How educational psychologists use cognitive behavioural therapy interventions: a systematic literature review

ABSTRACT

This systematic literature review explored how educational psychologists (EPs) use cognitive behavioural therapy (CBT) interventions within school settings, with whom and to what effect. The information available on the intervention design and delivery was benchmarked against Fuggle et al.‘s framework that identified CBT competencies required for working with children and young people (CYP). Research into school-based CBT interventions involving EPs between 2013 and 2021 was systematically identified and assessed for quality using quantitative and qualitative frameworks. The aggregative analysis highlighted the heterogeneity of CBT interventions being used to support a range of outcomes. The heterogeneity of interventions made it difficult to draw decisive conclusions about the impact of CBT. Although intervention impact varied, overall a positive impact was observed particularly for CYP with anxiety. Benchmarking interventions against a CBT framework enabled competency profiling within EP practice, suggesting EPs could potentially work more collaboratively with school staff and parents. Implications for future research and practice are discussed.

KEYWORDS:

• Cognitive behavioural therapy
• CBT
• educational psychologist
• intervention
• school-based

Introduction

Educational psychologists’ (EPs’) role in supporting mental health and wellbeing

There is broad recognition of the role of EPs in supporting mental health and wellbeing (for example, Grieg et al., 2019). Indeed, in a United Kingdom (UK) survey involving 680 schools, EPs were identified as the most frequent provider of mental health and wellbeing support for individual children and young people (CYP) (Sharpe et al., 2016). In a review of international literature, Simpson and Atkinson (2021) examined therapeutic interventions delivered by schools and EPs, finding that all had at least some positive effect on intended outcomes. Of the 22 therapeutic modalities used, cognitive behavioural therapy (CBT) was the most frequently used, in nine of the 16 research studies. Simpson and Atkinson (2021) called for more focused and specific research into the different therapeutic approaches identified within their review.

CBT in schools

A narrow definition of CBT is as a specific therapy or intervention that aims to change thoughts or behaviour; while a broader definition views CBT as recognising the relationships and interplay between thoughts, feelings and behaviours (Graham, 2005). Stallard (2005) described how CBT is “based on the underlying (psychological) assumption that affect and behaviour are largely a product of cognitions, and as such, that cognitive and behavioural interventions can bring about some changes in thinking, feelings and behaviour” (p.1). The evidence base for CBT is strong for supporting CYP with anxiety (Fuggle et al., 2013); it was highlighted as an evidence-based intervention for CYP with anxiety and adolescents with depression in the UK green paper (Department of Health and Department for Education, 2017) on mental health.

In schools, CBT has been used to support a range of positive outcomes including general anxiety (for example, Ruocco et al., 2018), exam anxiety (Gregor, 2005), and depression (Garvick et al., 2016). Targeted CBT interventions have been used with specific groups or vulnerable young people such as CYP who are on the autism spectrum for example (Higgins et al., 2019), and those experiencing emotionally based school avoidance (Lee, 2019). CBT has also been used in schools to support staff wellbeing and confidence (Squires, 2010).

The majority of evidence for the effectiveness of CBT comes from clinical research randomised control trials, often referred to as the “gold standard” of research, but which can lack examination of implementation factors or consideration of why CBT is effective for some individuals and not others (Williams, 2015, p. 344). Dunsmuir and Hardy (2016) suggested that EPs should exercise caution when interpreting clinically based findings, as school-based therapeutic interventions are not delivered in ideal conditions or with homogenous populations.

CBT competencies

Roth and Pilling (2007) developed a competency framework for practitioners working with adults, to provide a benchmark for what a CBT intervention should involve. Fuggle et al. (2013) later devised one for practitioners working with CYP, which included appropriately adapted techniques, and 24 competencies organised into six overarching domains, consistent with other CBT frameworks (Blackburn et al., 2001; Roth & Pilling, 2007). These are as follows: setting the right context, therapeutic alliance, collaborative practice, structuring the therapeutic process, CBT skills aimed at facilitating understanding and CBT skills aimed at facilitating coping, acceptance and change.

EPs’ use of CBT

Position papers (Pugh, 2010; Rait et al., 2010; Squires, 2010) discussed further applications of CBT, including training, supervising others, and indirect casework such as consultation or building staff capacity. Squires (2010) described these wider applications as “cognitive behavioural psychology” (p. 280). Simpson and Atkinson’s (2021) systematic literature review reported that use of CBT with CYP was most frequently via targeted group interventions, where it was also combined with other modalities, including attribution retraining.

Rationale for the current literature review

Building on the previous literature (cf Simpson & Atkinson, 2021; Squires, 2010), this review focused specifically on EPs’ use of CBT within UK schools. In doing so, it aimed to address the following research questions:

1. How is CBT being used by EPs, with whom, and to what effect?

2. What components of CBT are the most prominent within EPs’ practice?

Materials and methods

Databases were systematically searched for research where EPs had used CBT in school settings. A clear search strategy and classification criteria (Gough et al., 2017) helped ensure the quality and replicability of the review design, as outlined in the Preferred Reporting Item for Systematic reviews and Meta Analysis (PRISMA) diagram (see Figure 1).

Figure 1. PRISMA diagram.

