Developing a model of mental health self-care support for children and young people through an integrated evaluation of available types of provision involving systematic review, meta-analysis and case study

Population

Our population was ‘children and young people’, defined as those under the age of 18 years. This reflects the definition of the United Nations Convention on the Rights of the Child,42 of which the UK is a signatory. Because of differing international views of the age range for a ‘young person’, there is often ambiguity at the top of the childhood range (17+ years), thus we also considered studies with populations that included young people up to 25 years of age, so long as we had evidence that the mean age of participants was, or the majority of participants were, under the age of 18 years.

Interventions

Using our definition of self-care (see Chapter 1 ), we included any health, social care or educational intervention or service designed to support or facilitate CYP (or their parents) to take action to promote their mental health, prevent mental ill health, or maintain or enhance their mental health and well-being following recovery from mental ill health. We therefore included self-care support interventions and services which targeted CYP with specific mental health problems as well as those designed to improve the general mental health of CYP. As we were interested in self-care support rather than self-care per se, a support ‘agent’ (e.g. a health, social care or education professional, peer or lay person) who worked with the CYP and/or their parents needed to be present for the intervention or service to qualify.

Comparators

For the effectiveness review, we were unsure whether or not we would find sufficient trials with a control or other such comparison group (see Study design below), and so did not specify a comparator at the outset. As the perceptions review was concerned with absolute, rather than relative, service user views of specific self-care support interventions and services, this aspect of the PICOS framework was disregarded for the perceptions review.

Outcomes

For the effectiveness review, we were interested in whether or not self-care support interventions brought about a demonstrable positive change in mental health. Included studies, therefore, needed to contain a valid standardised mental health measure. We also considered, where available, a range of relevant secondary outcomes (measures of general functioning, general well-being and self-esteem, for example). For the perceptions review, we were not interested in outcomes per se, but in qualitative and quantitative data that captured service user or service provider views.

Study design

For the effectiveness review, we were initially interested in studies containing trials, with ‘trial’ being defined as any study in which there was, at minimum, a relevant pre- and postintervention outcome measure. This meant that, initially, uncontrolled pre/post designs, non-randomised controlled trials and randomised controlled trials (RCTs) were all included. We did, however, add a caveat that should there be sufficient RCTs, we would restrict the effectiveness review solely to RCTs. For the perceptions review, we included any empirical study (qualitative or quantitative) in which service user and/or service provider views about any element of the self-care support intervention or service were available.

Study selection

Of the 380 full-text articles in the initial pool for the effectiveness review, a further two were identified as duplicates, leaving 378 articles for consideration. These 378 articles represented 335 separate studies (some studies generated more than one paper) and two members of the study team (Elvey and Pryjmachuk) independently considered each study against the inclusion criteria. Eighty-six studies were excluded on content grounds – including a few studies concerned with enuresis and encopresis, which the study team and stakeholder group felt did not really fit in with our conceptualisation of mental health problems in CYP – and one on the grounds that it was published before 1995, leaving some 248 content-eligible studies ( Figure 4 ).

FIGURE 4.

Flow diagram of eligible studies for the effectiveness review.

Eight of the content-eligible studies had no empirical data, leaving us with 240 trials (‘trial’ meaning, at minimum, an uncontrolled pre/post design). Given the small proportion of uncontrolled (5.4%) and non-randomised trials (19.2%), we made a post hoc decision to restrict to formal RCTs, leaving us with 181 studies. The very large number of trials meant that it was not possible to extract and analyse data from all of the studies, and we made a second post hoc decision to include only those studies that met the following criteria:

• RCTs with a ‘control’ comparator (defined as a waiting list, usual care or attention/placebo control, for example), as these provided the best evidence concerning the primary research question (evidence for the effectiveness of mental health self-care support for CYP)

• RCTs with long-term follow-up (defined as follow-ups longer in duration than the immediate postintervention follow-up, as these provided the best evidence for the effectiveness of mental health self-care support for CYP in providing enduring gains).

This left us with 71 RCT studies focusing on the longer-term effectiveness of mental health self-care support interventions for CYP and their families. A list of these studies can be found in Table 2 .

Data extraction

Non-English-language papers were translated prior to data extraction. For the effectiveness review, descriptive data for all 71 RCTs were extracted onto a study-specific data extraction sheet (see Appendix 3 ) by one team member (Pryjmachuk), with other team members (Elvey, Kendal and Kirk) providing a second, independent extraction. The two independent extractions were subsequently combined into a separate document (see Appendix 4 for an example) with any discrepancies being resolved through discussion, with referral to a third member of the team for arbitration where a consensus could not be reached.

Quality assessment

In reviews published by The Cochrane Library, quality is assessed by the application of the Cochrane Collection risk of bias tool, a tool that explores several factors known to introduce biases into trials, including the way in which participants are allocated (randomised) into groups; the extent to which this allocation is concealed; the blinding of all those involved, including participants and researchers; the way in which incomplete outcome data are dealt with; and the use of selective outcome reporting. 140

For the current analysis, as well as describing the overall quality of the trials in the review, we sought to assess the impact of study quality through formal quantitative assessment of the relationship between quality and outcomes. To conduct these analyses, we chose a dichotomous measure based on allocation concealment, as this is the aspect of quality most consistently associated with outcomes in trials141,142 and is particularly relevant when outcomes are subjective, as is the case with the bulk of mental health assessments in the current review. 143 Other measures in conventional risk of bias assessments, such as blinding, are less relevant in the current context, as the nature of self-care support means that the conditions for blinding are rarely achievable. Allocation concealment was judged as adequate or inadequate according to the relevant section from the Cochrane risk of bias tool. Two members of the team (Pryjmachuk and Bower) independently assessed all of the included studies on this criterion so that each study could be categorised as either a high-quality (adequate) or a low-quality (inadequate or unclear) study. As with other elements of data extraction, any discrepancies here were resolved by discussion between the data extractors.

As sufficient RCTs were available, our primary data synthesis for the effectiveness review was meta-analysis, although it is augmented by a brief descriptive synthesis of the original 71 included RCTs.

Data extraction

Outcome data for the meta-analysis were independently extracted into a Microsoft Excel 2010 spreadsheet (Microsoft Corporation, Redmond, WA, USA) by two team members (Pryjmachuk and Bower). Data were extracted only for those outcome measures identified as ‘direct mental health’ measures ( Table 3 ; see also Appendix 6 ).

The data extracted for each outcome measure included, where available:

• the sample sizes for the intervention and control groups at randomisation

• whether or not a clustered RCT design had been employed and, if so, the number of ‘units’ (schools, clinics, etc.) in each cluster

• the numerical data necessary to calculate a standardised effect size at each declared follow-up time point, such as the sample sizes, means and standard deviations for the intervention and control groups, or the sample sizes and an event count (e.g. numbers ‘diagnosed anxious’) for each group.

In two studies (Barrett 200547 and Cowell 2009;57 see Table 2 ), discrete subsamples had been used which necessitated splitting each study into several comparisons. Similarly, in six multi-intervention studies (Kendall 2008,83,84 King 2000,86 Pfiffner 1997,101 Simon 2011,112 Stice 2008118,119 and Sánchez-García 2009b;139 again, see Table 2 ), more than one intervention was identified as a self-care support intervention and so these interventions were separated out into individual comparisons. Where a single study included multiple comparisons, the control group sample was split proportionately between each comparison to avoid double counting.

Some notes need to be made about the data extraction for a few other studies. Bernstein 200549,50 was a multi-intervention study [with both cognitive–behavioural therapy (CBT) and ‘parent training’ as self-care support interventions], but the study combined the data for the two interventions at the longer-term follow-up point, meaning that only the combined CBT/parent training data were useable in our meta-analysis. In Hahlweg 2010,73 only the data from mothers were useable. Both Lowry-Webster 200139,90 and Spence 2003113,114 had ‘high anxiety’ subsamples within their overall samples, so a decision was made in each case to use only the total sample. In the multi-intervention Multimodal Study of Children with ADHD (MTA) 1999 study,95–98 only the ‘intensive behavioural treatment’ arm qualified as a discrete self-care support intervention. Similarly, in Rosa Alcázar 2009,106 only ‘IAFS’ (Intervention for Adolescents with Social Phobia) qualified as such. Sheffield 2006110 was a complex four-arm trial which merged universal and high-risk samples across both a targeted and a universal intervention, so a decision was made to use only data from the full sample (universal intervention).

On completion of the outcome data extraction, the two independently extracted Excel spreadsheets were compared and any discrepancies and errors in the extraction were discussed and resolved by the two data extractors.

Primary outcome measure selection

Table 4 outlines the distribution of the various outcome measures used in the 71 included RCT studies. The distribution of outcome measures will be discussed further in the next chapter, but for now it is worth noting that the most frequent direct measure of mental health symptomatology in CYP was one which was obtained from the CYP rather than any other source.

In principle, meta-analysis requires the selection of a single measure from each comparison. As almost 79% of the included studies included CYP self-report data (see Table 4 ), preference was given to a validated CYP self-report measure based on a continuous score directly related to the condition on which the intervention was focused (e.g. a depression outcome measure for an intervention designed to help CYP with depression, an externalising outcome measure for an intervention designed to help CYP with behaviour problems, and so on). If more than one measure qualified on this basis, a global or total symptoms measure was chosen in preference to a specific (subscale) measure, and if more than one global measure was available, we made a judgement about which was the most established.

If no CYP self-report, continuous measures were available, parent-reported measures were selected; for the few studies that had neither child- nor parent-reported measures, clinician-reported measures (usually measures eliciting a diagnosis) were employed. Where a choice needed to be made between intensity/severity scales and frequency/number of problems scales, intensity scales took precedence. The primary outcome measures selected from each study are listed in Table 3 .

