Therapeutic Goal Types in Young People’s Mental Health Providers and Changes in Anxiety and Mood

Clinical goal setting is an important area of practice within mental health interventions for children and young people (CYP). The present research aimed to explore the type of intervention goals set by CYP during mental health interventions. Further, changes in goal progress from pre- to post-intervention were compared between externally observed behavioural goals and subjective feelings-based goals. The relationship between therapeutic goal achievement and changes in symptoms of depression and anxiety (measured using the Revised Children’s Anxiety and Depression Scale) was also investigated. Data were collected from 792 participants aged 3 to 18 years (mean age: 13.8, SD: 2.85) across all gender identities and ethnicities, who received low-intensity cognitive behavioural therapy interventions from Children’s Wellbeing Practitioners in the Southwest of England. A mixed methods approach involving conceptual content analysis, correlational analysis, and group comparisons was utilised. Eleven categories of child-rated goals were identified. The most commonly developed goals focused on understanding, managing, and expressing emotions and feelings. No significant differences were found between behavioural and feelings-based goals (p = 0.061). Partial correlations demonstrated that improvements in goal progress were significantly associated with reductions in depression and anxiety symptoms (rs > −0.157, p < 0.001), even when controlling for the number of sessions attended (rs > −0.146, p < 0.001). Integrating goal-based outcomes with standardised measures could enhance a comprehensive approach to service delivery and evaluation. The most commonly generated treatment goals focussed on understanding, managing and expressing emotions and feelings, followed by doing better at school. The majority of children and young people (75%) showed reliable improvement in their therapeutic goals from pre-to-post CBT intervention. Therapeutic goal achievement showed a significant negative association with symptoms of anxiety and low mood, whilst controlling for number of sessions attended. Combining goal-based outcomes with standardised measures promotes an individualised approach to intervention, whilst facilitating wider therapeutic outcomes. The most commonly generated treatment goals focussed on understanding, managing and expressing emotions and feelings, followed by doing better at school. The majority of children and young people (75%) showed reliable improvement in their therapeutic goals from pre-to-post CBT intervention. Therapeutic goal achievement showed a significant negative association with symptoms of anxiety and low mood, whilst controlling for number of sessions attended. Combining goal-based outcomes with standardised measures promotes an individualised approach to intervention, whilst facilitating wider therapeutic outcomes.

The majority of children and young people (75%) showed reliable improvement in their therapeutic goals from pre-topost CBT intervention.
• Therapeutic goal achievement showed a significant negative association with symptoms of anxiety and low mood, whilst controlling for number of sessions attended.