Search strategy

The search strategy attempted to ensure that relevant research published within the period 2013–2021 inclusive was identified, following the publication of specific guidance for supporting children and young people in schools (Fuggle et al., 2013). Systematic searches were completed between September 2021 and January 2022 on ASSIA, ERIC, EThOS, Web of Science and Google Scholar. The search terms educational psychologist/educational psycholog*, CBT and school/school* were used in advanced searches. According to Google Scholar, exclusions were applied for drug*, smok*, alcohol*, obes* weight, HIV and medical and screened to the point where relevant papers were no longer shown from 200 consecutive results. UK educational psychology journals Educational and Child Psychology; Educational Psychology Research and Practice and Educational Psychology in Practice were also hand searched. While included theses were not peer-reviewed, it was felt the academic rigor/quality of research was maintained through the doctoral examination process. Other non-peer-reviewed literature was not included. Study inclusion criteria were:

• Interventions must have a CBT component.

• CBT intervention takes place in a school.

• Participants are of school or college age (4–18 years).

• EP involvement in the intervention is defined as delivery or supporting the delivery.

• Published 2013–2021, following publication of Fuggle et al. (2013) guidance.

• Empirical research.

• Peer-reviewed through publication or doctoral examination.

• Written in English.

• Completed in the UK.

Weight of evidence

Gough et al. (2017) weight of evidence (WoE) criteria were applied against each paper identified as meeting the initial search criteria. The approach closely followed that used by Simpson and Atkinson (2021), who reviewed therapeutic practice by educational and school psychologists more generally. Ratings were based on information reported, to avoid unfair bias that could emerge from authors not responding to corresponding queries. Where thesis findings also appeared within peer-reviewed publications, WoE criteria were applied to both documents and the highest rated included.

WoE A evaluated the research for methodological quality using critical appraisal frameworks developed to support the review of qualitative and quantitative research (Woods, 2020). Where research was mixed methods, both frameworks were applied and the highest rating used. WoE B evaluated methodological appropriateness, by considering which key components were evidenced, using Fuggle et al. (2013) competency framework for CBT with CYP. WoE C determined whether the study's focus was relevant to the research questions, by scrutinising: the nature of EP involvement; whether intervention was CBT or composite intervention; whether the CBT process was considered; and the extent to which intervention content was specified. For WoE D, each study was classified as either low, medium or high for each WoE and then aggregated studies with two or more low ratings were excluded from the final synthesis.

Papers were initially co-rated by the first author, with three co-rated with the second author, in the role of research supervisor. From the ensuing discussion, some WoE C criteria were revised. An inter-rater agreement of 94% was deemed sufficient for the researcher to rate the remaining research papers independently.

Data synthesis

An aggregative method of analysis was used to synthesise the data (Gough et al., 2013). Similar to the findings of Simpson and Atkinson (2021) studies were heterogeneous in nature, meaning that statistical methods of aggregative analysis would not be meaningful. Data were synthesised against the identified foci (Gough et al., 2013) described in the headings of Table 1 including EP role and intervention criteria. Full references cited by authors or research in Table A1 papers are recorded in the Appendix.

Table 1. Study characteristics.