Once the primary outcome measures for each study had been determined, a combined Excel spreadsheet was prepared ready for transfer to Comprehensive Meta-Analysis (CMA) (Biostat, Inc., Englewood, NJ, USA), the meta-analysis software we employed. In preparing this spreadsheet, studies were also grouped according to the follow-up time points for which data were available. We were able to categorise the studies into three broad follow-up periods relatively easily: approximately 6 months, approximately 12 months and approximately 24 months post intervention.

Data preparation: missing data, cluster adjustments and effect direction

Where sample sizes were missing at follow-up, they were estimated by assuming an arbitrary 70% follow-up from baseline (such imputation will only influence the weight associated with the comparison in the meta-analysis, not the estimate of effect of that study). Where there were missing standard deviations, these were estimated using included studies that had the same outcome measure. Missing standard deviations for the Center for Epidemiologic Studies-Depression Scale (CES-D) at 6-month and 12-month follow-ups in Compas 200956 were estimated from other studies with 6-month CES-D data (two studies: Clarke 199553 and Garber 200967) and 12-month CES-D data (four studies: Clarke 1995,53 Clarke 2001,54 Clarke 200255 and Sawyer 2010108,109).

Three studies (Evans 2007,65 Compas 200956 and Tol 2008128) reported only a raw effect size which needed to be converted to a standard error in order to be compatible with CMA.

Cluster trials are frequently used in trials of psychological and mental health interventions as a way of avoiding bias associated with contamination. In line with guidance from the Cochrane Handbook for Systematic Reviews of Interventions,140 the sample sizes of approximately one-third of studies using cluster randomisation were adjusted by dividing the original sample by the ‘design effect’, which was calculated using the formula 1 + (M – 1) × ICC, where M is the mean cluster size (total sample size divided by total number of clusters) and ICC is the intracluster correlation coefficient, which we assumed to be 0.02.

The effect direction for each primary outcome was coded in CMA such that effects in favour of the intervention (self-care support) were coded negative. All coded effect directions were compared for consistency with what was reported narratively in the papers describing each study.

Primary and secondary analyses

Analyses were conducted in CMA using a random-effects model in each case. Our primary analysis focused on the long-term effectiveness of self-care support interventions in mental health symptomatology in CYP. We also explored subgroup analyses using a variety of prespecified ‘moderators’ (i.e. variables that were associated with the benefit of self-care support) that were chosen on the basis of theory and practical considerations. These moderators were the intervention level (see Chapter 1 , Table 1 ); the mental health condition and the age group the intervention was designed for; the theoretical model underpinning the intervention; the recipients (child/young person, parent or family); whether the intervention was school-based or not; group versus individual interventions; whether the agent was mental health trained or not; and the length of the intervention. Concealment of allocation was also included as a moderator.

There were relatively limited follow-up data at 24 months, so for all analyses we restricted the data to two follow-up points: around 6 months and around 12 months. In all cases, the meta-analysis results are presented in terms of a standardised mean difference (effect size) with 95% confidence limits, together with the relevant I ² statistic which is a measure of consistency (heterogeneity) across the included studies. I ² scores above 75%, 50% and 25% are said to have high, moderate and low heterogeneity (inconsistency) respectively. 144 Interpretation of the pooled outcome in meta-analyses with high I ² scores should be treated with caution. For the primary analyses only, we have additionally provided forest plots, produced via Stata Version 12.0 (StataCorp LP, College Station, TX, USA).

Study selection

For the perceptions review, 187 individual studies were identified from the pool of 193 full-text articles. Forty-six of these studies were excluded on content grounds and seven were identified as non-empirical. This left 134 studies with perceptions data for consideration and quality appraisal: 30 studies with qualitative perceptions data, 101 studies with quantitative perceptions data and 3 studies with both ( Figure 6 ). The application of quality criteria (discussed shortly) ultimately reduced the number of studies to 12 with qualitative perceptions data, 35 with quantitative perceptions data and one with both, making a total of 48 included studies.

FIGURE 6.

Flow diagram of eligible studies for the perceptions review.

Data extraction

As with the effectiveness review, non-English-language papers were translated prior to data extraction. Descriptive data for the 33 qualitative perceptions studies (of which 30 were solely qualitative studies and three were mixed methods studies) were extracted onto a study-specific data extraction sheet (see Appendix 5 ) by one team member (Pryjmachuk), with other team members (Elvey, Kendal and Kirk) providing a second, independent extraction. As with the effectiveness review, the two independent extractions were subsequently combined into a separate document, with any discrepancies being resolved through extractor discussion with referral to a third member of the team for arbitration where a consensus could not be reached. For pragmatic reasons, the 104 quantitative studies (101 solely quantitative and three mixed-methods studies) were filtered for quality prior to data extraction (see below), with data from the 36 quantitative perceptions studies eligible at this point being extracted directly into a Microsoft Excel spreadsheet by one team member (Pryjmachuk) and subsequently verified by a second (Elvey).

Quality assessment

For the qualitative perceptions studies, study quality for all 33 studies was assessed during the overall data extraction by one team member (Pryjmachuk) and independently by the second person extracting (Elvey, Kirk or Kendal). Quality was assessed using a modified version of the quality assessment tool used in our previous NIHR study on self-care support,15 which in turn was adapted from tools developed by Dixon-Woods et al. 145 and the Evidence for Policy and Practice Information and Co-ordinating Centre (EPPI-Centre). 146 A copy of the quality assessment tool can be found in Appendix 5 , embedded into the perceptions review data extraction sheet. Again, any quality assessment discrepancies were resolved by discussion between the extractors and, where necessary, through arbitration by a third member of the team. The tool comprised seven criteria, and for the purpose of this review we decided that a paper had to meet four criteria for inclusion.

We identified 104 studies as containing quantitative perceptions data. This was a substantial number of studies to consider for data extraction so, in contrast to the qualitative studies, we made a pragmatic decision to apply quality criteria prior to, rather than during or after, data extraction. During the process of scanning papers for perceptions data (see Search results and study selection and Box 2 earlier in this chapter), it became evident that trials were the sole source of quantitative perceptions data, i.e. all of the 104 studies containing quantitative perceptions data were controlled or uncontrolled trials. In the main, the quantitative perceptions data came from satisfaction surveys nested within these trials or from statistics concerning intervention uptake or adherence, though we did not consider the latter to be perceptions data. Thus, to manage the number of quantitative perceptions studies, we applied our PICOS criteria (see Inclusion criteria earlier in this chapter), selecting only randomised trials with a true control. Accordingly, only 36 trials with quantitative perceptions data (one of which was a mixed-methods study that also contained qualitative data) were eligible for the perceptions review.

Details of the included studies for the perceptions review can be found in Table 5 .

Intervention level

This analysis considered whether or not there were differing degrees of effectiveness when the interventions were categorised according to the levels of the intervention hierarchy discussed in Chapter 1 . A summary of the results can be found in Table 8 .

An interesting pattern emerges from this analysis in that the effectiveness of the self-care support intervention appears to increase as the interventions become more targeted and specialised. Those classified as treatment interventions are the most effective at 6-month follow-up, with a moderately large effect size, and indicated interventions, the prior intervention level, have a reasonable effect. At 12-month follow-up, selective, indicated and treatment interventions are modestly effective. At both time points, prevention interventions appear to be relatively ineffective. This statement, however, needs to be treated cautiously because (1) this analysis is exploratory and there may well be other factors confounding the results; and (2) the prospective nature and purpose of prevention interventions may be such that it is inappropriate to compare them with interventions designed to treat specific conditions, a point that we will pick up in the discussion.

Mental health condition

Across the two follow-up points, only four conditions had a meaningful number of studies available for analysis: anxiety disorders, depression, serious behaviour problems and eating disorders. The results of the comparisons across these four conditions can be found in Table 9 .

At 6 months, self-care support interventions were most effective for behaviour problems (a moderate effect size), with interventions for anxiety disorders and depression also showing a small effect. The effects appear to be maintained for anxiety disorders at 12-month follow-up. For depression, the effect seems to have attenuated at 12 months. For serious behaviour problems, there was still an effect at 12 months but it was small compared with the effect at 6 months. There were very small effects for eating disorders at the 6-month follow-up point and, interestingly, there is a hint that effects may increase with time.

Age

We divided the studies according to a commonly assumed demarcation between ‘children’ and ‘young people’: the age of 12 years. A summary of the results can be found in Table 10 .

At 6-month follow-up, self-care support interventions for ‘children’ (aged under 12 years) are slightly more effective than those for ‘young people’ (those aged 12 years and over), with both having small effects. The difference between the SMDs in the two groups (0.06) is, however, negligible. At 12-month follow-up, the effects are attenuated but there is still a modest effect for both age groups.

Theoretical model

As discussed earlier in this chapter, self-care support interventions based on cognitive–behavioural principles and parent training dominated the included studies. This meant that there were few studies considering other models, so this analysis is confined to those interventions using cognitive–behavioural principles, parent training or a mixture of both. This analysis examined whether the principal models identified were effective alone or in combination. A summary of the results can be found in Table 11 .

At 6-month follow-up, both models seem reasonably effective, but it is interesting that the effect is much larger in self-care support interventions where the two approaches are combined. This result is not reflected at the 12-month follow-up, however, where all of the approaches, including the combined approach, elicit only modest effects.

Recipient

This analysis examined whether self-care support interventions were more effective when targeted at the CYP alone, the family as a whole or solely at parents. Table 12 contains a summary of the results.

No clear pattern emerges from these results. At 6 months, self-care support interventions appear to be equally effective (with a small effect size), regardless of whom the recipient of the intervention is. At 12 months, family-targeted self-care support interventions appear to have a slight advantage over the child- or parent-only interventions.