•
Combining goal-based outcomes with standardised measures promotes an individualised approach to intervention, whilst facilitating wider therapeutic outcomes.
Historically, the evaluation of service-user progress in psychotherapy has predominantly relied on nomothetic patient reported outcome measures (PROMs).Nomothetic PROMS contain standardised items which are based on population norms and focus on symptoms of mental health.Questionnaires are usually supported by strong psychometric evidence, are concise, well received by clients, and contain items that encompass a wide range of challenges and experiences (Sales et al., 2023).A commonly used nomothetic PROM assessing anxiety and low mood within youth mental health services is the Revised Child Anxiety and Depression Scale (RCADS; Chorpita et al., 2000).Nomothetic PROMS are useful in making comparisons at the population level and are centred around the medical mental health model, which emphasises the treatment or management of symptoms of a disease (Yanos et al., 2008).However, although nomothetic PROMS are valuable, it is widely accepted that every individual receiving psychological intervention possesses an individualised set of attributes, difficulties, strengths, individual needs, and contextual circumstances (Cooper & McLeod, (2011)).In recent years within children and young people's (CYPs) mental health research, idiographic patient reported outcome measures (I-PROM) have been advocated alongside standardised symptom-based instruments (Jacob et al., 2023).I-PROMS are individualised measures that allow clients to actively participate in their own assessment criteria, formulating specific areas in which they require support and wish to improve (Sales, 2017).They recognise and focus on the importance of individual experience and are rooted in the understanding that "recovery" is complex and individualised, which cannot be minimised to predefined syndromes measured using symptom-based checklists.The Goal-Based Outcome (GBO) tool is an example of an I-PROM commonly used to facilitate shared decision making in CYPs mental health settings (Sales et al., 2023).Goal setting enhances the therapeutic relationship (Martin & Feltham, 2020), helps young people to feel more included in their care (Abrines-Jaume et al., 2016) and reduces non-attendance rates (Cairns et al., 2019).This paper aims to provide an overview of goal setting within CYPs mental health services.
GBOs aim to evaluate progress towards objectives that young people have chosen to work towards in collaboration with their practitioner (Law & Jacob, 2013).The basis of this method stems from Goal Attainment Scaling, which was developed four decades ago to address the need for personalised evaluation in mental health services (Cytrynbaum et al., 1979).GBOs are often created at the beginning of intervention as part of a collaborative process between the young person/parent and the practitioner.By using the GBO tool, practitioners can engage in meaningful discussions with young people, formulate goals together, and track progress over time.This process can facilitate further goal-focused conversations (Law, 2019).Although goals are highly individualised, previous research has explored the commonalities between goal types that CYP set themselves during mental health interventions.For example, in 441 goals set by 180 CYP across eight mental health services in England, the most commonly set goals focused on coping with specific difficulties, personal growth and independence (Jacob et al., 2016).In a similar study of 241 goals created by 80 CYP from England Child and Adolescent Mental Health Services (CAMHS), the top five goals were related to personal growth, functioning, and coping with specific symptoms and problems (Bradley et al., 2013).These studies, however, did not explore the relationship between goals and presenting difficulties and were conducted in CYP from specialist CAMHS settings, including those with more complex presenting difficulties.To our knowledge, previous research has not evaluated goal setting in CYP with mild to moderate mental health difficulties.Goal setting within these populations may differ due to differences in context and presenting severity.As such, in order for practitioners to better understand how to formulate effective goals within the context of mild to moderate mental health difficulties, it is important to evaluate CYPs collaboratively agreed intervention goals.
Setting precise, challenging, and measurable goals is considered "best practice" (Locke & Latham, 2013).Practitioners are often trained to follow the SMART framework when guiding therapeutic goal formulation (Lawlor & Hornyak, 2012), making goals; Specific, Measurable, Achievable, Realistic and Timed (SMART).In a study exploring 187 goals across two young people's mental health services in Australia, 95 goals (57%) were specific, 23 (14%) measurable, but none were timed (Cairns, Kavanagh, Dark & McPhail, 2019).This highlights that goals can vary drastically in terms of how specific they are and how clearly they can be tracked over time.However, their study did not evaluate whether goals that were more measurable may have been achieved to a greater degree.As such, one fundamental research question remains unanswered: Are goal types that focus on externally observable behaviour (e.g., to take part in classroom conversation by putting hand up three times), and therefore more measurable in nature, more attainable than more subjective (and less easily measurable) feelings-based goals (e.g., to be less sad)?
Goal setting within CYP mental health services has been found to positively influence emotional well-being.For example, young people who have accessed mental health services report positive emotions following therapeutic goal achievement, with many experiencing increased positive affect, productivity, a sense of hope, and a positive view of the future (Penno, Hetrick & Christie, 2022).These findings are in line with motivational theory (Lunenberg, 2011), such that goal setting builds on self-efficacy, selfdetermination and motivation, acting as a self-regulation strategy (Harkin et al., 2016).Goals appear to latently capture more 'existential' and 'distal' factors, such as confidence, resilience and understanding (Batty et al., 2013), however, goals are often formulated within the context of well-being (Jacob et al., 2016).As such, there is evidence that goal objectives often overlap with items from nomothetic PROMS (Weisz et al., 2011).When comparing therapeutic goals across eight CAMHS, 20 out of 27 goal themes were congruent with items on at least one commonly used standardised outcome measure (Jacob, Edbrooke-Childs, Law & Wolpert, 2016).This suggests that therapeutic goals appear to draw upon similar processes as standardised measures.Further, research demonstrates a significant association between goal progress and nomothetic PROMS.For example, in a sample of 137 young people from England CAMHs, significant associations were highlighted between progress towards goals (as measured using the GBO tool) and psychosocial difficulties, functioning, and impact on daily life (Edbrooke-Childs et al., (2015)).Taken together, these findings suggest that the GBO tool should be used to complement standardised measures of mental health, providing a more integrative approach towards service delivery and evaluation.The current study aims to explore the associations between goal progress and changes in symptoms of anxiety and low mood.

