Evidence Based CBT
29 Jun 2021

Dissertation Findings on Evidence Based CBT in Child and Adolescent Mental Health Services

ICHAS Student Dr. Hazel Fernandes recently completed her dissertation on evidence-based CBT in child and adolescent mental health services. Her main focus was through the lens of therapeutic alliance and the current trends and future challenges in this area. Here are some of her key findings. 


Reflective practice has become an integral part of the process on the journey to professional and personal growth and development. In thinking about what we do, rather than just doing what we do, that we can recalibrate what we do, in order to deliver the best of an acceptable practice to our clients and through that achieve a myriad of benefits to increase personal and professional competence (Schon 1983). This sense of competence can be affected by variables such as treatment outcomes of patients, amongst others (Ginzburg et al.,2012). 

However, it has also been demonstrated that competence has no significant effect on the outcome of treatment (Webb et al,2010). In Psychotherapeutic settings, the relationship between the clinician and the client is essential for effective practice and although complex, is an important factor.

Outcomes in therapy have been consistently related to therapeutic alliance (Arnow and Steidtmann,2014). Bachelor A. (2013), concluded that the client’s view of the therapeutic process significantly influenced treatment outcome and recommended seeking client feedback to achieve superior treatment outcome. 

This relationship progresses on a continuum, and like any relationship, there is progress either positively or negatively. When the client seeks therapy and meets with the therapist, the seed of a relationship is sown. As with anything with growth potential, the conditions need to be optimised, and in therapy, both therapist and client need to collaborate, innovate and re-evaluate goals and find common ground to allow for the achievement of goals agreed at the beginning of therapy.

Cognitive Behavioural Therapy (CBT), a collaborative intervention therapy (Easterbrook and Meehan, 2017), is traditionally seen as manual-based and appeared to allow little opportunity for a therapeutic relationship to develop. It was the view that this lack of attention to the relationship tended to cause CBT therapists to lose interest in it (Sanders and Wills, 1999).

Whilst others reported that too much emphasis was placed on “the technical aspects of therapy, rather than focussing on the client’s perception of the “relationship with the therapist” (Duncan et al., 2010), and it would be detrimental to ignore the insight that can be attained through observation and interpretation of the way the client and therapist interact (Leahy, 2008).

It would be equally wrong to interpret any therapy from the confines of a single variable and one of those variables that define a therapeutic relationship is “Therapeutic alliance”. Wolfe & Goldfried, (1988), called therapeutic alliance “the quintessential integrative variable” of therapy. 

Norcross, (2010), defined the therapeutic alliance as referring to “the quality and strength of the collaborative relationship between client and therapist, typically measured as agreement on the therapeutic goals, consensus on treatment tasks, and a relationship bond.”



Literature reviews are completed mostly in the context of conducting a research study, or prior to commencing a research study and this can lead to a somewhat disorganised review due to constraints of time and the quality and agreement on the issues discussed can raise potential issues of trustworthiness (Snyder, H.,2019.

A sound basis for research can be constructed through a robust literature review and can be achieved through the process of reviewing previous research in that particular area (Baumeister & Leary, 1997; Webster & Watson, 2002). 

Therefore, the extended literature review as a methodology can help identify vital methodological processes and derivatives within existing studies, not with the intent of “derailing” existing research but to go back to the “grassroots” so to speak. 

Having considered the systematic review, my focus being the understanding of the concept of Therapeutic Alliance led me to choose the integrative review. The integrative method will allow her to assess, critique and synthesize the literature to allow for a new theoretical framework to emerge (Torraco,2005).





There was a significant relationship between therapist-rated alliance and symptom reduction in that the strength of the ratings of all three raters: child, parent, and therapist, directly related to the decrease in symptomatology. However, there was a significant observed decrease in parent rating of therapeutic alliance towards the latter part of the intervention; the exposure work did not impact on the ability of the therapeutic alliance measure to predict treatment outcome (Keeley et al., 2011).

Anxiety Disorder

High alliance rates and treatment adherence was found with completion rates of more than 80% in the online CBT intervention (Carpenter et al, 2018). 

There was no difference in online CBT alliance ratings and face-to-face CBT alliance ratings of both parent and child in the study by Anderson et al.,(2012), but the parent-rated alliance was significantly higher than their counterparts in the face-to-face CBT.

Furthermore, although the anxiety symptoms improved significantly, these changes were not predicted by the working alliance even though the working alliance predicted youth alliance significantly. 

Compliance in the treatment did not significantly predict the outcome but parents and youth alliance were significantly related (Anderson et al., 2012).

There was a significant association between parent, child, and therapist ratings of alliance and youth reporting higher rates of supportive relationships, established stronger alliances (Levin et al, 2012).

