Goal formulation and tracking in child mental health settings: when is it more likely and is it associated with satisfaction with care?

The use of goal formulation and tracking has been implemented in mental health settings in the United Kingdom (UK) and North America for some time (e.g. Hopes and Expectations [69]; Goal Attainment Scaling (GAS) [46]; Top Problems [9]) and is seen as an important part of many evidence-based practices (e.g. [7, 17, 32]). However, other than general life goals (e.g. school attendance) set in accordance with personal characteristics (e.g. previous experiences, presenting difficulties and developmental stages; [23, 56, 57]), very little is known about the use of this approach with children in therapy. This includes what characteristics (including demographics such as age, gender, ethnicity or presenting problem) (Fig. 1) may determine the propensity to formulate goals and track goal progress and the potential impact doing so might have on experience of care received. In the present article, goals are specific outcomes a child, young person, or family wants to achieve in accessing mental health services [48]. Commonly set goals by children accessing services include coping with specific difficulties, personal growth, and independence, and commonly set goals by parents accessing services include managing specific difficulties their child has, parent-specific goals such as increased knowledge of their child’s difficulties, and improving self or life [44]. It has been suggested that goal formulation and tracking may be especially useful for service users with particular needs, where progress may not be expected in terms of symptom reduction [6], which may include children with learning disabilities or developmental difficulties.

There are a number of barriers to formulating and tracking goals in child mental health practice, including the child’s capacity to do so [29, 64]; involvement of parents who may have different experiences and perceptions to the child or their own emotional and behavioural difficulties [75]; lack of resources to record goals formally (or the pressure to do so) and training on setting and tracking meaningful goals [34, 61]; and safeguarding concerns around balancing the involvement of the child with their vulnerability [2, 71]. Some models of therapy (e.g. solution-focused therapy; [66]) use the ‘miracle question’ as a means to conceptualise what would be different if one’s difficulties magically disappeared. The ‘miracle question’ can be one way of helping to formulate goals [48]. However, there is no governance about how treatment goals should be set and tracked in child mental health practice. This means that very little is known about how practitioners help to formulate and track goals with children.

Adults accessing mental health support have been found to highly value working towards their own goals [50]. Parents of children attending UK child and adolescent mental health services (CAMHS) have advocated for more collaborative methods of outcome monitoring [54]. Practitioners and children report reviewing goals and tracking them as helpful to assisting communication, motivating children, and giving them ownership of their care [61]. Evidence also suggests that agreement amongst families and with practitioners about desired outcomes may in turn be associated with better outcomes [52] and reduced attrition [14, 70]. However, goals are still not widely routinely recorded and tracked, as demonstrated through the high proportion of missing data of this kind in routinely collected datasets (e.g. [73]).

Still, therapeutic goal setting in general is argued to be a way to focus all parties to work together towards reaching common, concrete and measureable objectives. Practitioners have also suggested that it is useful to track outcomes such as goal progress over time in order to monitor treatment progress [6, 61]. Setting and tracking goals with families has also been found to be a valuable way to gather further clinical information [27] and an explicit focus on patients’ goals has been identified as key to effective care [30]. An assessment process that incorporates goal formulation and tracking has been shown to help the practitioner to understand the patient’s difficulties in more depth and may also lead to positive alliance with the practitioner [3]. This kind of collaborative working is likely to affect the way parents and young people view and experience the care provided in terms of feeling listened to, feeling supported and feeling like the care addressed the issue they were concerned about. The impact of goal setting and tracking on experience of care has not yet been explored in detail; therefore, it has been chosen as a focus for the present research.

Experience of care is also a key dimension of healthcare quality. Groups of practitioners offering provision to children, young people and their families with mental health and wellbeing difficulties are striving to improve experience by providing more person-centred care and tailoring delivery to an individual’s needs [22, 43]. Experience of care has been shown to have good face validity and may be better understood by professionals than more complex outcome information related to symptomology [33, 59]. Ratings of satisfaction are also sometimes the only proxy indicator available to assess the quality of a mental health intervention [8] which allows the service users’ voice to help inform an evaluation of support received [33].

Satisfaction with treatment has been demonstrated to be a complex factor, which may not have strong links to child-reported clinical outcomes [33] but may be associated with treatment adherence, perhaps due to its close links to shared decision-making. While links between outcome and satisfaction have not been found from a child’s perspective, they have been found from a parent’s perspective [10]. We are also interested in parental satisfaction in the present research due to the central role parents play in their child’s care, particularly in terms of engagement and adherence to treatment. Moreover, the therapeutic context is particularly complex as it involves a multidisciplinary, systemic approach with a number of stakeholders, perspectives and priorities. Evidence suggests that children, their parents, and therapists may have different perceptions of presenting problems and reasons for attending therapy [11, 40, 45]. Parents may set different goals to children, [35, 39] such as the management of specific symptoms, behaviour management, improving self-care and goals for parents themselves to achieve [44], which raises the important question of who is the patient: the child or the parent?