Context
Early intervention, in the United Kingdom (UK), is increasingly seen as a central provision for children, young people, and families to build resilience and minimise physical, social, and educational inequalities into adulthood (Department of Health and Social Care and Department for Education, 2017; NHS, 2019).The UK government expanded the provision of evidenced-based interventions, aiming to reach 70,000 more CYP in England by 2020 (NHS England, 2016).This led to the establishment of the Wellbeing Practitioner for Children and Young People (CWP) programme as part of the CYPs Improving Access to Psychological Therapies initiative (CYP-IAPT; Department of Health, 2011).
The CWP workforce expansion programme was established in 2017 to help improve access to early intervention for the most common mental health difficulties experienced by CYP, including anxiety, low mood and behavioural difficulties (Ludlow et al., 2020)

Aims
The aims of the current study were to: 1) Explore therapeutic goals set by children and young people within Children's Wellbeing Practitioner services.
2) Compare subjective feelings-based goals with more objective behavioural goals in terms of achievement.
3) Explore associations between therapeutic goal achievement and changes in symptoms of anxiety and depression.

Method
Judges Two researchers (one male and one female) from the University of Exeter served as judges in this study.Both authors developed the coding rule document, and both judges conducted the analysis.The classification system was refined by the whole research team (consisting of a Postdoctoral Research Fellow, a Senior Lecturer and a Professor) via video conferencing.After discussing the classification system and reviewing coding examples developed by both authors, each researcher separately evaluated all of the goal descriptions.Disagreements were discussed by both researchers until consensus was reached.Cohen's kappa was calculated to evaluate interrater agreement for the analysis.Results showed a satisfactory consistency of 0.95.

Design and Data Collection
This retrospective cohort study design was conducted as part of a routine service evaluation between June 2017 and December 2019.Data from 20 children's mental health services in the Southwest of England was collected as part of routine practice using a secure online database designed for community-based mental health services.CYP engaged with routine outcome measures aided by a low intensity trained Children's Well-being Practitioner (CWP).CWPs underwent a year-long postgraduate certificate at the University of Exeter.All trainees received the same level of training on setting SMART goals with CYP, specific to obtaining GBOs, delivered by a course lecturer experienced in clinical practice.Measures were collected at the first and last point of contact with the CYP, with session-by-session measurements collected for symptom-specific (e.g., RCADS-Child) and idiographic measures (e.g., GBO).Goal progress was collected by the CWP on paper at the beginning of each session by using the script provided on the paper based GBO form.CYP rated their goal progress in front of the practitioner, and parent (if present).Data was exported and checked for missing values.Cases without post-intervention data were excluded before analysis.11 of the 20 services used an alternative means of data entry for which goal description was not collected; these services were therefore excluded.Due to limitations with the patient management system used by the included services, only the goal description of the young person's first goal was recorded and subsequently included in the analysis.

Participants
Of the 4098 CYP accepted within CWP services, 792 CYP were included in the analyses (see Fig. 1 for full procedure of inclusion/exclusion criteria).Demographic characteristics of CYP included in the study are presented in Table 1.Overall, 67.4% of the sample were female, with a mean age of 13.8 (median 14, SD 2.85, range: 3-18).The majority of CYP were referred for anxiety (43%), followed by low mood (17%), behavioural difficulties (5%), and other reasons (2%), with 32% of referral data missing.

Goal-Based Outcomes (GBO)
The GBO tool (Law & Jacob, 2013) allows self-determined goals to be formulated at the start of intervention, collaboratively set by the CYP, parent/carer and practitioner.The internal consistency of goal progress measured using the GBO tool has been found to be 'acceptable', with a Cronbach's alpha of 0.71 and 0.73 for pre-and post-intervention respectively (Edbrooke-Childs et al., (2015).In this study, the majority of goals were set by the CYP (55%), with the rest set by the practitioner (13%), the parent/carer (4%), or under mutual agreement (6%), with 22% not specified.This refers to the primary creator of the goal for data entry  purposes, whereas in reality, all goals included in this analysis were collaboratively agreed between CYP, practitioner and parent (if present).Goals included in this analysis were all focused on the child/young person.Goal progress was rated on a scale from 0 (not met at all) to 10 (goal achieved).The difference in GBO score between the first and last intervention session was calculated to measure each CYP's goal progress.