School-based interventions

This analysis examined whether or not school-based self-care support interventions were more effective than interventions conducted in other settings such as hospitals, clinics, other community-based locations and the child or young person’s own home. The results are summarised in Table 13 .

Whether self-care support interventions are school based or not seems to have no influence on their effectiveness. There is a slight advantage to school-based interventions at 6-month follow-up (both exhibit a small effect) but the difference between the SMDs (0.04) is negligible; at 12 months, they both have a modest effect.

Group versus individual intervention

This analysis examined whether or not group self-care support interventions were more effective than individual interventions. Two sets of analysis were conducted, one looking at those interventions using a group, individual or mixed group/individual format regardless of the recipient (child/young person, parent or family; Table 14 ), and one looking at those interventions targeting only CYP ( Table 15 ).

For all recipients, individual interventions appeared to be slightly more effective at 6-month follow-up, though both types of intervention could be described as eliciting a small effect. Mixing the two types of intervention elicits the biggest effect but this finding needs to be treated with caution given the number of comparisons (n = 3) available. At 12-month follow-up, there was little to distinguish between group and individual interventions (both having modest effect sizes) and mixing the two appeared to produce no effect.

The findings are similar when only CYP recipients are considered, except that the effect for individual interventions is perhaps more pronounced at 6 months. Again, there is little to distinguish between the two types of intervention at 12 months.

In all, this analysis hints that, despite being in the minority of studies, individual self-care support interventions may be more effective than group interventions.

Mental health or generic agent

This analysis examined whether self-care support interventions delivered by mental health-trained staff (such as clinical psychologists or mental health nurses) were more effective than those where the agent was not mental health trained (general nurses, teachers and educational psychologists, for example). The results are summarised in Table 16 .

At 6-month follow-up, there was little to distinguish between the SMDs for mental health and non-mental health-trained staff. At 12-month follow-up, self-care support interventions delivered by mental health-trained staff were only marginally more effective than interventions delivered by non-mental health staff. This suggests that, to be effective, self-care support interventions do not have to be delivered and facilitated by mental health-trained staff, although this is not the same as saying that the agents do not need to have some understanding of mental health in CYP.

Length of intervention

Length of intervention – or perhaps more correctly, contact time with the agent – was examined by making an arbitrary split at 12 hours of contact time so that a ‘short’ intervention was 12 hours or less and a ‘long’ intervention, more than 12 hours. The results of this analysis are presented in Table 17 .

At 6-month follow-up, the length of the intervention appears to be related to the effectiveness of the intervention, though the effect is attenuated at 12 months. This gives some hints, at least as far as the short-term effectiveness of self-care support interventions is concerned, that length of the intervention or amount of contact time may be important, in that longer interventions appear to be more effective.

Semistructured interviews

Semistructured interviews were used as the main data collection technique. They were conducted with staff at all six sites and with CYP and/or parents at all sites except the eating disorders online support site. Pragmatism largely dictated the way in which participants were interviewed in that they were given choices as to which approach they preferred (e.g. telephone vs. face to face; individual vs. group). Forty-two interviews involving 52 participants were conducted; 37 were conducted in person, either at the site’s premises or at the service user’s home, and five were conducted via the telephone. The majority of interviews (35/42) were individual, five were joint interviews with a child or young person plus parent, one was a group interview with a child, parent and sibling and one was a focus group of four young people. The interviews ranged in length from 10 minutes (for some of the youngest children, aged 5 and 6 years) to 1 hour and 20 minutes. Interviews were facilitated by the use of topic guides, copies of which can be found in Appendix 12 . All of the interviews were digitally audio recorded with the permission of participants and the recordings were subsequently transcribed verbatim by a professional company.

Documentary review

Where available, relevant documentary evidence was collected from each site. For sites 5 and 6, documentary evidence was obtained simply from the relevant websites of the services. More comprehensive data were obtained from the other four sites, including a referrer’s leaflet for site 1 as well as a printed self-injury ‘toolkit’ used in providing the self-care support service there; a Microsoft PowerPoint general presentation about the Incredible Years services at sites 2 and 3, and one about the funder’s strategic plan regarding these services; an information leaflet about Dina School training for teachers at site 2; and information leaflets about the site 4 service for referrers, parents and CYP. The documentary evidence largely served to provide a contextual background to, and additional understanding of, the six sites.

Virtual non-participant observation

For the eating disorders online support site (site 5), we collected the data via ‘netnographic’ non-participation observation, adopting the principles of a method designed specifically for health-care research. 188,189 Netnography is a form of ethnography used in the study of online behaviour, and as ethnography concerns everyday routine behaviours in a natural setting,190 netnography was an entirely appropriate method of observing a service that operated entirely online. The service centred around discussion threads that were organised by the providers into nine boards, following themes such as ‘recovery’, ‘caring about someone’, ‘introductions’ and ‘poetry’. Two boards were excluded from the study as the content merely described how to use the boards. All postings over a 4-month period (August to November 2012) were collected from the remaining seven boards. These comprised 114 discussion threads with more than 500 individual messages, each of which was screened for relevance. The postings were copied from the website and pasted into Microsoft Word 2010 documents (Microsoft Corporation, Redmond, WA, USA) prior to analysis.

Regarding management and coding of the data, the three data sets – interview transcripts, background information about the sites (documentary evidence) and the online group postings – were imported into NVivo 10 (QSR International, Warrington, UK), computer software designed to assist in the collection, organisation and analysis of qualitative and mixed-methods data.

The Incredible Years

The group for children with emotional and behavioural problems (site 2) and the parenting group (site 3) were interconnected, but separate, services. Both operated as part of Webster-Stratton’s The Incredible Years programme183 that had been adopted region-wide (across the local authority area) by the commissioning bodies in the region. The Incredible Years is an evidence-based programme, influenced heavily by social learning theory. It emphasises attachment, relationship building and emotional coaching, with a focus on children learning through play and parents and teachers spending time with children, listening to them and giving them positive feedback. The overall Incredible Years programme comprises universal and targeted (indicated) interventions, both of which were available in the region. At the time of data collection, around half of the region’s schools were Incredible Years schools. In these schools, all staff members were trained in the approach and were thus able to implement the programme universally (in a regular classroom setting, for example) or in a more targeted way.

Dina School

Site 2 was an example of a targeted intervention operating in one of the Incredible Years schools. Known as Dina School, it involved children taking part in 18 weekly, small group sessions, which took place in a room at the children’s school. The groups were facilitated by two classroom assistants, who delivered the programme as set out in the Dina School manual. The programme makes use of a dinosaur puppet called ‘Dina’ (which was also used universally in the main classroom setting) as well as two additional character puppets, ‘Wally’ and ‘Molly’. These puppets were used in role plays, demonstrations of behaviours and communication techniques and in video vignettes. Each session followed a similar format, with a review of the ‘homework’ tasks that were set at the previous session, followed by activities and games and the setting of a further homework task at the end. Parents came to collect their children at the end of the session which coincided with the end of the school day.

Understanding the mental health problem

The conceptualisation of CYP’s difficulties specifically as mental health problems appeared to be connected to the CYP’s age. The youngest children interviewed, aged 5 and 6 years, did not describe having any mental health problems or difficulties themselves. School staff and parents tended to think that these children mainly lacked confidence or had difficulties expressing or managing their emotions. Some parents described emotional outbursts and tantrums and two described their children as being ‘unhappy’.

I had concerns about his behaviour; he would break down into quite severe tantrums that would last for 20 minutes . . . he would spit all over the seats . . . strip naked and hit and lash and kick; hurt himself, hurt me. And generally I just felt that I had this unhappy child on my hands … I couldn’t go out with him . . . So I was kind of feeling trapped in my own home because of his behaviour.

Parent of younger child

Older, primary school-age children and their parents described various emotional and behavioural problems: being unsettled or disruptive at school; having emotional outbursts at home; having problems with family members or in making friends; lacking confidence; and having communication problems, especially in expressing emotions.

Me and my mum didn’t really use to get along, and not able to cope, like, and with my brother and sister, I didn’t know how to be responsible and stuff.

Child

Young people – that is, older children – who had used the self-harm (site 1), eating disorders online support (site 5) and depression/anxiety (site 6) services had a wide range of experiences, from severe mental health problems and diagnosed conditions, to milder symptoms and difficulties. Some attendees at the self-harm site had been diagnosed with depression and others described feelings of low mood, anxiety and intense feelings. Staff at the self-harm group mentioned a variety of self-harm that users presented with, including cutting and burning. Some CYP attending the self-harm (site 1) and family support (site 4) services had experienced more than one type of difficulty or problem, including being bullied at school, not attending school, displaying signs/symptoms of autism, family relationship problems and misuse of alcohol and/or drugs.

Self-care and self-care support

Just as participants described a wide range of mental health problems and difficulties, their experiences of self-care and support for their problems and difficulties prior to accessing the services were also mixed. Obtaining support was sometimes related to their understanding of the specific mental health problem, as in the case of one young person who had struggled with depression over an extended period:

So had you had any help with your depression before you went to the centre?

No, none, not at all. I had little understanding as well, it wasn’t something that had been spoken about.

Participants generally described undertaking little self-care before coming into contact with the services. Nonetheless, some parents whose children had attended the Dina School groups (site 2) had also previously attended an Incredible Years parenting course and had used some of the approaches at home. A few parents mentioned using techniques from books or television programmes such as Supernanny. Across most of the sites, the situation for many participants was that they came into contact with the services at a point when their problems had emerged, but they had rarely engaged in self-care themselves or received support they were happy with.