Parental investment in and agreement of goals may be crucial. Parental satisfaction has been indicated to be a key part of therapy completion in child mental health [5] due to the integral role parents often take in the management of a child’s emotional and behavioural difficulties [62]. A parent’s experiences of the support received and their relationship with the practitioner have been shown to be associated with parental satisfaction with child mental health care [42, 62].

Further, much of the work done related to child mental health is family-centred work, e.g. systemic family therapy in the UK and family systems interventions in America [24, 53], and there has been a call to explore other types of outcome measures for use in this work [41]. Evidence has shown that aside from demographic characteristics, the majority of variance in parent- and child-rated satisfaction of mental health treatment remains unexplained. As such, satisfaction may be influenced by factors that are relatively unexplored, such as ‘customer relations’-type variables, therapeutic alliance (e.g. a positive association has been found between parental satisfaction and length of treatment [13, 36, 49]) or physical characteristics of the experience [33] which could be due to service-level variance, i.e. that there are things specific to the service visited or received that have an impact on satisfaction.

Goal setting and tracking may promote collaborative practice by enhancing child and parental communication with practitioners [61]. In good working relationships, patients and practitioners work together to agree goals, discuss options for preference-based care and agree ways forward [12, 16, 20, 27, 55, 61, 65], which is also a predictor of positive treatment outcome [9, 40, 68].

The above evidence suggests that goal formulation during therapy may be associated with a better experience of support received, as it might promote communication and collaborative practice between practitioners and children, young people and parents, and therefore more positive experiences of and satisfaction with care. To the best of our knowledge, there is no existing evidence on whether goal formulation and tracking is more likely in therapeutic work with certain children and young people, in terms of demographic and case characteristics, or on whether goal formulation is associated with satisfaction with care received.

The aim of the present research was to explore whether goal formulation and tracking was more likely in therapeutic work with children and young people with certain demographic (i.e. age, gender, ethnicity) and case characteristics (i.e. presenting problems). We also sought to explore the association between goal formulation and tracking in therapeutic work with children and young people during therapy, and carer-reported satisfaction with care received.

To explore the relationships between demographic characteristics, presence vs. absence of goal formulation and tracking, and satisfaction with care, two sets of analyses were performed.

First, to explore differences between children with goal formulation and tracking present and children with goal formulation and tracking absent, a multilevel (with children clustered within services) logistic regression predicting the presence vs. absence of goals with the demographic characteristics (age, gender, ethnicity, presenting problems) was conducted. Here, a stepwise model selection approach was used, meaning predictors were only retained in the model if they improved the model fit. This means that all demographic and case characteristics (including the presence of possible combinations of presenting problems recorded for at least 20 children; e.g. hyperactivity × conduct difficulties present vs. absent) were initially entered in the model but were only retained and reported in the Results section if they improved the model fit.

Second, to explore differences between children of parents who were completely satisfied with care and children of parents who were not completely satisfied with care, a zero-inflated Poisson regression predicting parents’ satisfaction scores with demographic characteristics and goal formulation and tracking present vs. absent was conducted. This regression model has two parts: a logistic model for predicting parents being completely satisfied with care (scoring 0) vs. not being completely satisfied with care (scoring 1–18) and a Poisson count regression predicting the actual satisfaction with care score. The choice of this model was motivated by the high number of parents being completely satisfied with care [2415 of 3757 (64.5%)] and by the distribution of scores greater than zero.

This dovetails with previous research, showing that service users are likely to report extremely high levels of satisfaction with support received [15]. In both regressions, stepwise model selection was used, whereby predictors were only retained when their inclusion improved the overall model fit.

Listwise deletion of parents’ satisfaction with care was used rather than substitution (e.g. with mean scores). Attrition in mental health settings has been shown to be particularly high, and CAMHS is no exception, with up to 40% of children and parents not completing treatment [58]. Attrition in child mental health services is complex and the reasons children and their parents may stop attending services are multi-faceted [4]. Parents who are less satisfied may be more likely to stop attending and complete measures [5]; therefore, mean substitution did not seem appropriate.

Table 2 shows the results of the multilevel logistic regression for demographic characteristics predicting the presence vs. absence of goal formulation and tracking. Table 3 shows the results of the presence vs. absence of goals predicting parents being completely satisfied with care vs. not being completely satisfied with care, and also their satisfaction with care score (on the logarithm scale), after controlling for demographic characteristics.

The first aim of the present research was to explore whether goal formulation and tracking was more likely in therapeutic work with children with certain demographic (i.e. age, gender, ethnicity) and case characteristics (i.e. presenting problems). Regarding demographic characteristics, children aged 6–12 were less likely (OR = 0.61, p = 0.05) to have goal formulation and tracking present than children aged 0–5. Regarding presenting problems, children with learning disability were more likely (OR = 8.13, p < 0.01) to have goal formulation and tracking present than children without learning disability.

The second aim of the present research was to explore the association between goal formulation and recording in therapeutic work with children during therapy and satisfaction with care received. Parents of children with goal formulation present reported on average one point (on the log-scale β = –0.4, p < 0.01) higher in satisfaction with care compared to parents of children without goal formulation, even after controlling for demographic and case characteristics.