Procedure
A deductive content analysis (i.e., exploring the existence or frequency of concepts in a text; Christie, 2007) was conducted in the qualitative GBO description data.This involved analysing textual data using a-priori categories taken from Jacob et al., (2016), who conducted a content analysis of intervention goals in a similar population of CYP aged between 4 and 17 years.However, categorisation remained flexible to allow for the development of new concepts or redefinition of existing categories throughout the analysis.Categories were peer reviewed ahead of finalisation and subsequent coding.In total, 11 categories were determined including one category for goals with unclear meaning.A coding rules document was developed, as a reference for coding, especially for incongruent codes.During the analysis, two categories were redefined due to a high level of overlap between symptoms and emotions.Consequently, the categories were distinctly separated into internalised feelings and externalised behaviours.Data were coded separately using a previously created conceptual framework (see Jacob et al., 2016) and was based on implicit rather than latent concepts (i.e., context is derived from the visible and literal meaning of the words-taken at face value, rather than applying a deeper, interpretive analysis that seeks to infer underlying meaning from the words or phrases).That is, the coding aimed to objectively focus on what the data presented, rather than imposing a subjective view based on personal opinions, experience or expectation.Any disagreements about the categorisation of goals were reviewed and jointly recoded to ensure all goals were appropriately categorised.Ambiguous goals or goals that did not fit into any of the predetermined categories were allocated to the 'meaning unclear' category.Goals with numerous aims were coded into the 'multiple goals' category.

Analyses
Prior to analyses, outcome variables were examined for outliers, missing values and violations of the normality assumption.Descriptive statistics were conducted to examine the average pre-and post-intervention GBO score, and the average GBO change score.To determine change over time, the principles of the Reliable Change Index (RCI; Jacobson & Truax, 1992) were used, resulting in reliable improvement rates.The Reliable Change Index calculation considers the reliability of standardised measures (i.e., internal consistency) and identifies changes that are not solely due to measurement error.In line with previous research, a change in goal progress of three points on the GBO tool from pre-to post-intervention was used as a proxy of reliable change (Jacob et al., 2021).A change in score of +3 or more was categorised as reliably improved, whereas a change in score of -3 or more was categorised as reliably deteriorated.Scores not meeting this threshold were determined as showing no change.Frequency statistics were generated to highlight the percentage of goals within each category of the content analysis.A Mann-Whitney U test examined group differences comparing objectively observable behaviour goals vs subjective feelings-based goals.A chi-square test of independence explored the association between primary reason for referral (low mood, anxiety and behavioural difficulties) and goal type, using categories generated from content analysis.Standardised residuals of ±2 were interpreted to explore the source of the difference.
The relationship between GBO change and RCADS-C change were explored through partial correlations, controlling for the number of sessions each CYP attended, calculated separately for each RCADS-C subscale.

Sample Characteristics
Chi-square tests of independence were conducted to compare demographic variables between those who completed GBO pre-and post-intervention and those who did not.
A significantly higher proportion of CYP without post-intervention GBO data had missing age data.There was a significant relationship between GBO completion and ethnicity (χ 2 = 29.014,df = 3, p < 0.001).A significantly higher proportion of CYP with non-paired GBO data were ethnic minorities.Mean and range of RCADS subscale scores from pre-and post-intervention are presented in Table 2.

Intervention Characteristics
Children and young people attended an average of 6.7 intervention sessions (SD: 3.04, range: 1-23).A range of different CBT-based interventions were delivered (see Table 3).

Goal Progress and Improvement
Results from descriptive analysis of GBO data (see Table 4) indicate improvement from pre-to post-intervention in GBO with an average change of 4.70 points.Reliable improvement was seen in 75% of CYP's GBOs, which was significantly higher than no change (24%) or reliable deterioration (1%; χ 2 = 675.630,df = 2, p < 0.001).

Conceptual Framework
A content analysis was conducted to examine the frequency of goal types.The frequency of goal themes is presented in Table 5.

