple sessions), the finding that they can be effectively delivered by non-health professionals is important, especially for low-and-middle-income countries where access to mental health professionals is often limited.
What do these findings mean in context?
In her current role as a Senior Health Program Officer of the Philippine Department of Health where she is responsible for the implementation of a whole-school systems approach to health and wellbeing (Healthy Learning Institutions) in rural, ‘last-mile’ schools, our co-author Dasha Uy reflects on the importance of this evidence for these school communities:
The Philippines has among the highest risks of negative effects of climate disasters such as typhoons or extreme heat (Adil et al., 2025), disrupting the health and education of Filipino children. For example, there is about a month of class disruptions every school year (Department of Education, 2024), and Filipino children show the highest rates of climate anxiety in the world (Aruta & Simon, 2022).
In response, our Healthy Learning Institutions program helps schools prioritise sustainable practices, disaster preparedness, and infrastructure that protects schools from disasters and promotes environmental protection, while also providing resources to train parents to foster their children’s resilience. This review affirms the Philippine government’s direction to prioritise interventions that respond to climate disasters and suggests that school-based mental health interventions may warrant resource allocation, particularly in LMICs. However, school-based health interventions must exist in tandem with a comprehensive, systems-level response (e.g., policy change) to the climate crisis with advocacy from health and public health practitioners.
Some last-mile schools can become completely submerged due to flooding and most have no access to an internet connection that would enable online learning.
Statement of interests
Dasha Uy, Monika Raniti, & Jennifer Dam – no conflicts of interest.
No AI was used in the writing of this blog post.
Edited by
Dr Nina Higson-Sweeney
Links
Primary paper
Okki Dona Laksmita, Min-Huey Chung, Joseph Kondwani Banda, Yann-Yann Shieh, Sumarni Djaka Waluya, Sri Warsini, & Pi-Chen Chang (2025). School‐based interventions for child and adolescent survivors of natural disasters–a systematic review and meta‐analysis of randomized controlled trials. Child and Adolescent Mental Health. https://doi.org/10.1111/camh.70029
Other references
Adil, L., Eckstein, D., Künzel, V., & Schäfer, L. (2025). Climate Risk Index 2026: Who suffers most from extreme weather events? Germanwatch e.V. https://www.germanwatch.org/sites/default/files/2025-11/CRI%20summary%20EN%202026.pdf
Aruta, J.J.B.R., & Simon, P.D. (2022). Addressing climate anxiety among young people in the Philippines. The Lancet Planetary Health, 6(2), pp.e81-e82. https://doi.org/10.1016/S2542-5196(22)00010-9
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd edn). Routledge Academic.
Phoenix Australia. (2021). Australian Guidelines for the Prevention and Treatment of Acute Stress Disorder, Posttraumatic Stress Disorder and Complex PTSD. Australian Government: National Health and Medical Research Council. https://www.phoenixaustralia.org/australian-guidelines-for-ptsd/
Hiller, R. M., Meiser‐Stedman, R., Fearon, P., Lobo, S., McKinnon, A., Fraser, A., & Halligan, S. L. (2016). Research Review: Changes in the prevalence and symptom severity of child post‐traumatic stress disorder in the year following trauma–A meta‐analytic study. Journal of Child Psychology and Psychiatry, 57(8), 884-898. https://doi.org/10.1111/jcpp.12566
Department of Education. (2024). Class disruptions due to storms rise to 35 for the school year. Republic of the Philippines. https://www.deped.gov.ph/2024/11/13/class-disruptions-due-to-storms-rise-to-35-for-the-school-year/
Wang, C. W., Chan, C. L., & Ho, R. T. (2013). Prevalence and trajectory of psychopathology among child and adolescent survivors of disasters: a systematic review of epidemiological studies across 1987–2011. Social Psychiatry and Psychiatric Epidemiology, 48(11), 1697-1720. https://doi.org/10.1007/s00127-013-0731-x