The aim of the present research was to explore whether goal formulation and tracking was more likely in therapeutic work with children with certain demographic and case characteristics. We also sought to explore the association between the presence of goal formulation and tracking in therapeutic work with children during therapy and satisfaction with care received. Routinely collected data from children attending child mental health services in the UK were analysed.

Pre-school children (0–5 years) were marginally significantly more likely to have goals formulated and tracked than older children, which may reflect goal setting and tracking filling a gap in standardised PROMs as there are few other tools available to measure change in this age group. While standardised measures do potentially measure different constructs to goals, self-report measures for young children are sparse due to the cognitive abilities of this age group; this can be evidenced by their scant presence in the Children and Young People’s Improving Access to Psychological Therapies (CYP IAPT) programme [21], a UK government-funded initiative seeking to transform CAMHS through routine outcome monitoring. Standardised self-report measures such as the Strengths and Difficulties Questionnaire [37] often have a lower age limit of 11 and thus parent reports are relied upon.

Possibly for similar reasons, children with learning disabilities were more likely to have goals formulated and tracked. Children in this group may be subject to similar challenges to younger children, in that the completion of standardised measures requires a certain level of literacy. This may in part explain the association between children with learning disabilities with presence of goal formulation and tracking. Alternatively, it may suggest that practitioners feel more structure is required with these populations and therefore actively choose to work on goals. It could also support the suggestion that goal formulation and tracking is helpful with children with whom standardised measures may be less applicable or less likely to capture changes in symptoms and functioning [6].

Children presenting with both hyperactivity and conduct disorder were more likely to have goals formulated than children without these presenting problems, and with these presenting problems alone. Children seen for hyperactivity along with conduct difficulties may have been seen for ongoing psychological intervention, meaning formulation and tracking of goals may have been more likely. This finding may also be due to practitioners providing additional intensive interventions to meet the needs of these children and their parents. For example, there may be differences in services providing care for children with multiple diagnoses to those seeing children with fewer diagnoses or different types of complexities.

There were large differences between services even when allowing for these factors. This suggests that there may be differences in organisational culture and training in relation to goal setting (and potentially collaborative practice) that may be of relevance. It is known that the move to more collaborative ways of working is novel for some practitioners and there are concerns about this approach (see [25]). In particular, the fact that the majority of cases did not have any evidence of goals being formulated (98.5% of the overall dataset) suggests that this practice has yet to find widespread adoption. There are promising indications that training practitioners to become more collaborative in their work with children can improve their skills and motivation [26]. The current results suggest more targeted training for practitioners working with key groups, particularly those not being currently worked with in terms of goal formulation, may also be warranted.

There may of course be other reasons for the high proportion of cases where goals were not set and tracked, including a lack of resources to record goals formally and a reliance on the subjective views of children [61]. More generally, practitioners may have concerns that outcome measurement of this kind does not capture case complexity, will increase burden or will be misinterpreted [74]. Others have reported barriers from children themselves, related to not feeling that they have the capacity to set goals [67].

Moreover, there may be overarching service-level reasons for not using a particular form of outcome measurement. For example, a commissioning body might make funding decisions based on specific measures, which are often more generalisable and normed measures. Alternatively, practitioners may not feel as though goal setting and tracking fits with their client group; for example, it is unlikely that goals would be set for a child attending a service every 6 months for a medication review.

The findings of the present research suggest that formulating and tracking goals may be associated with higher levels of parental satisfaction, as parents were much more likely to be completely satisfied with care and, on average, rated satisfaction one point higher when goals were formulated and tracked; however, this does not suggest causation and the meaning of a one-point difference in satisfaction is not clear and perhaps likely to be small. What is less clear is what might contribute to this effect. It may be that the mere act of goal formulation communicates a message of collaboration to families that increases satisfaction with the support received. However, it is likely that there are more active ingredients at play: due to the variance between services found, the absence or presence of goal formulation in the dataset may be a reflection of a more or less positive culture of individual therapists or teams. The helpful clinical processes that go on around goal formulation are not yet understood, and further research is needed to understand the ‘active ingredients’ that contribute to effective goal formulation.

Although overall rates of satisfaction were already high, this sample may be positively skewed due to differential drop-out, and the increased satisfaction of those with goals may be even higher in the sample not captured. This finding may suggest that goal formulation and tracking may be associated with higher levels of parental satisfaction, potentially through higher levels of parental involvement and collaborative working, as suggested by prior research [2, 42], and may highlight the association between therapeutic relationships and the importance of goal formulation and tracking [12, 16, 20, 27, 55, 61, 65]. It may of course also be possible that this association is more to do with another common factor such as greater therapeutic alliance to which both higher levels of goals and satisfaction relate, but given the existing literature on how goal setting might help support this it is not unreasonable to at least consider goal setting as enhancing satisfaction and collaborative working.

The finding that parents of children with self-harm were less likely to be completely satisfied with care may be an artefact of the intrinsic nature of those difficulties; for example, parental ambivalence towards receiving treatment for their child’s self-harm has been discussed elsewhere, which may in turn affect their perception of satisfaction of this care [60]. However, there may be other factors related to treatment received, individual differences in patients and practitioners (e.g. ethnicity as also found here), or organisational culture that influenced the findings.