Relationship Between Goal Type and Primary Reason for Referral
There was a significant association between primary reason for referral and goal type (χ 2 = 72.304,df = 18, p < 0.001).CYP referred for low mood reported a significantly higher frequency of 'personal and hobby-related goals', whereas CYP with behavioural difficulties were statistically more likely to report goals related to 'improving family relationships' and 'child-parent collaboration'.

Externally Observable Behaviour vs Subjective Feelingsbased Goals
Goals were grouped/categorised into externally observable behaviour vs. subjective feelings-based goals.Nine goal types were organised into one of the two overarching domains (see Table 5).Goals related to increasing confidence, independence, self-esteem, and self-organisation did not fit perfectly into one category, therefore, each goal was independently reviewed and allocated to the appropriate category.Two categories (child-parent collaboration goals & multiple goals) were omitted from analysis.Overall, there were a total of 512 goals related to externally observable behaviour, and 223 goals related to subjective feelings.A Mann-Whitney U test showed that GBO scores did not differ significantly for externally observable behaviour goals (Mdn = 5) when compared to subjective feelings-based goal types (Mdn 4; U = 55552.5,p = 0.061).

Goals Progress and Changes in RCADS-C
Table 6 highlights a correlation matrix between GBO change and changes in symptoms of depression and anxiety, as measured using change in RCADS-C scores for each subscale.There were weak but significant negative correlations between all RCADS-C subscales and GBO change scores.After controlling for the number of intervention sessions attended, the relationship between GBO change and RCADS-C change scores remained significant across all subscales.

Discussion
This paper is the first to evaluate GBO data routinely collected from CWP services in Southwest England.CYP showed a positive change in their goals after a course of treatment, in line with previous literature (Edbrooke-Childs et al., (2015), Jacob et al., 2020).This is evidenced by a greater average post-intervention GBO score relative to average pre-intervention scores, with an overall average GBO change of 4.70 in our sample.The reliable improvement rate was high in this study compared to previous research (Jacob et al., 2020), with 75% of CYP showing reliable improvement, 24% showing no change and only 1% showing reliable deterioration.The most commonly developed goals fit into the category of 'understanding, managing and expressing emotions and feelings', followed by 'doing better at school (including participation and behaviours)'.There was a significant   association between goal type and primary reason for referral, with low referrals more likely to generate 'personal or hobby-related goals' and those with behavioural difficulties being more likely to generate goals related to 'improving family relationships' 'child-parent collaboration goals'.Further categorisation showed no significant difference in GBO change scores with respect to externally observable behaviour versus subjective feelings-based goals.Improvement in GBO scores were associated with a significant decrease in child-reported symptoms of depression and anxiety.The majority of CYP made significant progress towards their goal.This indicates the CWP service is helping to address areas of change that are identified as being most important by the CYP themselves.GBOs allow the CYP to be active agents within their own treatment, by allowing them to independently, or in collaboration with practitioners and/or carers, generate the focal point of their therapy (Law & Jacob, 2013).A 'client-orientated' approach to intervention allows idiosyncratic factors of therapeutic change to be evaluated.These areas may not be assessed by other normative measures currently used to measure outcomes in children's mental health.For example, goals may latently capture more 'existential' and 'distal' factors, such as confidence, resilience, and understanding (Batty et al., 2013), which could be more congruent with the clinical reality of psychotherapeutic work (Edbrooke-Childs et al., (2015); Sales & Alves, 2016).
A large number of goals focused on understanding, managing and expressing emotions and feelings.This strengthens findings from Edbrooke-Childs et al., (2015) and is to be expected in a clinical sample in comparison to the general population (Bradley et al., 2013).This shows that CYP are particularly motivated to improve aspects of their mood and emotional state.The second most common type of goal was 'doing better at school (including participation and behaviours)', followed by 'behavioural management/cooperation', 'seeing friends and improving peer/ social relationships', and 'increase confidence, independence and self-esteem'.These findings are in line with those of Hawley and Weisz (2003), which showed poor schoolwork and disobedience at home and school to be the highest two identified problem areas by CYP, followed by trouble getting along with family members and not getting on with other children.The overlap between identified problems in Hawley and Weisz (2003)'s research and the treatment goals in the current study highlights the consistencies across similar populations of CYP.The goals generated are appropriately wide-ranging and cover the broad scope of individuals, presenting difficulties, and context of referrals within services providing care for CYP experiencing mental health difficulties (Bradley et al., 2013;Edbrooke-Childs et al., (2015)).
Goal types were associated with referral reason.Low mood referrals were significantly more likely to generate 'personal or hobby-related goals'.Behavioural theories of depression state when one becomes depressed, many of their behaviours function to avoid and escape aversive thoughts, feelings or external situations (Veale, 2008).Consequently, those with low mood or depression engage less frequently with rewarding or satisfying activities and receive less positive reinforcement than those without depression, further exacerbating their symptoms.In the current study, CYP with low mood may have been motivated, or collaboratively encouraged with the practitioner to reconnect with their hobbies, which might, in turn, lead to an increase in positive reinforcement in their lives.
In contrast, those identified as having behavioural difficulties were more likely to generate goals related to improving family relationships and child-parent collaboration goals.Research has shown behavioural difficulties to be directly associated with high parent-child conflict in adolescents (Marmorstein & Iacono (2004)).The results of this study indicated that improving family relationships is identified as being the highest priority area to address from the perspective of children primarily referred for behavioural difficulties, whereas parents of those with behavioural difficulties are particularly motivated to collaborate with their child and better understand their behaviour compared to other referral types.
Improvement in goal attainment was significant but weakly associated with a decrease in symptoms of low mood and anxiety, after controlling for the number of attended treatment sessions.In the current study, the majority of goals within the understanding managing and expressing emotions category were formulated within the context of improving mental health symptoms (e.g., "to be able to recognise and stop myself having negative thoughts that give me more anxiety").This may have confounded the association between goal progress and change in RCADS-C scores, however, post-hoc analyses indicated weaker but significant correlations after excluding this category.This suggests that irrespective of the goal's focus, improvement in goal attainment is significantly associated with a decrease in child reported symptoms of low mood and anxiety.These results are supported by research which shows an association between change in progress towards goals and a change in clinician-reported functioning over treatment (Edbrooke-Childs et al., (2015)).The current study highlights that achieving therapeutic goals correlates with improved mental health and well-being.However, the direction of causality is unclear.Achieving goals may lead to a reduction in symptoms of low mood and anxiety, whereas a reduction in symptoms of low mood and anxiety, resulting from intervention, may lead to young people being more able to achieve their goals.Goal achievement may be integral to treatment outcome, however, weak correlations suggest that other factors may largely be at play.