Research study	Participants			Intervention						Outcomes
Author and year	Sample size	Age	Gender	Inclusion criteria	Intervention focus	Intervention programme	Intervention delivery	EP role	Research design	Qualitative	Quantitative
Boden (2020)	1	16	Male	Discussion with SENCO	Anxiety and depression	EP developed bespoke CBT intervention using Fuggle et al. (2013) and specific treatment protocols; no fidelity measures	8 × 45 minute weekly sessions. Individual. Targeted intervention delivered by TEP.	Intervention	Mixed methods retrospective case study	Therapeutic diary, case notes, narrative summary detailing implementation	No overall clinically significant changes on the BYI (Beck et al., 2005) & RCADS-P (Chorpita et al., 2005). Some scores on individual subscales increased or decreased. Weekly TME (Dunsmuir et al., 2009) – no overall changes but some weekly change.
Brightmore (2016)*	85	9–10	Male and female	No information available	Emotional literacy and wellbeing	Manualized CBT based intervention using Think Good Feel Good Workbook (Stallard, 2002). Fidelity measures.	6 × 2 hours weekly sessions. Universal intervention delivered to class by TEP (with support) or class teacher.	Intervention delivery and evaluation.	Fixed experimental design	No qualitative measures	Significant positive differences between CBT and control groups for emotional literacy (SELR) (Faupel, 2003) and SDQ (Goodman etal., 2000). Negative outcomes when CBT was combined with SEAL (Department for Education and Skills 2007).
Clarke et al. (2017)	14	11–14	Male	ASD diagnosis. High anxiety levels (discussion with SENCO	Anxiety and coping skills	Manualized Adapted Attwood (2004) Exploring feelings: CBT to manage anxiety for children aged 10–12; No fidelity measures.	6 × 1 hour weekly sessions; Targeted group intervention delivered by TEP.	Participant selection, screening	Mixed methods quasi experimental deisgn	Semi structured interviews with parents and children. A range of implementation factors identified.	SCAS-C (Spence, 1997) and SCAS-P (Spence 1997). Significant differences maintained at follow up (6–8 weeks); CSCY (Brodzinsky et al., 1992): Significant differences in behavioural avoidance and problem solving. No significant differences for cognitive avoidance and assistance. Less cognitive avoidance at6–8 week follow-up.
Collins et al. (2014)	317	9–10	Male and female	No criteria	Anxiety and coping skills	Manualized locally developed mental health CBT-based programme (Waters et al, 2010) underpinned by CBT principles. No fidelity measures.	10 × weekly sessions (length unspecified). Universal intervention delivered to class by school psychologist or class teacher	Intervention delivery and evaluation.	Mixed methods experimental design	No qualitative measures	CSI (Amirkhan, 1990) and SCAS-C (Spence, 1997). Significantly improved coping skills and reduced anxiety post intervention maintained at follow up (6–8 weeks).
(Edgington, 2014) *	12	11–16	Male	ASD diagnosis. IQ above 70	Sensory. experiences/sensory sensitivity. Secondary measures; repetitive behaviour and anxiety	EP developed bespoke CBT-based intervention based on the literature; no fidelity measures.	8 × 45-minute weekly sessions; targeted group intervention delivered by TEP	Intervention design. Participant screening. Intervention delivery and evaluation.	Mixed methods single-case experimental design.	Focus group with parents. Seni-structured interviews with participants. Enhanced cognitive awareness and coping behaviours.	Primary measures: SSP (Dunn, 1999), AASP (Brown & Dunn, 2002); secondary measures: SCAS-P (Spence, 1997) and RBQ (Turner, 1999). No significant statistical difference at post intervention or follow up (8 weeks) at group or individual level on all measures.
Green and Atkinson (2016)	5	11–13	Male and female	School staff identified following TAHMS training.	Anxiety, emotional distress and coping skills	Manualized CBT-based intervention. FRIENDS with some adaptations/omissions. Fidelity measures.	9 × 1 hour weekly sessions; targeted group intervention delivered by two learning mentors.	Training; Intervention evaluation; Analysis of implementation factors.	Mixed methods single case experimental design.	Semi-structured interventions with learning mentors identified a range of implementation factors.	SCAS-C (Spence, 1997): Thesis data shows non significant reduction in anxiety except for one CYP. No reliable changes in coping skills measured by Kidscope vignettes or emotional distress measured by PI-ED (O’Connor et al., 2010). No change at follow up (unspecified time).
Grimes (2015)*	41	11–16	Male and female	Specified within ART handbook; not receiving another intervention; perceived ability to reflect.	Social competence: Pro social skills	Manualised CBT based intervention– Aggression Replacement Training	10 × weekly sessions (time unspecified); Targeted group intervention.	Intervention delivery; Intervention evaluation	Mixed methods quasi experimental design.	1:1 semi-structured interviews with pupils and facilitator questionnaire. All perceived intervention to have positive impact. Factors contributing to success identified.	SRM-SF (Gibbs et al., 1991): Significant change in moral reasoning. No significant progress in social skills or problem-solving skills.
Kite (2020)*	8	14–15	Male and female	TEP developed inclusion and exclusion criteria. Discussion between TEP and school staff.	School work related anxiety; General anxiety; Meta-cognitive learning strategies	TEP developed bespoke CBT intervention: Individualised behaviour experiments following a 15-step plan. No fidelity measures.	6 × 30 minute weekly sessions; Targeted individualised intervention delivered by learning support assistants.	Intervention design; Participant selection; Staff training; Staff supervision; Intervention evaluation.	Mixed methods quasi-experimental case study.	Consultation/supervision notes; LSA completed online questionnaire: Identified implementation factors.	SCAS (Spence, 1997): Significant reduction in generalised anxiety except increase for one CYP; SDQ (Goodman etal., 1997): Non-significant reduction in perceived difficulties except increase for one CYP; MSLQ-SV (Pintrich & De Groot 1990): Mixed outcomes in adaptive and productive strategies for learning; Liker scale: Mixed outcomes CYP schoolwork anxiety and confidence.
Lake (2014)	18	12–13	Male and female	Elevated but not clinically significant anxiety on SCAS	Reduce anxiety	TEP developed bespoke CBT based intervention. Based on Stallard (2005); Stallard (2008) and Fuggle et al. (2013). Fidelity measures.	6 × weekly sessions (time unspecified); Targeted group intervention delivered by two of four TAs.	Intervention design; Participant screening; Staff supervision; Implementation evaluation	Mixed methods quasi experimental design.	Nominal group technique with CYP. Positive effects: Aided development of problem-solving skills and coping skills including seeking social support and identifying triggers. Focus group with CYP: Some positive and some negative coping strategies; implementation factors identified.	SCAS-C and SCAS-P (Spence, 1997): Some reduction in anxiety, on SCAS-P this was for some CYP (not statistically significant).
Luxford et al. (2017)	35	11-	Male and female	Diagnosis of ASD; SCQ high score; verbal and total IQ above 70; clinically significant anxiety of SCAS-C and SCAS-P.	Anxiety; social worry. Secondary: social responsiveness, attentional control and attention to threat.	Manualized CBT intervention; Attwood (2004). No fidelity measures.	6 × 90 minutes weekly session. Target group intervention delivered by TEP (supported by teaching assistant).	Participant screening. Intervention delivery and evaluation.	Randomized control experimental design.	No qualitative measures.	SCAS-C and SCAS-P (Spence, 1997); significant positive change, maintained at follow-up (6 weeks). SAS (Lyneham et al., 2008); some improvement (not statistically significant). SWQ (Spence, 1995); some significant changes on pupil and teacher reported worries. SCQ (Rutter et al. 2003); marginally increased teacher reported social responsiveness; Erickson flanker task and emotional stroop task; no change in attentional control or attention to threat.
O’Callaghan and Cunningham (2015)	9	9–11	Male and female	Mild, moderate and severe anxiety on BYI. Teacher-identified as having anxiety or depression.	Depression, anxiety and self-concept.	Manualised CBT programme; Cool Connections. Fidelity measures.	10 × 1.5 hr weekly sessions; targeted group intervention delivered by two of four facilitators (school staff).	Participant selection; staff supervision; staff training; intervention evaluation.	Mixed methods experimental design	Written feedback questionnaires completed by CYP. All CYP reported	BYI Scales (Beck et al., 2005); depression: statistically significant improvements; anxiety: statistically significant improvements; self-concept: some improvement (not statistically significant).
Tudor (2014)	15	12-	Male and female.	School referral: at risk of exclusion; pastoral support plan; estimated IQ greater than 8-	Conduct and behaviour	EPS team developed bespoke CBT-based intervention; fidelity measures	6 × 1hour weekly sessions; targeted group intervention, delivered by TEP with additional school staff support.	Participant screening; intervention deliver and evaluation.	Mixed methods experimental design	Semi-structured interviews with teachers: implementation recommendations; focus group with pupils: intervention received positively	Significant reduction in clinical severity for conduct disorder; no significant change across all other measures; positive, non-significant changes for mean scores (ODD, PCS & SDQ; Goodman etal., 2000). No changes for ICU (Frick, 2004); KSads PL Version 1.0 (Kaufman et al, 1997); BYI-II (Beck et al, 2005) (disruptive behaviour subscale); reported behaviours on SIMs, IRI and CSGM (Lochman et al., 1993).
Weeks et al. (2017)	19	11–14	Female	Staff identified CYP who would benefit from intervention to reduce anxiety	Anxiety	EP developed bespoke CBT-based intervention based on Seiler (2008), Stallard (2008) and Squires (2001)	6 × weekly sessions (length unspecified); targeted group intervention, co-delivered by an EP and co-facilitating TA in one school.	Intervention design, delivery and evaluation	Mixed methods quasi experimental design	Semi structured interviews with school staff and pupils; parents focus group; intervention feedback questionnaire. Some positive impact on anxiety. Process and implementation variable identified	No group outcomes reported on measures taken. SCAS-C, SCAS-TR (Spence, 1997), CATS (Schniering & Rapee, 2002), SDQ (Goodman etal., 2000), SAS-TR (Lyneham et al., 2008). One student increased in anxiety on SCAS-C but reduction on SCAS-TR.

