The Adults Have Reviewed the Evidence, The Children Were Not Consulted: From RCT-Only Research to a Real Evidence-Based Standard for Children

A new systematic review was published a few months ago in Clinical Psychology Review, one of the most prestigious journals in the field (April, 2026). Alkærsig and colleagues set out to do something genuinely valuable: map the evidence base for 12 psychological treatments used with trauma-exposed children aged 12 and under. It is a serious piece of work, methodologically ambitious, and worth engaging with carefully.

But when I reached the play therapy section, I had to stop and read it again.

Not because the findings were surprising. Because the framing was.

This Substack article is my attempt to think through what that framing reveals: not to dismiss the review, but to name what it cannot see, and why that matters for every play therapist, researcher, and advocate working to strengthen this field.

I should be transparent about my position. I am a Registered Play Therapist-Supervisor (APPTA: RPT-S™) and the founding director of a Play and Filial Therapy practice. I reviewed the 4th edition of the Australian Psychological Society’s Evidence-Based Psychological Interventions review (2018), and provided forward feedback into the 5th edition, and am listed as a key stakeholder contributor in the published 5th edition acknowledgements. That context is directly relevant to what follows.

Alkærsig et al. searched six databases for quantitative pre-post studies across 12 treatments. For play therapy, they identified 18 eligible studies published between 1999 and 2025, involving 711 children. The meta-analysis of controlled studies (k = 10) yielded a small-to-moderate effect size (Hedges’ g = 0.39).

On the surface, this places play therapy in a cluster with EMDR and CPP, described as “emerging but promising”, behind TF-CBT, which earned the paper’s recommendation as first-line treatment.

That framing will be cited. It will shape funding decisions, training priorities, and possibly Medicare and NDIS policy conversations here in Australia. So, it’s especially important articles like this are engaged with through constructive review and critique.

Here is how the review defines play therapy:

“interventions making use of play as the central psychotherapeutic elements”

That is not a widespread clinical definition of play therapy. It is a descriptor so broad it could encompass almost anything involving a toy, a sand tray, a puppet, a drawing, or a game. And indeed, across the 18 included studies, the authors themselves note that 12 different play therapy approaches were represented. Twelve.

Here is what those twelve approaches actually were, according to the review's own study characteristics table (Alkærsig et al., 2026): child-centred play therapy, filial therapy (including the manualised Child-Parent Relationship Therapy protocol), intensive child-centred play therapy, intensive sibling group play therapy, Adlerian play therapy, cognitive behavioural play therapy, a hospital-specific educational play package, parent-child interaction therapy, a manualised programme combining CBT with cooperative play and creative-expressive exercises, an integration of play and experiential techniques with TF-CBT, studies using multiple or unspecified play therapy methods, and play therapy left otherwise unspecified. These are not variants of the same thing. They differ in theoretical orientation, relational stance, degree of directiveness, caregiver involvement, mechanism of change, and training requirements. And importantly, some, such as parent-child interaction therapy, would not be classified as play therapy by a play therapist at all.

Two modalities absorbed into this generic category deserve specific mention: filial therapy and group play therapy. Both carry their own independent evidence bases. Filial therapy, which positions the trained parent as the therapeutic agent under therapist supervision, has decades of controlled outcome research demonstrating effectiveness across a range of child presentations including trauma. Group play therapy has a well-developed evidence base particularly relevant to school and community settings, with studies demonstrating outcomes for children exposed to disaster, domestic violence, and abuse. Collapsing them into a single category does not just create a heterogeneous effect size. It obscures evidence that already exists.

What makes this especially striking is that a more rigorous, clinically grounded definition of play therapy was already available to these authors, in a document they themselves cite.

The Australian Psychological Society’s Evidence-Based Psychological Interventions review, now in its 5th edition (officially dated 2024, released 2025), defines play therapy as follows:

“Play therapy uses play modalities to engage children (and adults) in therapy and provide them with age-appropriate language and context to communicate with the clinician. Clinicians trained in play therapy use a systematic approach to identify patterns and themes in a child’s play. The clinician’s skill is to analyse what occurred in the session e.g., verbal and non-verbal cues, information that the child communicated through symbolic use of toys, drawings, and other play activities. In play therapy, the clinician must skilfully use play that is tailored to the child’s presentation.”