With some of the older children (young people), there was some evidence that they could make rational choices regarding self-management, especially when it came to medication:

I usually forget to take it, or I intentionally go out my way not to take it, because I feel as if I don’t need it and it makes me feel different . . . I was on Sertraline, but it kept me awake and then I was on Mirtazapine and then some other things . . . I didn’t like them . . . I don’t feel as if meds help me.

I’m on medication at the moment, because my depression has peaked at the moment . . . so it [the medication] does work.

However, these choices were not always adaptive, as in these examples of young people ‘self-medicating’ with alcohol:

I used to drink and hide in a tree to drink, because my mum wouldn’t allow it, so I used to hide and had to go in a hole and fall out of a tree regularly drinking, that’s how I coped.

Yeah. I used to think alcohol helped me, but it used to just make me worse.

Regarding the support received, there were participants at all of the case study sites with experience of accessing some form of health care or support prior to attending the self-care support service; CAMHS, social services, GPs, psychiatrists, paediatricians, school counselling services and social services were all mentioned. For example, some families had a history of contact with social services due to family situations and some parents had sought help because they were concerned that their child might have an autistic spectrum disorder. There were some reports of positive experiences, such as helpful school counsellors, health visitors and other workers:

My outreach worker who works with me to integrate into the community, because I’m, kind of, agoraphobic, I don’t like people, crowds, so he works with me to try and get me into the community.

Young person

However, many negative experiences were recounted where families had felt dismissed when they raised concerns, for example, about their children’s behaviour or social understanding and had found support hard to access. The mother of one child who had previously been referred to community-based mental health services had found the care provided inadequate:

It didn’t help much, she used to just enjoy going there because I was doing all the talking, they would watch her play, so it wasn’t . . . [helping her with] expressing her feelings . . . and then it just kind of stopped, they were thinking she didn’t need it, but she’d got so much anger at the time, she wanted to leave the house . . . I was worried for her.

Parent

Two settings in particular were singled out for criticism by young people: mainstream schools and inpatient hospital care. Young people had experienced bullying at school, including being bullied about their mental health problems by other students. Some had stopped attending mainstream school and were at school units and felt that staff at mainstream schools often struggled to help students who self-harmed:

They’re not trained in mental health and are only really equipped to do with stuff, like, things to do with education and . . . bullying and peer pressure . . . they’re not equipped . . . they really don’t have a clue when it comes to stuff like mental health and things like self-harm . . . I had a really bad experience with my mentor.

Young person

In terms of hospitals, participants described how support for self-harm in inpatient settings focused mainly on preventing physical harm. For example, two participants reported being restrained and put into seclusion (actions which could both be seen as punitive) when staff found that they were in possession of objects which the staff thought the young people would use to self-harm. Participants had found that they could access either group ‘talking therapies’ or an individual to talk to in the units, but had found this problematic because the therapy was not suitable for them at that stage, or because staff did not spend enough time with them:

There’s people there all the time that you can talk to, [but] I think the therapy puts quite a lot more stress on you at the very beginning . . . you’re taken out of your house . . . kind of throw you in with a bunch of new people, that’s hard to deal with . . . it’s a lot harder to then feel comfortable . . . I think, I had therapy twice, it didn’t work, so they basically just left me.

[My named nurse] was on nights and then when she was on days, she didn’t speak to me anyway, that was so unhelpful. She came to my room at like half past 10 and I’d be talking about stuff and getting upset and minutes later I had to try and sleep.

Yeah I had some nights when mine was . . . on nights and the head of [name of department], so he really didn’t have time [for me].

Running throughout the narratives of the young people was a recurring sense of anxiety around accessing support. Feeling alone, not knowing where to turn for help, discomfort and worry about discussing their problems with family or friends were mentioned repeatedly. The following data extract was taken from one of the site 5 message boards:

When you realise youve got some kind of (eating disorder) what was the first thing you all did? It’s just im stuck and going nowhere, the idea of food and excercise is going round and round in my mind, having crazy thoughts in my head all the time- but i cant tell anyone or even talk about it to anyone because i could shy when it comes to spilling out information about myself . . . any help? Pleasepleaseplease . . . ALSO, if you call the youthline what do they ask you and stuff? because I really wanna call them but im scared of the reponse and over the phone (procedure).

Young person, verbatim message board posting

Service development

Staff at the sites became involved in the services for a variety of reasons: a desire to change things or innovate on the basis of personal, often negative, experiences of mental health services; wanting to improve the evidence base for practice; and often just sheer enthusiasm to help CYP. For example, the manager of one site had been motivated to found an independent alternative to traditional ‘clinical’ approaches on the basis of personal negative experiences elsewhere:

I started to experience the onset of depression and anxiety and looked at what sort of support or help was available to me and there was nothing that I felt I would have been likely to access or appealed . . . so I . . . basically came up with a service that I would have liked to see when I was in that situation and applied for funding and got it.

Service lead

At another site, the service lead had been influenced by observations from her own practice, at a time when there was a perceived ‘epidemic’ of self-harm among young people and generally insufficient support available, and when anxiety about the risk of serious harm, including suicide, was high among staff who worked with young people. The rationale behind setting up a group therapy service was that a group approach could be appropriate for young people because of the influence and importance of peer groups to this age group. Moreover, improvements in peers would be clearly visible in a group setting and so serve as a vehicle for hope and optimism in other group members.

The Incredible Years programme, on the other hand, had been adopted at two sites because of its perceived strong evidence base:

I felt really strongly that I knew that the programmes were blueprint [evidence-based] programmes . . . that met the high standards for replicability and for research . . . I really, really like the Incredible Years model.

Staff member

Moreover, service development seemed to be coupled with service leads who were enthusiastic and highly motivated, even to the point of being ‘on a mission’:

So it kind of became my mission and it’s grown into the mission for the Authority to develop all of those programmes.

Service lead

Across all six case study sites, the attributes of the service leads in particular were suggestive of a high level of leadership skills, in that there was evidence of innovation, planning, empathy (for CYP, parents and colleagues), motivation and communication.

Training and supervision

All of the sites except the online eating disorders site (site 5) were manualised, that is there was a specific written manual in place for how the service should be delivered. The Incredible Years sites (sites 2 and 3) and the anxiety/depression site (site 6) were guided by manuals that prescribed the topics to be covered in each session; the self-harm (site 1) and family support (site 4) services had manuals that were less prescriptive.

At the self-harm site (site 1), all staff running the groups were provided with the service manual and, as part of their training, they also observed groups prior to becoming a group leader. Staff running the groups met monthly for group supervision with the service lead. On its initial formation, staff at the family support service (site 4) had been trained in the resilience approach by its architects. New staff members were first introduced to the model and subsequently trained by the service’s current staff, as well as receiving clinical supervision from a psychologist. At the anxiety/depression site (site 6), training was provided by the company supplying the CBT programme and all volunteers had to work through the course before working as a volunteer; there was no formal clinical supervision in place at this site, however.

The two Incredible Years services (sites 2 and 3) were different in that, of all the sites, they had the most structured and formal arrangements for training and supervision. To deliver a programme, staff members were required to undergo formal training. Staff members were regularly supervised by local colleagues and there was ongoing monitoring from the programme base in the USA. The service lead explained why she thought the training and ongoing supervision and support were important:

We’re using an evidence-based programme and actually unless you really are delivering with fidelity we know that you can’t guarantee that you’re going to get the same results as [the] research . . . everybody who delivers the group in [area name] can be sure they’re either going to be able to work with [or get] supervision from someone who’s accredited either as a peer coach or as a mentor . . . I think that is really important in making it effective. These are hard groups to run, so it’s important ensuring that people do get the support.

Service lead

Fidelity was not emphasised as strongly at any of the other sites. This is perhaps because fidelity can be in opposition to flexibility, a characteristic seen by many as a key factor in a service’s accessibility, as will become apparent when the next theme, Accessing self-care support, is discussed.

Integration with other services

As outlined earlier in the site descriptions, the two Incredible Years sites (sites 2 and 3) had been widely adopted by local commissioning bodies. These services were closely integrated with local planning and commissioning structures and with the local education, health and social services sectors. The Incredible Years services were an exception, however. None of the other services had this level of integration with health, education and social care. The family support service (site 4) was integrated in as much as it was run by a voluntary sector organisation but funded by a local authority, with clinical supervision provided by an NHS psychologist. It also had close links with special educational needs co-ordinators (SENCOs) in schools and could form part of an action plan arising from national ‘Common Assessment Framework’196 assessments. Staff at the family support service, however, perceived that there were overlaps between their work and that of the NHS and social services, and that they fitted into a niche between the two.

The self-harm group (site 1) was run from a NHS mental health trust, at a CAMH day service. The service is attached to an inpatient unit with residential and non-residential care and education provision. Self-harm groups using the same principles are run at other NHS locations in the region, and although these groups are integrated with other services to the extent that GPs and consultants can refer CYP to them, they are usually run as ‘standalone’ services with little integration with the NHS CAMHS provider delivering them. The eating disorders (site 5) and depression/anxiety (site 6) services were provided by charities dedicated to helping people with particular difficulties in these areas. The depression/anxiety service was run by a small, relatively new organisation which provided some other activities as well as the supported online CBT, and to some extent could be seen as an alternative, rather than a complement to, statutory services. Regarding the eating disorders charity operating at site 5, integration tended to be limited to ‘signposting’ in that message board posters often offered advice to other board users about how to access statutory services, encouraging users to overcome their anxieties in accessing such services and offering opinions about those services.

Where integration was most evident was in the referral processes at the various sites. This is discussed further in the next section.