Notes. AASP =; ART = Aggression Replacement Therapy; ASD = Autism Spectrum Disorder; BYI = Beck youth inventory; CATS = Children’s automatic thought scale; CAYP = children and young People; CBT = Cognitive Behavioural Intervention; CSGM = Children’s Social Goals Measures; CSI = Coping Strategies Indicator; CSCY = Coping Scale for Children and Youth; EPS = Educational Psychology Service; ICU = Inventory of Callous Unemotional Traits; IRI = Interpersonal Reactive Index; KSads-PL = Kiddie Schedule of Affective Disorder and Schizophrenia – Present Lifetime; LSAs = Learning Support Assistants; MSLQ-SV = Motivated strategies for learning questionnaire – Student version; ODD = Oppositional Defiance Disorder; PCS = Peer Conflict Scale; PI-ED = Paediatric Index of Emotional Distress; RBQ = Repetitive Behaviour Questionnaire; RCADS = Revised Children’s Anxiety and Depression Scale; SAS = School Anxiety Scale; SAS-TR = school anxiety scale – TR; SCAS = Spence Children’s Anxiety Scale; SCAS-C = Spence Children’s Anxiety Scale – Children; SCAS-P = Spence Children’s Anxiety Scale – Parent; SCAS-TR = Spence Children’s Anxiety Scale – Teacher Rated; SCQ = SDQ = Strengths and Difficulties Questionnaire; SEAL = Social and Emotional Aspects of Learning; SELR = Southampton Emotional Literacy Resource; SENCo = Special education needs Co- ordinator; SIMs = School Information Management System; SRM-SF = Sociomoral Reflection Measure – Short Form; SSIS = Social Skills Improvement System; SSP = Short Sensory Profile; SWQ = Social Worries Questionnaire; TA = Teaching assistant; TAMHS = Targeted approach to mental health in schools; TEP = Trainee educational psychologist; TME = Target Monitoring Evaluation.

All references in the supplemental material within the body of the main reference list.

Results

Thirteen studies met inclusion criteria (see Table 1); with two studies having a high overall WoE, and 11 medium. Five studies were published in peer-reviewed journals and eight were theses; of which three had aspects of their research published in peer-reviewed journals. Where a thesis was used in the analysis and findings, this is indicated in Table 1 with an asterisk.