Strengths Limitations
These results should be considered in the context of several limitations.First, goal descriptions may have been abbreviated or altered by the therapist during data entry.Therewe cannot be certain the goal description recorded and submitted for analysis is a complete or entirely accurate representation of the actual wording of the goal set during intervention.The data was taken from the CWP programme, which should predominantly reflect the experience of CYP with mild to moderate presentations of anxiety and low mood, therefore, these results may not generalise to severe presentations, or other treatment programmes.Goal progress was rated by the CYP in front of the practitioner and parent (if present), therefore, responses may be subject to social desirability bias.A high proportion of our sample were white females aged between 13 and 18 years old, which may have biased the results and affected the reliability of the findings.In addition, the higher proportion of ethnic monitories with non-paired GBO data could be more system-related, as opposed to specifically relevant to any difficulties providing GBO data.This may be due to difficulties in recording data from specific centres, which may also be exemplified by high levels of missing age data.The goals generated may have been influenced by the therapy on offer, for example, behavioural activation may have caused those with low mood to generate goals related to reconnecting with and approaching positive and rewarding experiences, such as hobbies.As such, outcomes were uncontrolled, the difference in outcomes may have been due to the passing of time or extraneous variables that were present outside of therapy, rather than the specific effects of the psychological intervention.Goals were not always clear cut and did not always fit into categories, for example, unclear goals were missing the context surrounding the goal, therefore, categorisation was not possible.Furthermore, goals related to increasing confidence did not fit perfectly into externally observable behaviour or subjective feelings-based goals.These goals were independently reviewed and allocated to the appropriate category, depending on whether they were associated with a specific observable behaviour or more generally, an emotional internal state.In addition, there was a large amount of missing data in relation to goal description and pre-and post-outcome data.Consequently, we cannot be certain that these results reflect the experience of the majority of the CYP seen by these services, as literature suggests those who do not complete post-intervention measures have more severe problems at the outset (Stiles et al., 2003).However, this was beyond the scope of the current study.Lastly, this study only considered each CYP's primary short-term goal for analysis, whereas each CYP generated a maximum of three goals.It may be that by excluding the second and third goals, important information was lost, as goals may make sense within a wider network of goals.
This study also has various strengths.The mixed methods approach employed in this study allowed for a more indepth investigation into the goals that CYP set themselves during intervention.This provided a qualitative exploration of the type of goals being set, whilst quantitatively examining achievement rates and the association with low mood and anxiety.As data was recorded as part of a routine service evaluation, it was possible to relate both goal type and progress to demographic and referral information.There is a lack of literature exploring the types of goals being set by CYP within mental health services and examining any differences between referral types.This study shows that GBOs are helpful in assessing individual's motivations and focus and may contribute to the therapeutic process and progress.