Critically, the APS review also explicitly includes ‘filial therapy’, ‘child-parent relationship therapy’, ‘floor time’, and ‘theraplay’ within its search terms for play therapy, acknowledging these as related modalities warranting specific attention within a coherent definitional framework, not submersion into a generic category.

I reviewed the 4th edition of this document and provided input into the 5th. The definition above reflects a process of engagement with the play therapy profession that took the modality seriously enough to consult the profession on its definition. The contrast with “interventions making use of play as the central psychotherapeutic elements” could not be more stark. One definition was built with input from the profession. The other was built without it.

The authors do acknowledge the heterogeneity in their discussion, noting that variation in findings “may be attributable to and influenced by the different approaches provided.” But this caveat appears after the effect size has already been calculated, reported, and positioned in the hierarchy of recommendations. The definition problem is upstream of and impacts on everything else.

The review identified only 18 eligible play therapy studies published between 1999 and 2025. This number deserves examination, because it is not simply a reflection of play therapy’s evidence base. It is partly a product of how the review went looking, partly a product of the inclusion criteria that excluded most play therapy research, which is largely qualitative or mixed methods, an appropriate design for an experiential therapeutic modality, and partly a product of what it failed to consult.

The authors deliberately narrowed their play therapy search terms. Notably, this is the opposite treatment given to less-researched approaches elsewhere in the same review: for methods like the Sleeping Dogs Method and NARM, search terms were intentionally broadened to capture as many citations as possible, while terms for a well-established method like TF-CBT were left at standard specificity, since its literature was already dense enough to surface easily. Play therapy was the only approach in the review to receive an additional restrictive filter rather than a standard or broadened one. Here is the exact string used:

((Play therap* AND trauma) OR (Play therap* AND “adverse childhood event*”) OR (Play therap* AND “adverse childhood experience*”))

In practice, this means a play therapy study only surfaced if its title or abstract paired “play therapy” with the specific words trauma, adverse childhood event, or adverse childhood experience. But these are still relatively contemporary terms in the play therapy literature. Earlier studies, including some of the field’s foundational trauma-relevant work from the 1990s and early 2000s, were far more likely to describe children as “abused,” “maltreated,” or “traumatized” than to use “trauma” or “adverse childhood experience” as a keyword. A search built around this specific terminology risks quietly excluding the older literature it most needs to capture. Narrowing the search means narrowing what gets found and therefore what meets the inclusion criteria.

More significantly: the APS Evidence-Based Psychological Interventions review is cited in this paper, but only as a treatment guideline source. It was not searched or drawn upon to identify eligible play therapy studies. This is not a minor gap. It is a structural one with direct consequences for the conclusions drawn.

Here is why it matters: the APS review’s evidence summary table for children and adolescents places play therapy at Level I evidence for posttraumatic stress disorder, alongside CBT-T (individual and group) and EMDR. Level I. The highest evidence classification. In the very document these authors cite as a guideline source, play therapy has already been assigned its strongest evidence rating for the exact condition this review examines.

A research team that had engaged with the APS review as a source rather than merely a citation would have found eligible, high-quality play therapy studies, a rigorously developed definition, and an evidence classification that tells a substantially different story than g = 0.39 labelled “emerging.”

Beyond the APS review, the authors’ own inclusion criteria excluded all studies lacking pre-post quantitative assessment of trauma symptomatology. This criterion alone eliminated 50% of eligible full-text play therapy studies. Fifty percent. That is half the literature, set aside because it did not use the right kind of measurement at the right time points. The conclusion that follows, that the evidence is thin, conflates the absence of a particular kind of assessment and monitoring documentation with the absence of meaningful practice and outcomes. These are not the same thing.

This is the issue I want to sit with longest, because it connects to something fundamental about what play therapy is and why it exists.

Even with the methodological sophistication of this review, which goes well beyond simple RCT inclusion, the entire framework prioritises quantitative pre-post designs with standardised symptom measures. For children aged 0 to 12, this is not a neutral methodological choice. It is a developmentally loaded one.