Interventions

Ten studies took place with secondary school-aged children and three with primary children aged nine and over. The majority of interventions (nine) were described as CBT-based, with Kite’s (2020) the only study to use a single component of CBT (behaviour experiments) as the primary intervention focus. In six studies, EPs were involved in designing bespoke interventions based on pre-existing intervention programmes, with Boden (2020) and Lake (2014) referencing Fuggle et al. (2013) as a resource that informed intervention design. Five manualized interventions were used across six studies. These were Atwood’s Exploring Feelings (Clarke et al., 2017; Luxford et al., 2017), Cool Connections, FRIENDS for life (O’Callaghan & Cunningham, 2015), Aggression Replacement Training (Grimes, 2015) and Think Good Feel Good (Brightmore, 2016). Clarke et al. (2017) piloted an adaptation of Atwood’s (2004) Exploring Feelings intervention.

Dunsmuir and Hardy (2016) define three types of intervention – universal intervention, targeted intervention and specialist intervention. Universal interventions are defined as interventions that are “administered to every member of a population irrespective of risk level”, targeted interventions are “designed for at risk groups” and specialist interventions are “delivered to individuals and groups that require high levels of support from professionals with greater expertise” (Dunsmuir & Hardy, 2016, p. 7). Eleven interventions were targeted and two were universal (Brightmore, 2016; Collins et al., 2014). Boden (2020) delivered an individualised intervention with a 16-year- old who had anxiety and depression. Most interventions were group-based; two involved delivering CBT to individual children and two were whole class. The length of intervention ranged from six to ten weeks and session length, where specified, varied from 30 to 120 min. EPs or trainee EPs directly delivered the intervention in nine studies, of which three (Brightmore, 2016; Luxford et al., 2017; Weeks et al., 2017) were supported by school staff. In four studies, the CBT intervention was delivered by school staff, with the EP having a role in training (Green & Atkinson, 2016; Kite, 2020) and/or supervising (Kite, 2020; Lake, 2014; O’Callaghan & Cunningham, 2015).

In seven studies, decisions for participant inclusion were made by staff members, and in three of these, discussions were had with the EP about participant suitability. Kite (2020) included specific inclusion/exclusion criteria to ensure participant suitability and O’Callaghan and Cunningham (2015) employed additional screening using the Beck Youth Inventory (BYI, Beck et al., 2005). EPs were involved in screening in five studies; three of which involved intelligence quotient (IQ) screening (Clarke et al., 2017; Edgington, 2014; Tudor, 2014). Three studies used measures of anxiety or depression as inclusion criteria (Lake, 2014; Luxford et al., 2017; O’Callaghan & Cunningham, 2015). In Lake (2014), only CYP who had elevated anxiety, which was not clinically significant, were included in the research, whereas Luxford et al. (2017) only worked with individuals whose anxiety was clinically significant.

Focus and outcomes

Anxiety was the most frequently targeted outcome, within 10 studies. Kite (2020) focused on school work anxiety, although positive outcomes in relation to overall anxiety were found. Other outcomes targeted included self-concept (O’Callaghan & Cunningham, 2015); social responsiveness (Luxford et al., 2017); depression (Boden, 2020; O’Callaghan & Cunningham, 2015); moral reasoning (Grimes, 2015); sensory processing (Edgington, 2014); behaviour (Grimes, 2015; Tudor, 2014) and emotional literacy and wellbeing (Brightmore, 2016). Seven of the 13 interventions targeted multiple outcomes.

Some positive outcomes were reported in all studies. Although Green and Atkinson (2016) reported no overall positive effect of the FRIENDS (Barrett, 1996) intervention there were individual changes in anxiety on the SCAS-C (Spence, 1997); intervention fidelity and implementation issues were identified in the research. In three of the mixed method studies positive outcomes were reflected in qualitative, rather than quantitative outcome measures (Edgington, 2014; Green & Atkinson, 2016; Tudor, 2014); while Boden (2020) reported limited changes on some specific subscales. In 10 studies, there was a positive impact on anxiety, with this being reported to be statistically significant in five (Clarke et al., 2017; Collins et al., 2014; Kite, 2020; Luxford et al., 2017; O’Callaghan & Cunningham, 2015).

Two studies reported a positive impact on depression (Boden, 2020; O’Callaghan & Cunningham, 2015), although this was considered limited in Boden’s (2020) research. Two studies (Brightmore, 2016; Kite, 2020) reported changes in general wellbeing, measured using the Strengths and Difficulties Questionnaire (SDQ, Goodman, 1997); while O’Callaghan and Cunningham (2015) reported a positive, although not statistically significant, change in self-concept and Brightmore (2016) reported significant changes in emotional literacy. Other positive outcomes included positive social outcomes (Luxford et al., 2017; Tudor, 2014) such as a reduction in social worries (Luxford et al., 2017). Grimes (2015) reported changes in moral reasoning, but not in other behavioural measures. No change in at least some of the study measures was reported in nine of the studies, including in relation to sensory needs (Edgington, 2014), and attentional control/attention to threat (Luxford et al., 2017). Although Boden (2020) reported a positive effect of the intervention on anxiety and depression, this was described as only limited.

Some negative impact was reported. Kite (2020) and Weeks et al. (2017) both referred to individual CYP whose anxiety had increased during the time of the intervention, and Brightmore (2016) described a negative impact in some measures on the SDQ (Goodman, 1997) and the Southampton Emotional Literacy Resource (Faupel, 2003) when a CBT-based intervention was combined with the Social and Emotional Aspects of Learning (SEAL) curriculum.