Anxiety treatment outcome and alliance

Therapists and mother’s alliance ratings had a predictive effect on improvement in anxiety symptoms, in contrast, father’s and child’s ratings of alliance did not predict improvement in anxiety symptoms. In view of this finding, it was surprising to find that with respect to reverse causality, it was the father’s and therapist’s alliance ratings that were predicted by the reduction in anxiety symptoms and not the child’s and mother’s (Marker et al, 2013).

Hudson et al., (2013), found that therapist involvement directly influences child alliance and involvement and teachers observed a decrease in symptoms when therapeutic alliance improved.

But no significant differences were found between therapist and child-rated alliance scores and this did not significantly predict treatment outcome (Zandberg et al., 2015).

Depressive disorder

Early therapeutic alliance significantly influenced the latter severity of depressive symptoms, but the reverse was not true (Labouliere et al, 2017). 

Pre-treatment severity of depressive symptoms led to stronger therapeutic alliances as rated by independent observers, but the adolescents having worse depressive symptoms than anxiety symptoms were perceived as having weaker early alliances by the therapists (Levin et al., 2012).

Trauma and PTSD

Girls built early alliances more easily than boys and it was the strength of the alliance that led to greater improvement in internalizing symptoms than externalising symptoms, although both improved and the symptom change was not mediated by the change in alliance (Zorzella et al., 2015).

There was a significant positive progression in both the therapist and child ratings throughout therapy with the child’s ratings changing significantly from session 3 and session 8 to the final session (Zorzella et al., 2017).

Treatment expectancy and collaboration did not influence treatment outcome although it influenced the working alliance but parent/ carer working alliance affected treatment outcome (Kirsch et al., 2018).

The carers alliance rating was the highest in comparison to the child and therapist rating and although concordance was low between the therapist and carer with respect to ratings, alliance contributed to a reduction in symptom severity (Loos et al.,2020).


The results showed that although alliance was significantly higher in those engaged in CBT- PML (Structured CBT), it did not translate into better treatment outcome and did not play a role in symptom improvement (Boyer et al., 2018).

Eating Disorder

The study showed that the internal consistency of the ACF (psychometric scale being evaluated) was high and therapeutic adherence and alliance was excellent with alliance client having a positive impact on loss of control (bingeing) of patients but not on other variables. These include other BED disorder symptoms, mood symptoms, and treatment expectations (Puls et al., 2019).


Child-therapist alliance was a significant predictor of post-treatment anxiety scores, and parent-therapist alliance was only a marginally significant predictor (Klebanoff et al., 2019). 

In ASD, the child’s intelligence quotient positively affected task collaboration late in therapy and this was directly related to emotional regulation, child emotional lability, externalizing symptoms and therapist rated the severity of Autism, which predicted emotion regulation (Albaum et al., 2020).

Developmental age and gender

An increase in alliance influenced involvement in therapy and was significantly directly proportional to age, the older the child, the stronger was the involvement (McLeod et al, 2014).

In older youths there was a significant and positive relationship between parent alliance and treatment outcome, in that stronger parent alliance led to better treatment outcomes from pre-treatment to 6 months post-treatment but this was not significant for younger youths (Anderson et al., 2012).

Girls were found to establish stronger alliances early on than boys in trauma therapy (Zorzella et al., 2015), however, another study found that neither age nor gender had a significant effect on alliance in anxiety disorder (Chu et al., 2014).

Female gender predicted better treatment response to online intervention in anxiety disorders (Stjerneklar et al., 2019).

Interpersonal relationships

The stronger the interpersonal relationships and support they received from these relationships, the stronger was the alliance in children with anxiety and depressive disorder (Levin et al., 2012).


It was observed that the stronger the parent–therapist alliance was, the better the treatment outcome in OCD (Keeley et al., 2011).

In older youths but not in younger youths, there was a significant and positive relationship between parent alliance and change in symptom severity (Anderson et al., 2012).

Internet-based CBT

The compliance with agreed treatment goals was not found to be adversely affected by using online intervention and in fact was quite the contrary, in that more than 80% completion rate was achieved (Carpenter et al., 2018).

In the study that compared face-to-face and online intervention of the same treatment modality, it was found that there was no difference between the two in terms of alliance ratings. However, the ratings of parents attending the clinic were significantly higher than the parents of children attending online CBT where there was no face to face contact and minimal therapist contact. 

The study found that the age of the child (older the child) was a factor in the outcome being predicted by stronger alliance (Anderson et al., 2012).