Referral to services

Children, young people and their families had accessed the services via a range of routes, including self-referral, signposting and referral by professionals. The self-harm (site 1), family support (site 4) and parenting group (site 3) services were well integrated into referral pathways and several CYP and parents at these sites had been referred in this way. Several postings on the eating disorders message boards (site 5) mentioned being ‘signposted’ to the board by health professionals, including psychologists and doctors. The parenting group (site 3) and anxiety/depression (site 6) sites encouraged self-referral and publicised their services; some parents had found out about the service through leaflets in their child’s school bag, or through a friend. The anxiety/depression site was advertised through posters and leaflets in shops and bars as well as through presentations at schools and Sure Start centres. At Dina School (site 2), staff at the school had approached the parents of children who they thought would benefit from the service and asked these parents for consent to include their children in the group. Schools were a common source of referrals for the family support service (site 4) and two children who participated in the study had been referred by school staff, one by a SENCO and one by a school nurse. The young person interviewed at the anxiety/depression site (site 6) had self-referred to this service following a presentation at school; in interview, the service manager expressed disappointment that no school nurses had made referrals to the service.

One family who had attended the family support service recalled a long wait (around 2 years) from becoming aware of the programme to the time when they started receiving the support. Waiting times were not cited as a problem, however, by other study participants.

Building and maintaining engagement

In building engagement with health-care services, one of the first barriers to overcome can be the physical access to services. The participants at all of the case study sites seemed to encounter few physical barriers to access, though it should be added that we did not recruit those with probably the best information about barriers to access – ‘dropouts’ from the services – to our sample. For the physically provided services, none of the CYP or parents reported particular problems with travel to the sites. The virtual eating disorders support site could be accessed via the internet at any time, although messages were only uploaded when staff members were available at the service to moderate them (until 20.30 on weekdays and until the afternoon on Saturdays). The Incredible Years services (sites 2 and 3) covered a large rural area and were provided in convenient venues across the region. The young people at the self-harm group site (site 1) were all at school or college and often had to miss school or college to attend the service, though none of the young people interviewed raised this as an issue. These young people were mostly brought by car to the service by their parents. There were several examples of staff working to make services convenient to attend, or taking the service to the families. For example, staff at the family support service (site 4) worked with families in their homes and with children at school. The parenting group (site 3) was generally provided in a group setting but could be run by staff on an individual basis at people’s homes if necessary:

It’s very different to a traditional service where you might send out an invitation and if the parents don’t come then . . . they might say that you might not be able to access the service. With Incredible Years it’s more about going out and getting the parents really. So it’s about awareness raising, training lots of agencies who know about the programme . . . but [also] offering really nice coffee and biscuits. If parents miss a session, [it’s] really important that they have the hand-outs. So, if it’s possible, the leader goes and visits them at home.

Recently we’ve had a case [where] mum . . . is not ever there physically when we turn up for our appointment . . . so our worker . . . she’ll go to the child’s nursery . . . to try and catch mum at a drop-off, just to have that initial face-to-face engagement because we were mindful that this is a mum who is surrounded at the moment by professionals who are all breathing down her neck, and we wanted to make sure that mum had a fair view of where we were in that process, what our role is and that we’re not scary monsters who are trying to trip her up.

Flexibility in service provision seemed to be a key in not only building, but also maintaining, engagement with CYP and their families. As a member of staff at the family support service (site 4) outlines:

It’s looking at each case on an individual basis and thinking about what are the reasons for disengagement, why haven’t they engaged, is this a language issue, did they not understand when the appointment was, is it that the intervention isn’t working for them . . . It’s quite involved; but we don’t just do a kind of, if you don’t pitch up we send you a letter and then if you don’t pitch up again we bin you off to be picked up by some other agency – we will attempt always to get an answer. And then if it’s just that it’s not working, well, then we’ll have a conversation about that and let’s work it out.

Staff member

Another example of the flexible nature of the services was the between-session support that the sites offered. At the self-harm (site 1), family support (site 4) and parenting group (site 3) sites, support in addition to the scheduled sessions was available, including staff being available to parents over the telephone if they wanted to make contact. Staff at these sites emphasised that they worked hard to engage people in the services and to maintain engagement once a supporting relationship had been established. However, a participant who provided clinical supervision at one of the sites provided an alternative perspective on this, suggesting that it was possible to be overly flexible in that, for example, always bringing the service to a family’s home could potentially discourage independence:

I think, the fact that they [the organisation] go out and work with families in their local area, either at home, or in schools . . . that’s really important . . . something that families will like [but] if you’re going to visit somebody at home then . . . how do you evaluate their motivation to change? Because . . . if you’re at home and somebody comes to see you then actually you don’t necessarily need to do anything to engage, other than sit and nod and make the right noises . . . but if you’ve actually got to physically leave the house and go somewhere, then that suggests that your motivation might be greater to engage . . . it’s trying to get the match between that initial buy in to the service [and maintaining engagement].

Staff member (from outside the organisation)

Some staff expressed an awareness of working with parents who were used to being highly monitored by statutory services and said this could sometimes be a challenge when working to engage parents initially. A school-based member of staff, for example, who had referred several children to the service observed:

You do have the odd family who you refer who don’t engage. And it’s very sad really . . . it’s usually . . . parents who have already got to the point of social services . . . [people think] oh, don’t get social services involved, they’ll take my children from me. And once social services are involved . . . they’re breathing down your neck all the time then, aren’t they? I mean, they’re popping in and out of your house all the time, they’re watching your every move, very intrusive.

Staff member

Welcoming staff; welcoming organisations

Positive staff attitudes appeared to be particularly important. At every site (apart from the online eating disorders service), service users spoke spontaneously and positively about the staff running the service, using adjectives such as ‘nice’, ‘lovely’, ‘good’ and ‘thoughtful’ to describe them. Listening to CYP, allowing them to tell their own story and treating them with empathy and compassion, was also important. This was mentioned particularly by parents at the family support service (site 4) and the young people at the self-harm service (site 1), who liked the manner of the staff and contrasted this with previous, negative experiences at other services. Young people using the self-harm (site 1) and anxiety/depression (site 6) services in particular emphasised the importance of being able to trust staff in order to talk to them openly:

You have to feel comfortable talking to that person, if you don’t like them, then you’re not going to feel comfortable, so you’re not going to engage with them at the level that’s needed to help [you] recover.

Young person

At several sites, CYP and parents perceived staff as wanting to understand and help them and felt that they were treated with care and compassion:

The service worker talked to me as if he’d been there before . . . went out of his way for me.

Young person

It wasn’t just a job to her.

Parent of a younger child

Knowing that the service would accept people for who they were and be open to hearing about their problems was an important attribute of the case study sites operated by voluntary sector organisations. Being non-judgemental or offering non-judgemental support was a particularly positive aspect:

You could be honest and say, ‘This is what my child has done’, and you weren’t judged.

Parent of a younger child

Indeed, contrasts were sometimes drawn with other statutory or conventional health and social care services perceived (unfairly or not) as being judgemental. A worker who referred into the family support service noted:

You’ve got to the point where this has become . . . crisis time: ‘We’ve got a plan here in front of us, you will do this, you will do this, you will do this’ . . . whereas with [site name] it’s not, ‘you will do . . .’ it’s ’these are the suggestions that we can make, that will make things better for you’. So it’s not as judgemental, well, social services aren’t judgemental, I suppose, but it might feel like they are . . . I hear in lots of groups that I go to . . . ‘you’re all judging me, that I’m this, you’re judging me on that’.

Staff member

A similar feeling was expressed in an interview with a young person at another site:

[With the general practitioner (GP) and] even with CAMHS sometimes . . . it seems a bit like I’m being judged, or it doesn’t seem like they’re there for me in the way that I’d like, but it was different with [site name], like, it definitely felt a lot more welcoming.

Okay. Is it the people at the GP and CAMHS? Or is it more general?

I think, it’s more the method, like, I mean, it’s just all about a feeling, it feels like if I’m at the doctors . . . it becomes a thing that I’m ill, or there’s something wrong with me . . . you still sit there and you feel like you’re being judged . . . it definitely affects your confidence when you’re trying to answer, like, the questions, like, confidently and with, like, full honesty . . . I don’t know, like, when I saw my GP I didn’t tell him the whole truth, just because I didn’t feel that comfortable.

Organisational features of the sites were also important in determining how welcoming a service was. In the earlier Service development and Building and maintaining engagement subthemes, we discussed how the service leads appeared to have a high level of leadership skills, and described how staff often worked hard to make services convenient to attend, or took the service directly to the families. There were also some comments about the premises that the services were delivered from. At two of the voluntary sector sites, staff explained that particular efforts had been made to ensure that the physical surroundings were attractive to young people, featuring, for example, bright colours and soft furnishings or having music playing. On the other hand, CYP at one site commented that the rooms in which the groups were held were somewhat small. The physical features of the premises, however, were overshadowed by spontaneous references that CYP and parents made to the welcoming ambience. Indeed, there was often blurring between organisational and individual staff attributes: ‘welcoming’, ‘friendly’ and ‘non-judgemental’ were used by CYP and parents to describe both the staff and the general atmosphere or ambience of the service, often in contrast to other services they had experienced.

Activities to build skills for self-care

The use of skill-building techniques and opportunities to practise such skills in their daily lives was considered an important feature of all six services. Children, young people and parents outlined a variety of practical activities and techniques, including games and exercises, which they had used in formal sessions (in vitro) as well as at home and in school (in vivo). Fittingly, given the nature of this study, most of these activities and techniques were designed to assist the participants in managing (self-managing) the CYP’s condition or problems. For example, in the self-harm group (site 1), young people were supported in devising coping strategies other than self-harming; in site 2, children practised communication skills at home that they had learnt at Dina School; and in the family support group (site 4), families were encouraged to action plan and set goals. For younger children in particular, a large proportion of their narratives consisted of their recollections of the various games and activities in which they had taken part. The youngest children stated that they had gone to Dina School to learn things and that they got ‘prizes’ (plastic chips) for doing well at the activities there:

And what did you get a chip for?