Implications for Practice
This paper highlights the importance of goal setting within CYP mental health services.GBO monitoring may allow the therapist to better understand the context behind the CYP's presenting difficulties and can provide an opportunity to apply practice-based evidence alongside evidencebased practice.
Services could benefit from further improvements in the quality of routine data for children's mental health, whilst being consciously aware of potential areas of missing data.This is a significant component of providing the most effective care for CYP, creating a 'feedback loop', whereby data can be used at the service level, with practitioners being trained to set clear, measurable goals with the CYP.The progress of these can also be used at the clinician level, by reviewing goal progress alongside the CYP and thus, maintaining a virtuous cycle.
Services should be mindful of setting clear, directed and measurable goals.For example, there are difficulties in evaluating and understanding the focus of goals with multiple aspects and those without a clear context.This may also have negative impacts in terms of CYP goal progress and more generally, the overall performance of the service.

Future Research
Although behavioural goals were not achieved at a significantly higher degree than subjective feelings-based goals, this relationship approached statistical significance and requires further exploration.CWPs are trained to collaboratively work with the CYP to make goals objective and congruent with SMART criteria.There are few studies which focus on validating and testing SMART (Bexelius et al., 2018), however, there is a gap in the literature on SMART goals within the CYP mental health domain.As such, a goal assessment scale should be created, which can be used to both guide the development of intervention goals and to retrospectively evaluate goal description data within CYP mental health services.Such a measure could help practitioners collect GBO data in a more standardised way by making goals similar in terms of how specific, measurable, achievable, realistic and timed they are.The effectiveness of the SMART framework could also be evaluated.
These findings warrant the need for further research on CYP intervention goals.The current study identified negative correlations between intervention goal progress and changes in symptoms of low mood and anxiety, however, more in-depth analysis is needed for the purpose of developing a more comprehensive understanding of the nature of this relationship.
The GBO tool potentially facilitates feedback in areas specific to the individual, that may encourage motivation, engagement and a sense of achievement.However, further research needs to be carried out to investigate the underlying concepts that GBOs are measuring (e.g., resilience and understanding) to identify which factors may have the largest effect on intervention progress.

Concluding Comments
The CWP programme has demonstrated meaningful therapeutic progress through clear, measurable outcomes.However, clinicians may benefit from clear guidance on consistently setting SMART goals to reliably monitor goal progress.Although externally observable behaviour goals did not seem to affect outcomes in terms of goal achievement, this was approaching significance and therefore, requires further investigation.Goal progress was correlated with improved RCADS scores and may be an integral aspect of the therapeutic process.Whist the GBO tool weakly correlated with the RCADS, it may also be tapping into another important aspect of the therapeutic experience (i.e., more existential and distal factors, such as confidence, resilience and understanding).Measuring GBOs facilitates a feedback loop benefitting the service, the practitioner and the CYP.Combining idiographic and standardised measures may provide a more holistic approach towards service delivery and evaluation, allowing the CYP to have an active role within their therapy, whilst also demonstrating servicelevel impact.

Table 1
Demographic characteristics of CYP who completed GBO data before and after intervention Demographics Closed Cases with GBO Data pre and post-intervention (N = 792)

Table 2
Mean, standard deviation and range of RCADS-C subscale scores pre-and post-intervention

Table 5
Frequency of Goals Within Each Category of the Conceptual Framework