CBT competencies

CBT competencies within each intervention were identified as part of WoE B, using the Fuggle et al. (2013) framework to locate “essential elements of CBT practice with children and young people” (p. 266), within delivery descriptions. A summary of the findings are presented in Table 2. Only two of the studies (Boden, 2020; Lake, 2014) made explicit reference to Fuggle et al. (2013) competencies within their intervention design.

Table 2. Weight of evidence B research papers benchmarked against Fuggle et al. (2013) competency framework.

	Setting the right context^1			Therapeutic alliance^2			Collaborative practice^3						Structuring the therapeutic process^4			CBT skills aimed at facilitating understanding^5				CBT skills aimed at facilitating coping, acceptance and change^6					Total
Research paper	A1	A2	A3	B1	B2	B3	C1	C2	C3	C4	C5	C6	D1	D2	D3	E1	E2	E3	E4	F1	F2	F3	F4	F5	Total
Boden (2020)	1	1	1	1	1	1	1	1	1	0	0	1	1	1	1	1	0	1	1	1	0	1	1	1	20
Brightmore (2016)	1	0	1	0	1	1	0	0	0	0	0	1	1	1	1	1	1	1	0	1	1	0	1	1	15
Clarke et al. (2017)	1	1	1	1	1	1	0	0	0	0	1	1	1	1	1	1	1	0	0	1	1	0	0	1	16
Collins et al. (2014)	1	0	1	0	0	0	0	0	0	0	0	1	1	1	0	1	1	1	0	1	1	0	0	0	10
Edgington (2014)	1	1	0	1	1	1	1	1	0	0	0	1	1	1	1	1	1	0	1	1	1	1	1	0	18
Green and Atkinson (2016)	1	0	1	0	1	1	0	0	0	0	0	1	1	0	1	0	1	1	0	1	1	0	0	1	12
Grimes (2015)	1	0	0	1	1	1	0	0	0	0	1	1	1	1	1	1	1	1	0	0	1	1	1	1	16
Kite (2020)	1	0	1	1	1	0	1	0	0	0	1	1	1	1	0	1	1	1	0	1	1	0	0	1	15
Lake (2014)	1	0	1	1	1	0	1	1	1	0	0	1	1	1	1	1	0	1	0	1	1	1	1	1	18
Luxford et al. (2017)	0	0	1	0	1	0	0	0	0	0	0	1	1	0	1	0	0	0	0	1	0	1	0	0	7
O’Callaghan and Cunningham (2015)	1	0	1	1	1	1	0	0	1	0	0	1	1	1	1	1	1	1	0	1	1	1	0	1	17
Tudor (2014)	1	0	1	1	1	1	0	0	0	0	0	1	1	1	1	0	0	0	0	1	1	1	1	0	13
Weeks et al. (2017)	1	0	0	0	1	1	0	0	0	0	0	1	1	0	1	1	1	0	0	1	1	0	0	1	11
Total	12	3	10	8	12	9	4	3	3	0	3	13	13	10	11	10	9	8	2	12	11	7	6	9

Note. ¹ Setting the right context: A1 Ethical Practice) A2 (Active reference to and/or involvement of parent/carer/family members) A3 (Active reference to school/college/work factors and/or involvement of school/college staff/people at work).

^aTherapeutic alliance: B1 (Empathy), B2 (Child/young person centred), B3 (creativity).

^bCollaborative practice: C1 (Joint session planning), C2 (being goal focussed), C3 (providing a rationale), C4 (summarising), C5 (seeking feedback), C6 (monitoring and evaluating progress).

^cStructuring the therapeutic process: D1 (preparing for the session), D2 (pacing and time management), D3 (between session tasks).

^dCBT Skills aimed at facilitating understanding: E1 (psycho education about CBT), E2 (recognising emotions), E3 (discovering cognitions), E4 (developing a shared formulation).

^eCBT Skills aimed at facilitating coping acceptance and change: F1 (developing coping strategies and acceptance), F2 (problem solving) F3 (encouraging positive behaviour) F4 (specific behaviour change techniques) F5 (cognitive change methods).

Particularly salient competencies, addressed in either 12 or 13 studies, were as follows: ethical practice; child-centred practice; monitoring and evaluating progress; preparing for the session; and developing coping strategies and acceptance. Competencies evident in three or less interventions were as follows: active involvement of the parent/carer or family member; being goal focused; providing a rationale; summarising, feedback and developing a shared formulation.

In setting the right context domain (Fuggle et al., 2013) all but one study gave descriptions of how CBT was practised ethically. Most interventions (10) involved the school setting, with only Boden (2020), Clarke et al. (2017) and Edgington (2014) involving families. In terms of building a therapeutic alliance, all but one study (Collins et al., 2014) delivered CBT in a child-centred way. Eight papers referenced working in an empathetic way and nine gave at least one example of using CBT wcreatively.

Monitoring and evaluating progress was the most salient competency within the domain of collaborative practice. Although progress was mostly measured using standardised questionnaires, Boden (2020) used target monitoring evaluation to monitor progress. Four studies (Boden, 2020; Edgington, 2014; Kite, 2020; Lake, 2014) jointly planned sessions with CYP and, with the exception of Kite (2020), involved CYP in goal setting. Session feedback was sought in three studies (Clarke et al., 2017; Grimes, 2015; Kite, 2020), with Kite (2020) discussing the importance of seeking regular feedback in monitoring the intervention. No studies gave examples of summarising.