When I’m doing stuff nice.

Many of these activities focused on identifying thoughts and emotions or on learning techniques to relieve anxiety or calm anger. These had been introduced by staff at the sites and also practised by children outside the sessions. The older primary school-aged children seemed to have more insight into the purpose of the activities and techniques:

We used to do these words about temper and then after that to calm me down we used to play a few games.

Child

The parents who were interviewed also seemed to appreciate practical techniques and talked at length about implementing these at home. Reward systems including sticker charts were mentioned frequently, along with techniques designed to help children manage tasks and express their feelings:

I think it is important to have a task and try and complete it and stuff.

Young person

I think she is responding well to positive rewards . . . she loves the reward charts . . . I give her a sticker just on her T-shirt if she’s done something really nice.

Parent of younger child

And from an interview with another parent:

Are there any things in particular that you find helpful?

Well there was the words on the fridge . . . the fridge magnets spelling out, ‘I’m not happy’, ‘I’m . . .’

‘. . . sorry’.

‘Sad’, ‘I’m sorry’ and . . .

‘I love you’.

‘I love you’, yeah. We did that for a while. We lost all the pieces, like, we did it for a while.

Some games and techniques were ‘prescribed’ as part of the manual or workbook for the course. However, most of the activities allowed some personalisation such as tailoring rewards to the individual child. For example, staff at the family support service (site 4) found out what children were interested in or enjoyed and then looked for local activities or clubs they could join, such as football or the Boys’ Brigade. At the self-harm site (site 1), young people put together their own ‘tool boxes’ of distraction techniques, memorabilia and keepsakes that helped them reduce urges to self-harm. The service also provided support for caring for wounds that arose from cutting, and advice on ways to camouflage scars, with an organisation specialising in this visiting the group.

I’ve got this thing called a tool box in my room that I put all my different distraction techniques inside so, like, if I’m having a bad day, even if I’m not having a bad day . . . I could use the different things. And different things work at different times, so one time, like, writing out your feelings might be enough, or drawing might be enough, but sometimes you have to use a range of things to minimise that urge and sometimes nothing works, but . . . at least I’ve tried.

Young person, interviewed with parent

The programmes delivered at sites 2 and 3 (the Incredible Years sites) and site 6 (the anxiety/depression support service) were supplemented by videos for participants to watch that contained vignettes or examples of role plays to augment the situations or techniques that the programmes considered. These were criticised repeatedly by staff and service users for being outdated and participants disliked that they had been filmed in the USA as they would have preferred British accents. One service user found them ‘idealised’ in that the actors looked too ‘perfect’ and suggested that videos featuring real people, not actors, would be more helpful.

Sharing experiences and peer support

Although a key aspect of all of the services at the case study sites was introducing CYP and their families to stock tools and techniques to help them care for themselves, an important aspect of skills building was supporting CYP and parents to discover self-care techniques for themselves. Often, this happened with the support of their peers, especially in the sites operating group-based approaches where other group members could play a key role in generating ideas or giving feedback. For example, children in the Dina School groups (site 2) took part in role plays and other interactions that helped them to think through how they could handle things that children might find difficult, such as sharing:

[Using a puppet to act out] snatching a book, [then discussing], is that kind?, is that unkind? . . . then after a while you’d get them to bring their problems in, you know, if you have a problem on the yard [playground], you know, if someone’s not listening to you, or doesn’t want to play, well, eventually they would come up [and talk about it with the group].

Staff member

Similarly, users of the online support service (site 5) frequently exchanged tips and techniques. Postings on the message boards often included users sharing self-care tips and suggestions with each other, and describing various relaxation or distraction techniques and activities such as reading, knitting, writing, drawing and exercising:

Listening to music is a huge relaxation thing for me, also drawing/writing – either a story, poetry or in a journal. Do you enjoy reading? To begin with i didnt have much concentration for reading, but now I am really enjoying reading again, and i find if i start reading after a meal when i am anxious that I can get lost in the book, and before i know it half an hour has passed. I’ve also tried knitting, but i’m a bit of a perfectionist so that didnt go so well. Hope you are ok, and that some of these tips may help you.

Verbatim message board posting

The following were taken from two other sites:

They discuss coping strategies, like, they’ll say, like, if you get the urge to self-harm, what do you do? And they, sort of, come up with, ‘well, I’ll go and do this, I’ll go and put my favourite music on, my film, go for a walk, walk the dog, talk to my mum’ . . . and I do think it’s very beneficial.

Staff member

I enjoyed the discussion. And what was nice is the feedback we gave each other . . . Because I think, that does you good, to have other people sometimes go, no, you’re good at that or why don’t you try this or why don’t you try that? . . . because you put it into practice with other people, it works better than you just reading books.

Parent

Although the sharing of practical tips and techniques was an important element of peer support, the sharing of the experiences among CYP and parents was perhaps more beneficial. As discussed in the first theme, Having a mental health problem, some participants had not talked about their difficulties with other people prior to coming into the services and had felt alone with their problems. Staff members and attendees at several sites talked about the emotional benefit of being in a group with people who had had similar experiences, in that people who had previously felt isolated with their problems, or uncomfortable about seeking help for them, felt less alone:

It was actually amazing really; firstly because you realise that what your child was doing was normal and that all the other children were doing the same things.

Parent of younger child

I think they listen more to their peers, their peers is pivotal, isn’t it, in adolescence, so I think the peer groups are very important, and that sense of belonging is good. And often . . . I’ve found that every single youngster who attends a group, is excluded from some aspect of their life, either from their family or from a peer group, because they’re bullied or they bully or, you know, they don’t fit somehow . . . So I guess being together with other young people helps you feel, you know, you’re not on your own with this.

Staff member, another site

As well as receiving support from others, satisfaction gained from feeling that they had helped others was also important to some participants:

It was good for [my daughter] to see that she had helped other people, she liked that; that helped her as well.

Parent of older child

As well as perceiving that the sharing of experiences was helpful, staff and parents felt that having a mixture of experiences and people at different stages of illness or recovery within a group provided additional benefits. The self-harm and parenting groups brought together people from different social groups who would normally not socialise together, but who had experienced similar problems, and this seemed to be helpful in terms of giving people a wider perspective on their problems, or perhaps developing empathy:

[They] have a good understanding of each other’s needs . . . it runs well because it’s a mixture . . . they’re all at different stages of their recovery . . . we’ve got young people who have been established for a long time and . . . have almost recovered, others are at the stage of contemplating, sort of, change and in the process of recovery and some that are not ready to change. And that’s, as I say, a very supportive group . . . those that are almost recovered, they do get and understand where [those who are less recovered] are coming from.

Staff member

Very heterogeneous . . . you’d have youngsters in local authority care, and then you’d have very upper middle class young people, who were anxious about exams and stuff like that. And that actually works as well, because kind of sometimes it’s good to see people in other situations . . . you thought your situation was really bad, but actually, there’s people who are worse off, and something about that kind of gratitude about what you have got.

Another staff member, same site

The participant quoted below, who described herself as ‘middle class and educated’ seemed to agree that this could be a benefit:

It sounds awful, but people that you might not necessarily have mixed with before, because you wouldn’t have had the chance, and got to know them . . . different social spheres . . . all of that’s broken down . . . [there can be perceptions that] . . . some backgrounds are better, supposedly, than others. And actually, you know, one of the mums in particular, who hadn’t got any further education . . . was just a fantastic mum. Whenever we’d . . . learn a new principle, when we’d feed back the next week, her and her partner had discussed it, and had both tried it . . . And it was nice to hear about other people, and there is a range of us, you know, people with partners, people without.

Parent of younger child

The ‘sharing experiences’ aspect of peer support was complemented by others including the social aspect and what might be termed an ‘empowerment’ aspect. Regarding the social aspect, service users at the self-harm (site 1) and parenting (site 3) sites enjoyed attending the groups and especially liked the social aspect of getting to know others in the group. The parenting course (site 3) had an arranged ‘buddy system’ where parents were paired up and encouraged to telephone each other between sessions to talk about how they were getting on with their tasks or exercises, and some of these parents also reported forming friendships and continuing to meet socially after the programme had finished. From the focus group:

I prefer group talks, like, talking in groups, instead of taking medication . . . [at first] I was an inpatient so I was really struggling and . . . just having the support off people in the group was what I needed.

We’re good for each other!

We just, like, support people with what’s been going on in their week, and stuff, and it’s nice, because we’re like a little family.

We are like a family aren’t we?

The social benefit of groups was also reported by the staff member at the eating disorders message board service (site 5):

[T]he message board is a really good place to start forging links with other people in a very safe and supported way and it helps to build their confidence . . . when they suggest something to somebody [and they] come back and say, ‘that was really good, I did that and it really worked for me’ [they] make very strong bonds with each other.

Staff member

Some of the group activities also appeared to help empower the CYP. For example, the self-harm groups (site 1) helped participants develop empathy by encouraging the young people to take it in turns to chair the group. Each group began with each member giving an update on how his or her week had been, and a role of the chair was to ensure that each person contributed. Staff observed that some young people in the group seemed unhappy or annoyed when they would have preferred not to contribute but the chairperson still asked them to share their experiences with the group. The staff member who described this thought that this could help young people to appreciate the difficult task that staff sometimes faced in encouraging young people to engage in ‘talking therapies’, and that sometimes encouraging someone to contribute in a group, even if they seemed not to want to, could help develop empathy and leadership skills in the young person.