The play therapy studies in this review covered children aged 3 to 13 years. Yet the TF-CBT literature, which forms the backbone of the first-line recommendation, is heavily weighted toward older children. The paper reports that only 11 of 62 TF-CBT studies (18%) focused exclusively on preschool children aged 3 to 6, and only 24 (39%) included any preschool-aged children at all. Crucially, the paper provides no breakdown of how many TF-CBT studies included substantial numbers of children under 8: the developmental window where play-based approaches are most specifically indicated, and where the cognitive and verbal demands of CBT protocols require most significant adaptation. The 62-study headline is compelling. The evidence for the youngest children within it is far thinner than that number suggests.

Even the review’s own authors concede this point elsewhere in their discussion: play therapy and TF-CBT may simply suit different children within the same trauma-exposed population. They suggest that approaches leaning on developmentally sensitive channels, play, narration, drawing, bodily awareness, may be better suited to children with certain developmental capacities or trauma histories, including early-life trauma that disrupts development itself, precisely because they don’t depend on the language and cognition that TF-CBT requires.

Young children do not reliably self-report on symptom checklists. They process experiences through play, relationships, narratives, and their body. Caregiver reporting introduces its own layers of complexity, particularly where the caregiver is themselves traumatised, or where a child’s behaviour changes in ways caregivers do not immediately recognise as recovery.

The review itself calls for increased use of developmentally sensitive PTSD measures such as the DIPA, CATS-2, and the Darryl Cartoon test, a genuinely important recommendation. But it does not address the deeper epistemological question: what counts as evidence of change in a young traumatised child?

Meaningful change in young children shows up in relational security, immersive engagement in play, play complexity, narrative coherence, regulatory capacity, parental attunement, and much more. Processes that unfold over time and are often best captured through observation, case formulation, single-case experimental designs, and mixed methods research. None of these designs were eligible for this review. Process-oriented studies were excluded. Qualitative studies were excluded.

The authors acknowledge this limitation, and then continue to use the resulting hierarchy to recommend TF-CBT as first-line treatment and position play therapy as second-line. A recommendation built on a framework not designed to capture what play therapy does and how children actually engage in the therapeutic change processes is not a neutral evidence statement.

There is a question this review does not ask at all, and its absence is telling.

What do the children think?

Research ethics frameworks for adults include informed consent as a non-negotiable foundation. For children, the parallel concept, assent, requires that children are given age-appropriate information about their participation and that their willingness is meaningfully respected, in both clinical and research contexts. Yet the review contains no discussion of assent, child rights frameworks, or the ethical dimensions of conducting and evaluating trauma interventions with young children.

The UN Convention on the Rights of the Child is explicit: children have the right to express their views in all matters affecting them, with those views given due weight. Article 12 is not an aspirational principle. It is a ratified obligation in the majority of countries in which these studies were conducted or coordinated.

In a review evaluating 127 studies involving nearly 5,700 traumatised children, the complete absence of any engagement with children’s right to participation, in treatment decisions, in research design, in outcome evaluation, is a significant gap. It is also a symptom of the broader problem: a research framework built on adult epistemological assumptions, applied to children’s lives, without children’s perspectives integrated into the methodological architecture, and excluding qualitative and mixed methods studies that if any may have included a rights-based perspective.

To understand this problem clearly, it helps to look at who produced this review.

The author team is led by Professor Ask Elklit, one of Europe's most prolific psychotraumatology researchers, with an h-index of 47 (as of July 2026), over 340 publications, and more than 8,000 citations. He is head of the National Centre for Psychotraumatology at the University of Southern Denmark. Associate Professor Maria Louison Vang and Jesse Roest from the University of Applied Sciences Leiden provide meaningful mid-career research contributions. Dr Sille Schandorph Løkkegaard is an early-career lecturer in psychology. The remaining three authors, Mette Alkærsig, Trine Banzon, and Anni Ravn Stein, are PhD students, and it is their hands that did much of the primary data work

This is not a team that lacks rigour or seniority. Elklit’s research output alone places him among the most impactful trauma researchers in Europe. The team is accomplished, the methodology is sophisticated, and the review represents a serious scholarly contribution.