Within all studies, it was evident that sessions were well planned and prepared. Ten studies considered pacing and time management as part of intervention planning. For example, O’Callaghan and Cunningham (2015) delivered additional intervention sessions to ensure that all the planned weekly content had been delivered. Within 11 studies between session tasks were set. CBT skills aimed at facilitating understanding involved the inclusion of psychoeducation in 10 interventions. Nine interventions involved supporting CYP to recognise emotions and eight involved discovering cognitions. Only Boden (2020) and Edgington (2014) described the development of a shared formulation between the EP and CYP.

To support the development of CBT skills, all interventions involved the development of coping strategies and acceptance except for Grimes (2015). Eleven supported CYP in developing problem-solving skills, while seven referred to encouraging positive behaviour. Six involved specific behavioural change techniques and nine referred to cognitive change techniques.

Discussion

How are EPs using CBT, with whom, and to what effect?

CBT was used within each of the five EP functions – assessment, consultation, intervention, research and training (Scottish Executive, 2002). As well as using it directly as an intervention (Boden, 2020), EPs were also involved in training or supervising others although typically EPs delivered the intervention themselves. Rait et al. (2010) and Squires (2010) discussed about EPs having a potential role in upskilling and supporting school staff in the effective delivery of CBT interventions, enabling staff to apply the principles they have learnt with other CYP thus building capacity and potentially being more cost-effective.

There were examples of interventions being delivered at universal, targeted and specialised levels, although the majority were to small groups, echoing the findings of Simpson and Atkinson (2021). A range of standardised CBT intervention programmes were used, with seven studies involving bespoke EP-developed interventions. Pugh (2010) discussed EPs’ potential to deliver personalised as well as manualized CBT interventions.

The range of interventions indicated a continuum of CBT interventions (Rait et al., 2010) and that EP use appeared to extend from traditional (Graham, 2005), as exemplified in Boden’s (2020) study, through to interventions focused on skill development (Clarke et al., 2017; Grimes, 2015). Interventions tended not to be broader, such as approaches involving parent training and solution focused brief therapy (Graham, 2005).

EPs worked with young people from age nine to post-16, working towards a range of outcomes. Most studies were with school-aged children, with only one post-16 example (Boden, 2020), suggesting greater scope for involvement with older students, but also perhaps a training need in this area (Atkinson et al., 2015). CBT was predominantly used to support CYP with anxiety, reflecting the relevant evidence base (Fuggle et al., 2013; Van Steensel & Bögels, 2015). CBT was also used to support CYP on the autism spectrum in three studies, consistent with emerging evidence (Hillman et al., 2020). These findings suggest that EPs are adapting CBT interventions for CYP from specialist populations (Squires, 2010).

The wide range of techniques and methods encompassed within CBT (Corrie & Lane, 2021) and demonstrated within this study make it difficult to draw conclusions about its effectiveness, although notably, all studies reported some overall positive effect. Consistent with evidence that CBT is effective for CYP with anxiety (Fuggle et al., 2013; Neil & Christensen, 2009), the predominant positive impact was in this area. Where CBT was used to support social skills and anger/behaviour, the impact was less strong and CBT was not found to have a positive impact on sensory needs or metacognition. Given the limited evidence for the effectiveness of CBT for these difficulties, EPs may need to be clear in their rationale for using it, through practice-based evidence (cf Frederickson, 2002) or publishing the outcomes of EP casework (Simpson & Atkinson, 2021).

Some studies reported positive qualitative outcomes even if these were not necessarily reflected within quantitative measures, highlighting the importance of practitioners considering how to evaluate interventions (Dunsmuir & Hardy, 2016), perhaps by including affective measures (such as ratings), systematic feedback and behavioural outcomes.

What components of CBT are most prominent within EPs’ practice?

Squires (2010) discussed how EPs have a range of underlying CBT skills and competencies which they deliver as part of everyday practice, some of which were evident within the studies. Given the importance of ethical practice in EP therapeutic work (Dunsmuir & Hardy, 2016) it was perhaps unsurprising that it was a prominent competency within the research. Child-centred practice was particularly apparent, through EPs’ choice of developmentally appropriate materials, reflecting Squires’(2010) observation that EPs have the knowledge of child development to enable them to adapt to CBT interventions.

Fuggle et al. (2013) referred to the importance of considering both family and school factors when working with CYP. Given that good practice guidance for CBT (Maddox et al., 2021) recognises the importance of involving parents in formulation and intervention, it is notable that only three interventions involved families. Given reported difficulties engaging parents of younger children in parent-led CBT interventions (Ellins et al., 2021), there appears to be a dearth of the literature about how EPs can effectively involve parents in school- based CBT interventions.

Three of the Fuggle et al. (2013) competencies, evident in three or less studies – being goal-focused, providing a rationale and seeking feedback – related to collaborative practice; while only four studies involved CYP in joint session planning. This is particularly significant given that collaboration is a core principle of CBT (Corrie & Lane, 2021; Fuggle et al., 2013). Offering some potential insight, Fuggle et al. (2013) reflected on the challenges of building collaborative practice with CYP due to factors such as the “power imbalances” (p. 121) inherent in relationships between children and adults. It is possible that these issues are even more evident within a school, than in a clinic setting, where the impetus for referral may be more likely to have come from a family member.