Although our data demonstrate that peer support is beneficial, the risk of ‘contagion’, or triggering self-harm or problems with eating, was perceived as a key challenge by staff at the self-harm (site 1) and eating disorders (site 5) sites. Staff at both of these services saw risks in group work and worked actively to manage such risks. Unlike the parenting site (site 3), where parents were ‘buddied’ with another parent and encouraged to support each other between sessions, young people at other sites were discouraged from having contact outside of the service. At the self-harm site, young people were actively discouraged from meeting outside the group and, at the online site, interaction was controlled through moderation of the message board postings. This involved screening and editing posts to ensure that they did not contain contact information.

Some people who had attended the self-harm groups (site 1) and parenting courses (site 3) remembered feeling nervous about attending their first session. As it was the first time they had attended anything of that nature, they were unsure what to expect and were nervous about having to talk about their difficulties in front of other people. They felt that groups were only helpful for people who were at a stage of their illness or recovery where they were ready to attend a group. Our interviewees had all settled into the groups and had positive experiences, but they thought that the group setting would not be suitable for everyone; some people simply would not want or choose to discuss their problems in a group setting. Staff at the self-harm group identified managing dominant personalities as a further challenge that they had to be aware of, in order to make sure that all group members had a chance to participate during each session.

Time and attention

Giving CYP time and attention appeared to be a valued characteristic of the services. At four of the six sites, participants valued simply spending time with children, especially when this involved enjoyable activities. A key principle of Dina School (site 2) and the parenting course (site 3) was giving children positive attention. The Dina School groups allowed for each child to receive more attention than would be possible in a regular class. Parents felt that their children enjoyed being in the group and that the environment was more realistic than the regular classroom for them to learn about talking about their feelings. Parents who attended the parenting course found that setting time aside for their children was helpful, including scheduling time to play with them and also having activities together as a reward for good behaviour, and reported that their children were calmer. At the family support service (site 4), a worker described how during school holidays, instead of having appointments with children at their school, she had collected them from home and brought them to the centre, and had been surprised by how much they enjoyed this:

Often I’ll go and pick the children up and bring them here, which they just think is the best thing in the world . . . it’s not that exciting, but it’s just something different . . . that’s what some of them say.

Staff member

Participants of different ages seemed to enjoy positive attention, perhaps away from a school or the clinical or home environment:

Another thing as well that sticks in my mind about the group is when [staff member] actually took us out once into town . . . as a group . . . just to do something nice, instead of being stuck in a small room that’s quite clinical . . . it was good.

Young person

At one site, young people recalled their experiences of being inpatients, when some staff would come to talk to them, but often at inappropriate times or not for long enough. At another, family members of younger children appreciated project workers talking to the children themselves.

I think that young people often really have appreciated the fact that someone has taken the time to listen to them and hear their side of the story.

Staff member

An older sibling of one of the children at the same service singled out the way that the project worker had helped her brother to understand his behaviours and the emotions attached to them, and what might affect these:

When [child’s name] used to get angry we used to talk to him and try and calm him down . . . but we never used to speak to him about why he got angry . . . I think because [project worker’s name] spoke to him and broke it down into steps that he could understand . . . and then she used to ask him stuff about his friends and his family.

Sibling

Although CYP and parents appreciated the time and attention that staff at the sites gave them, staff mentioned that operational constraints sometimes prevented them from providing as much time as they would have liked. The manuals for the Incredible Years services (sites 2 and 3, Dina School and parenting groups) were prescriptive, with themes and associated activities being set out for each session. However, although staff at these sites liked the fact that all the materials for the sessions, and also the letters to send home to parents, were provided (as this made the course easy to deliver), they felt that there was too great a volume of material, and that they had often had to leave out certain aspects in order to complete the sessions on time.

Sometimes it was hard to keep their attention, because we’ve got a programme to deliver and it’s quite a lot to fit into those 2 hours, so sometimes [we] would look at it before and think we’re never going to fit all that in so we’d think, right, we won’t do that activity today, we’ll do that next week and try to fit in an activity where they’d be moving a bit more and try to adapt it a little bit.

Staff member

Both CYP and parent participants at several sites mentioned that they would have preferred more sessions or time with the service – a statement also echoed by staff at some sites – but financial constraints prevented this from occurring. In particular, staff at one of the voluntary sector sites outlined how a scheduled 15-week programme had to be reduced to 12 weeks because of financial and resource restraints.

Now we have this 12-week programme and it used to be a bit more like 15 weeks and we used to be able to be a bit more flexible with what each different family needed . . . I feel I used to do a lot more . . . therapeutic work with the young people and creative work, and some of that has had to be cut because we’re having to be a bit more focused on where we need to get to, which is a bit of a shame . . . I used to do a lot more kind of crafty things . . . than I feel that I’ve got the time to really do now.

Staff member

Sustainability

An evaluation of CAMHS innovation projects over a decade ago203 identified a number of key elements in services that work, including having a capacity to keep in touch with CYP over the long term and offer additional, short-term intervention or support if needed. This message of sustainability in self-care support interventions has been repeated more recently in both adult mental health12 and CYP’s mental health,204 in mental health promotion in schools200 as well as in a general review of the evidence on self-management. 205 From a service user perspective, it has also been neatly summarised by a SAG member who argued that the principle should be one of ‘give us the skills, remind us of the skills and help us when we get stuck’. Given that our effectiveness review focused on short- to medium-term follow-up (6- and 12-month time points), it can, to some extent, also be seen as a ‘sustainability’ review. Our findings thus suggest that few of the self-care support interventions examined were sustainable, as effectiveness was only moderate at 6-month follow-up and it subsequently declined at 12-month follow-up. Moreover, our effectiveness review found that interventions that were longer also tended to be more effective, an observation that further suggests that a longer-term commitment may be beneficial.

Different perspectives

One clear finding from our data is that, in evaluating mental health self-care support for CYP, a range of different perspectives surface. Not only is the perceived value of an intervention or service dependent on who is asked, but some biases in the overall literature imply that certain information about interventions and services is valued more than others. Regarding the former, our satisfaction data, for example, noted that adults (parents and teachers) often rated interventions better than CYP, a finding that has been observed elsewhere,206 and in both our satisfaction and qualitative data, there was evidence that CYP did not agree among themselves about the value of a particular intervention or which specific components were the most valuable to them.

Regarding the latter, the dominance of trial studies compared with perceptions studies suggests that an effectiveness perspective is generally more valued. In addition, the observation that most outcome measures in the effectiveness studies were concerned with mental health symptomatology (see Chapter 2 , Table 4 ) also demonstrates a particular (i.e. clinical) perspective regarding the importance of particular outcomes. This contrasts with our previous NIHR study exploring children’s long-term physical health conditions,15 where a much wider range of outcome measures were considered, including health service use, which was notably absent from all of the studies we included. Moreover, the effectiveness of an intervention or service can be measured across many domains, such as a child or young person’s ability to cope or move on from a difficult situation, and not just in improvements in symptoms, and there is evidence that CYP’s problems can be addressed more effectively if their strengths, self-efficacy and resilience are promoted. 204 This is a point that the recent CYP’s IAPT programme seems to have conceded, given that the programme considers a range of outcomes including strengths, general well-being, employment, and education uptake and attendance, as well as mental health symptoms. 31

Flexibility versus fidelity

Flexibility appears to be a key element of self-care support. Inflexibility of services was a common criticism made about standard CAMHS services by CYP and parents in the case study sites, and flexibility was identified as a valued feature of self-care support in both our previous NIHR report15 and the recent NIHR project on self-care support in adult mental health. 12 However, almost all of the interventions considered in our effectiveness review were manualised, a status that implies that there are ‘rules’ as to how an intervention should be delivered. This raises the issue of fidelity, a concept that it is somewhat at odds with flexibility. Compare, for example, the case study sites that used The Incredible Years, where fidelity to the manual appeared to be paramount (to the service lead, at least), to the perceived inflexibility of some of the manualised interventions that was evident in the perceptions review.

The issue of fidelity in manualised interventions is interesting as there is no consensus on its importance in achieving the best outcomes in psychosocial interventions. The evidence for fidelity in parenting interventions, in particular, is equivocal, with some arguing that it is a necessary component for effectiveness,207 while others argue that there is scant evidence that manuals improve treatment effects. 208 A recent discussion paper from the British Psychological Society209 attempted to reconcile these two positions in the context of the real-life service delivery of parenting programmes, arguing that fidelity is important but not at the expense of acceptability, and that fidelity should be redefined as those principles critical to effectiveness, rather than being seen as the duplication of an original in vitro model.

Choice versus constraint

In attempting to integrate the effectiveness and perceptions reviews at the end of Chapter 3 , we concluded that a key message might be one of service providers constrained by homogeneity, in terms of largely providing manualised, group-based, face-to-face (cognitive–)behavioural interventions. Homogeneity of intervention erroneously assumes homogeneity of recipient, an assumption that has been challenged in many aspects of CYP’s health-care delivery. For example, this has been pointed out in relation to human immunodeficiency virus/acquired immune deficiency syndrome (HIV/AIDS) prevention programmes for CYP210 and the mental health needs of homeless CYP,211 and in a review of CYP’s general perspectives on health and health services. 212 Treating patient groups as homogenous purely on the basis of qualifying for a specific diagnosis has also been critised,208 a fair point given that CYP diagnosed with a specific mental health condition such as anxiety will vary immensely in their needs, their experiences and the resources they have, as will their families. In terms of effectiveness, it might therefore be more appropriate to ask ‘what works for whom and in what circumstances?’ rather than merely asking ‘what works?’.