Higher levels of computer comfortability were associated with increased treatment response as were self- and clinician-rated severity of depressive symptoms (Stjerneklar et al., 2019).

online cognitive behavioural therapy

Conclusion and Recommendations on Evidence-Based CBT

This review was undertaken to evaluate the current state of Research in Child and Adolescent Mental Health in the context of evaluating Therapeutic Alliance when using Cognitive Behavioural Therapy as an intervention.

The aims were manifold, one of which was to evaluate the influence of developmental age of a child and social constructs of the child’s own relationship in therapeutic alliance.

I was really interested in understanding the importance of parent alliance not only with the therapist but also if the alliance with their own children in many interventions, particularly in those of younger age affected a change in the treatment outcome of many mental health disorders.

I was also interested in understanding if the rating scales used to evaluate Therapeutic Alliance were adequate to capture the three pillars of Therapeutic Alliance:

  • Congruence
  • Empathy
  • Unconditional positive regard,

As posited by Rogers,1951.

Undertaking the Master’s Course in Cognitive Behavioural Therapy fuelled my interest in alliance studies, particularly due to the course content where reflective accounts were deservedly valued as essential in enhancing psychotherapeutic skills and therapeutic alliance being one of the variables that can inform treatment outcomes which in turn serve as a compass to assess gains thus assisting therapeutic interventions and improving therapist skills and confidence, thus this review fulfilled many of my learning objectives.

Various obstacles encountered during this process ranged from choosing the methodology to acquiring publications for review all with time constraints whilst working full time made this review quite a laborious task but one when finished, enriched knowledge. Maintaining objectivity of the review and resisting the introduction of subjectivity without making it overly a commentary but also injecting it with pertinent subjective learning outcomes was difficult to achieve.

What is essential to observe about the current state of research in this field, is that small steps are being undertaken, questions are being asked and answers are provided with no fear of being judged.

Although all the questions can never be answered and future research could unravel all previous theoretical frameworks, what is undeniable is that the studies discussed in this review will transform and enhance our understanding of our ability to transform child mental health outcomes more positively.

Even though many of the studies were underpowered, the results could help us, especially when seeing patients whose presentation may remind us of a cohort in this review and help guide our engagement in view of study outcomes.

The review observes that working with parents to collaborate and engage them when treating the more severely affected children, to the awareness that perhaps engaging the male gender in interventions is helpful even if the statistical power of these findings is low.

The knowledge that Internet-based therapy should not only be reserved for mild severity of illness but can also work in severe mental illness especially in debilitating social anxiety is a welcome finding from this review, particularly during the current pandemic. This comes with the caveat that choosing patients for internet-based intervention, one needs to ensure good internet connectivity as it would be self-defeating to allocate internet-based CBT to a patient with no or poor connectivity.

This review continues to espouse the individualized treatment plans for all patients through the authors’ acceptance that larger studies need to be conducted before their findings are generalized.

There is however one aspect of child mental health that was excluded as a variable in all the studies and that is “suicidal ideation” or “suicidality” in children with mental health issues. Another aspect of possible variables which could predict Therapeutic alliance is the use of Medication which needs further research.

Including these two variables in future research studies would help clinicians working with children in preventing adverse outcomes and inform prescribing of medication in children.

It would really help to include this symptom in future studies to address this gap in knowledge.

To summarise, World Health Organization (WHO) defines mental health “as a state of well‐being in which the individual realizes his or her abilities, can cope with the normal stresses of life, can work productively and fruitfully, and is able to make a contribution to his or her community” (World Health Organization, 2005).

It, therefore, stands to reason that both Therapeutic Alliance and mental health should be viewed over time rather than as “a snapshot” of a point in time (Kendall et al, 2009). It is probably only fitting to view Therapy as a ship with its’ passengers sailing in a “Sea of Therapeutic alliance”, which ebbs and flows like the tide but eventually reaches the shores with patient-carer-therapist navigating the course.

About Dr. Fernandes, Consultant Child & Adolescent Psychiatrist

Qualified from the oldest Medical College in Asia, The Goa Medical College, in Goa, she completed her postgraduate specialising in Obstetrics & Gynaecology. In pursuit of further developing her skills and she continued her clinical practice in Obstetrics & Gynaecology working in the major Hospitals in Dublin. However, following the birth of my son, she decided to switch to Psychiatry, completing her Basic Psychiatry Scheme through the Dublin University Rotational Scheme.

Child Psychiatry was a natural choice given her links to her previous specialty and therefore she chose to complete the Masters in Cognitive Behavioural therapy at ICHAS.

Dr. Fernandes is now pursuing new frontiers, having had the privilege of being appointed as a Consultant Child & Adolescent Psychiatrist, in an award-winning NHS Trust, Hertfordshire Partnership Foundation Trust, where she will continue to utilise the psychotherapeutic skills acquired through a very rich, well-supported programme of Masters in CBT through ICHAS.