Summarisation, which involves discussing goals and supporting a joint formulation (Fuggle et al., 2013), was not evident in any of the studies. The development of a joint formulation was only evident in three research studies. Interventions seemed more focused on psychoeducation, teaching CYP thinking skills and problem-solving skills. Seiler (2008) discussed how the main distinction between CBT and cognitive behavioural interventions was that CBT involves developing a shared formulation with the client about the areas of difficulty, which is then used to inform treatment that will promote positive change. A key factor, given the focus on knowledge transfer, rather than collaboration, is likely to be the fact that only two of the studies described work at the individual level, meaning opportunities for therapeutic partnership may have been limited.

One element demonstrated comprehensively within the research studies was monitoring and evaluating progress, with EPs assessing impact using a range of methods, although it should be noted that all of the interventions took place within the context of research studies, where data collection would be conventional. However, practice did not reflect ongoing monitoring, which is important in order to ensure that the intervention is having an ongoing positive effect (Dunsmuir & Hardy, 2016).

Limitations and opportunities for future research

The lack of control groups within the studies makes it difficult to ascertain whether CBT was the “active ingredient” or whether progress was due to other factors, given that Neil and Christensen (2009) found non-CBT-based interventions as effective as CBT-based interventions for CYP with anxiety. However, as Dunsmuir and Hardy (2016) discussed, it is important for EPs to balance between randomised control trials and the insights into real-world research. The mixed methodology employed by the majority of studies provides understanding about how CBT interventions were experienced, as well as individual and systemic level factors which impacted on outcomes but made it difficult to aggregate results. Future research could analyse factors that impact on CBT intervention implementation in school settings at different levels (cf Durlak & DuPre, 2008). The heterogeneity of CBT interventions used is also a limitation with regard to drawing definite conclusions about the use of CBT in EP practice.

As Simpson and Atkinson (2021) also discussed, it is difficult to draw conclusions about long-term impact due to limited follow-up or longitudinal research within this review. This highlights the importance of EPs planning research that evaluates the long-term impact of CBT interventions. This is particularly important given that research suggests that the positive outcomes from school-based interventions for anxiety and depression, the majority of which involved CBT, might not be sustained beyond sixmonths (Gee et al., 2020).

Future systematic literature reviews could also address limitations inherent within the exclusionary criteria, perhaps by widening the scope to cognitive behavioural interventions, and third generational models of CBT which may have wider ranging applications such as supporting staff wellbeing in schools (Gillard et al., 2021). The exclusion of international literature also meant that potential applications of CBT may have been missed within the practice of school and educational psychologists.

WoE B, using the Fuggle et al. (2013) criteria, enabled the analysis of key components of CBT, prominent in EP practice. However, as Simpson and Atkinson (2021) discussed in a similar study, competencies may be under-reported, given the brevity of research articles. Conversely, without fidelity measures, the described competencies are based on self- report and may be overstated. Finally, the Fuggle et al. (2013) framework was developed for individual work, and competencies for group or whole class work, at targeted or universal levels, may require different benchmarks for effectiveness.

Conclusions and implications for EP practice

Overall, the literature review suggests that EPs should be mindful of the evidence base for using CBT with CYP, including its particular effectiveness with anxiety, and more limited outcomes for supporting CYP with other difficulties. If EPs are using CBT, potentially looking to understand processes and mechanisms for change within a practice-based evidence approach (Dunsmuir & Hardy, 2016), may help illuminate further potential advantages and limitations of using CBT within casework, and lead to a more cautious, considered and responsive intervention, especially given the limited research available of EPs using CBT in practice. The evident heterogeneity of the CBT research in EP practice is a limitation of this literature review which makes it difficult to make definite conclusions.

Despite these issues, this study highlights that CBT is an intervention approach that EPs can use to support positive outcomes for CYP, particularly anxiety; however, other therapeutic interventions and modalities have been found to be effective in other available literature. Although therapeutic practice can be considered along a continuum, the research suggests that EPs are predominantly involved at the intervention level rather than using CBT as a specialised therapeutic approach.

There remains a relative dearth of literature regarding EPs’ use of CBT interventions with CYP, particularly in relation to individual, specialised interventions, and also universal approaches. This poses challenges for EPs when making informed, evidence-based decisions. This review highlights a limited but growing body of research evidence into how EPs can upskill and support school staff to implement CBT intervention, either through training or supervision. Echoing the recommendations of Simpson and Atkinson (2021), practitioners using CBT interventions could contribute to the dearth of literature by publishing practice outcomes.

Future research could place particular emphasis on understanding implementation factors, perhaps using in-depth case study and mixed-method designs. Research into operationalisation as well as outcomes will support EPs and schools to make informed choices about when and when not to use CBT interventions.

Findings also indicate that EPs have underlying competencies to deliver CBT interventions for CYP, although further training and skill development may be needed to ensure delivery in line with good practice guidelines (Maddox et al., 2021; Squires, 2010). Finally, this research highlights the limited collaborative practice in EPs’ delivery of CBT interventions, perhaps confounded by the prevalence of group or whole-class interventions. However, even for interventions delivered at these levels, attention to Fuggle et al. (2013) might promote effective practice.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This project was funded through England’s Department for Education (DfE) National College for Teaching and Learning (NCTL) ITEP award 2017-2020