But look more closely at the composition of that research team. Every author sits within psychology and psychotraumatology. The gravitational pull of Elklit’s research framework, trauma epidemiology, PTSD measurement, adult and adolescent populations, quantitative outcome designs, will have shaped the methodological architecture of this review from the ground up.

There is no registered or credentialled play therapist anywhere in this team. No allied health practitioner with experience of play-based intervention. No one whose training would have flagged that “interventions making use of play as the central psychotherapeutic elements” is not a clinical definition of play therapy. No one who would have known to look at the APS review as a source rather than a citation. No one who would have recognised that filial therapy and group play therapy have independent evidence bases. No one who would have thought to ask about child assent and how child rights can be upheld within the research methodology.

This is not a criticism of the individuals. It is a structural observation about what happens when specialist allied health modalities are evaluated entirely within another discipline’s epistemological framework, and why that framework, however rigorous internally, may not know what it does not know from another discipline.

I am not arguing that play therapy should be exempt from evidence scrutiny. Quite the opposite. The profession needs rigorous, cumulative, well-designed outcome research. We need standardised outcome measurements integrated into clinical practice. We need the longitudinal, controlled studies that will allow play therapy to occupy its rightful place in treatment guidelines, which, as the APS review already demonstrates, it can and does.

What I am arguing is that the framework for that evidence needs to be developmentally appropriate, methodologically integrative, children’s rights informed, and constructed in partnership with the clinicians who hold the practice knowledge.

This connects to a broader point about what counts as evidence for practitioners. Frameworks such as Integrating Theory, Evidence and Action (ITEA) argue that current models of evidence-based practice lean too heavily on quantitative research, when research is only one of several ways of knowing that clinicians draw on to guide their practice. A useful evidence base for the profession has to deliver genuine evidence-based practice (EBP), not a narrow, quantitative-only imitation of it (Hitch et al., 2014).

A review that cites a document rating play therapy at Level I for PTSD in children without appearing to notice, excludes half the available literature, does not transparently account for what happens to children under 8, and never once asks what children themselves experienced, will not give us an accurate map. It will give us a map of what adult measurement instruments can see, from the vantage point of adults looking down.

The children who come to play therapy: the three-year-old who has stopped speaking, the six-year-old who re-enacts violence they have witnessed in the sandtray, the eight-year-old who finally, after months, engages in reciprocal relational play, deserve a research framework built to see them, and built, in part, around asking them to share their insights.

Building that framework is the work of this generation of play therapy researchers, clinicians, and advocates. It will not come from waiting for someone else to count us.

Alkærsig, M., Banzon, T., Roest, J., Elklit, A., Stein, A.R., Vang, M.L., & Løkkegaard, S.S. (2026). Psychological treatments for young children suffering from trauma-related symptomatology: Systematic review and partial meta-analyses of the current evidence-base for 12 methods. Clinical Psychology Review, 125, 102725. https://doi.org/10.1016/j.cpr.2026.102725

Australian Psychological Society. (2018). Evidence-based psychological interventions in the treatment of mental disorders: A literature review (4th ed.). APS.

Australian Psychological Society. (2024, released 2025). Evidence-based psychological interventions in the treatment of mental disorders: A literature review (5th ed.). APS.

Hitch, D., Pépin, G., & Stagnitti, K. (2014). The Integrating Theory, Evidence and Action (ITEA) method: A procedure for helping practitioners translate theory and research into action. British Journal of Occupational Therapy, 77(12), 592–600. https://doi.org/10.4276/030802214X14176260335183

Dr Kate Renshaw (PhD, RPT-S™) is a play therapist, researcher, educator, author, and advocate with over 20 years of experience working therapeutically with children and families. She is the developer of the TORA and the Multi-Tier Play Therapy (MTPT) framework, and is the founding director of Play & Filial Therapy. Her work spans clinical practice, doctoral research, sole-authored practitioner guidebooks, parliamentary advocacy, and international conference presentations. She is based in Ballarat, Victoria, Australia.

🌐 playandfilialtherapy.com

💼 linkedin.com/in/drkaterenshaw