However, assuming that different things work for different people brings about the question of choice. Choice does not necessarily mean choice only in terms of specific interventions or services. It might also mean choice in terms of specific elements of the service: face to face or telephone, or group or individual. It might mean choice in terms of location: home or school. It might mean choice in the referral process including the opportunity for self-referral, a feature of two of our case study sites and a salient observation in the recent NIHR study on self-care support in adult mental health. 12 Finally, it might mean choice in terms of who makes that choice: the child or young person, a parent or, as is too often the case, the staff providing the service. In other words, choice should mean taking into account the CYP’s and their families’ individual needs. 203

In raising the issue of choice, it is not to say that effectiveness is unimportant, just that some generally effective interventions might not work for some CYP. Standard effectiveness tools such as meta-analysis are not able to assess the issues of accessibility and engagement very well, where choice might be a more significant issue and driver of outcomes. A pragmatic compromise, therefore, is that services should still consider effectiveness when providing interventions but that there should also be a reasonable choice in evidence-based approaches. To some extent, this has been recognised by the CYP’s IAPT programme,31 in that initially only cognitive–behavioural and parent training (the dominant models in terms of effectiveness data) were available but the programme has recently been extended to include other interventions with evidence of effectiveness, such as systemic family therapy and interpersonal therapy. However, even if a range of evidence-based interventions are available, the issue of who makes the choice remains. Moreover, if it is to be CYP and their parents who make that choice, there is a need to ensure that sufficient information about interventions and services, including information about their effectiveness with different CYP, different problems and different circumstances, is available so that CYP and their parents can make informed choices.

A further issue in discussing choice, and a key element of the overall acceptability of a service or intervention, is an appreciation of the CYP’s or their parents’ readiness to self-care. 12,24 Although there was little explicit reference to readiness in our findings, there were definite hints from the case study data that readiness could be an issue in terms of accessing self-care support, and it was a noticeably absent concept in most of the interventions examined in our literature reviews. Our previous NIHR study15 also picked up some hints about readiness in relation to decisions about whether to access services or not. A related concept to readiness is health activation,213 a concept we will discuss further in the Conceptualising self-care support in CYP’s mental health section later in this chapter.

Choice and groups

While discussing choice, it is worth talking about one particular feature of our overall findings. Group interventions and services dominated all of our data sets, although the groups varied from relatively small groups to groups as large as whole primary or secondary school classes. It is not difficult to understand the popularity of groups, given the peer support and shared experience advantages of group approaches in health care identified in this study, our previous study15 and elsewhere,214–216 their relatively low costs compared with individual attention, and the evidence that social support networks are associated with better mental health outcomes, fewer problem behaviours and more health-promoting behaviours in CYP. 217

However, there was some evidence from our meta-analysis that individual interventions may be more effective than group interventions for CYP, though the difference in effect size between the two (a difference in SMD of –0.11 in favour of individual interventions at 6 months) may be insufficient to counter the additional benefits that groups provide and the additional costs involved.

Furthermore, there is evidence that there are some risks to CYP in group interventions. A recent review on the role and impact of social capital on the health and well-being of CYP217 reported that, in certain circumstances, social support networks could increase participation in health-risk behaviours such as alcohol and tobacco misuse, an observation backed up by our case study reports concerning the risk of ‘contagion’ in eating disorders and self-harm groups.

Location of service

The evidence on where services should be located is somewhat equivocal. Our effectiveness review findings suggest that the setting (clinic, home, school, etc.) is a largely irrelevant factor in the effectiveness of an intervention and no argument in favour of any particular location emerged from the perceptions review. This suggests that other factors need to be taken into account when considering the location at which self-care support should be delivered. The case for hospital-based services is generally weak. A recent King’s Fund paper219 argues that health service delivery in general needs fundamental change and that hospital-based care should be reserved for those who cannot be treated more appropriately elsewhere. Our case study data also suggested that inpatient hospital care was not particularly well liked by participants. In terms of mental health provision, current government policy9 is that people with mental health problems – including CYP – should be treated in the least restrictive environment. For CYP, the least restrictive environment usually means school, home or some other community-based location such as a clinic or community centre. All of these locations have advantages and disadvantages. Clinics, community centres and schools often have town centre locations close to transport hubs, a factor seen as important in terms of access214 though our case study data implied that physical access was not necessarily a critical factor for participants. Formal clinics, however, may, at best, reinforce the medicalisation of the CYP’s problems and, at worse, stigmatise the CYP. The family’s own home offers convenience and few access problems but not everyone is comfortable with professionals intervening at home. Perhaps the best candidate for location is schools: schools are part of the CYP’s normal experience, and the fact that most of the studies in the effectiveness review were school based demonstrates that schools are feasible places for delivering such interventions. Moreover, there is evidence that schools are substantially more cost-effective at delivering mental health interventions for CYP than routine CAMHS. 220 On the other hand, there was some evidence in our case study data that schools are not necessarily prepared for the task (as they came in for criticism, along with inpatient hospital care, from some of our participants), and there is some evidence, at least in terms of preventative approaches,200 that schools may need to have the right ethos in place to be centres for mental health support.

Use of technology

Given the commonly held view that CYP are inherently comfortable and confident with new technologies – that they are ‘digitally native’ – there was surprisingly little use of computers and internet technology in the interventions and services we examined. None of the interventions examined in the effectiveness review used computer or internet technology, there were only three such interventions in the perceptions review and only five were identified from the service mapping. This might be because the peer support and shared experience elements of face-to-face, group self-care support are harder to replicate online (though not impossible, as the eating disorders online support case study site has demonstrated); or it might be because NHS services in particular are inherently conservative and have not yet caught up with the technology. 221 On the other hand, it might be that the assumptions about CYP being digitally native are incorrect180,222 – that CYP have no greater a preference for computers and internet technology than adults or that, again, homogeneity is being assumed in CYP when in fact some may embrace technology more than others. This is a salient point because it brings about once more the issues of choice and readiness, in that some CYP may not be ready to use technological interventions and some – as we found in our perceptions review – may not particularly like them. Nevertheless, there is evidence from elsewhere that technology can help in engaging CYP in mental health services: for example, the use of text messaging to reduce ‘did not attend’ rates. 214,223

Conceptualising mental health

When considering how self-care support might be conceptualised in CYP’s mental health, it is not surprising to find that conceptualisations of ‘mental health’ can also be somewhat vague. For example, when reporting the case study findings, we noted that conceptualisations of mental health seemed to be connected to the CYP’s age, with mental health being conceptualised in younger children as a lack of confidence or difficulties with emotion and behaviour, and in older children as diagnosed conditions or severe mental health problems. The reasons for this are far from simple; being ‘diagnosed’, for example, may be down to capacity and configuration of local services, the expectations of the child and/or parent, and social constructions of mental illness. There is, nevertheless, a balance to be struck in conceptualising a child or young person’s difficulties as mental health problems or not. A diagnosis is often a criterion that triggers access to a potentially helpful service or it can act as a filter for different interventions and services, as there is evidence in self-care support that different conditions benefit from different approaches. 205 Yet a diagnosis can also have negative connotations (leading to stigmatisation and exclusion, for example), as one of our case study sites made clear when it asked for the removal of any reference to ‘mental health’ from our recruitment materials, even though it was delivering an ostensible mental health service. Maybe choice needs to be considered here too in that CYP should be allowed a choice as to whether they accept a diagnostic label or prefer to focus on, for example, identifying specific needs, feelings and opportunities to develop strengths. This is also reflected in the current debates expressing concerns about the increasing ‘medicalisation’ of emotions and behaviours in CYP. 225,226

It is also worth asking whether CYP’s mental health problems and needs should be considered as long-term conditions because the focus for many families is, as it is for the families of CYP with long-term physical health needs, often more about quality of life and living (well) with a condition than it is about ‘curing’ the condition. Mental health self-care support interventions for CYP need to be considered in a context where the majority of adult mental health problems have their origins in developmental problems in childhood;227 yet, as our study illustrates, what tends to be on offer are relatively short interventions with erratic follow-up periods (often merely for research purposes) for CYP who have problems and mental health needs that are likely to extend into adulthood. Clearly, there is a relationship here with our earlier discussions on sustaining self-care, and a conceptualisation of CYP’s mental health in a longer-term context may help address some of the shortcomings in the sustainability of specific interventions and services.

A conceptual model of mental health self-care support in children and young people

In modifying the typology of self-care support (see Chapter 4 , Table 19 ), we argued that self-care support in CYP’s mental health can be seen as a process. There needs to be the acquisition of knowledge and skills, which require input and motivation on the child or young person’s (or parent’s) part; opportunities to practise and consolidate these requisite skills, with support and facilitation from a self-care support agent; and some positive outcomes for the child or young person and his or her family. Moreover, as we earlier argued, there is some evidence that this process should be long term rather than short term. On the basis of our findings and the discussion herein, a conceptual model of mental health self-care support is proposed in Figure 14 .

FIGURE 14.

A conceptual model of mental health self-care support in CYP.

We have discussed a significant number of the elements of this model already, including effectiveness, acceptability, readiness and outcomes. Choice and flexibility, though not explicit in the model, permeate the whole model in that various elements of it, such as the types of skills on offer and the means of acquiring them, the outcomes used and even the self-care support agent, can all be tailored according to the individual needs of the child or young person and his or her family. Two critical elements of the model – skills and the self-care support agent – warrant further discussion, however.

Regarding skills, it seems that skills and the activities that support their development are a central feature of effective self-care support programmes12,15,200,203,205 and of recovery in CYP’s mental health in general. 214,217 However, as mentioned above, choice and flexibility need to be considered when identifying these skills and providing the support CYP and their families need in acquiring them. This raises the issue of the personalisation of health-care services and the associated issue of personal health budgets,219,228 although it seems these two issues raise tensions similar to the ones we identified when discussing effectiveness and acceptability, in that the patient preferences inherent in personalisation have to be reconciled with a co-existing framework of ‘payment by results’ where effectiveness is the key concern. 229

Regarding the self-care support agent, the principal issues that need discussing are the attitudes and attributes of the professional or lay person taking on this role. This involves some discussion of both the attitudes of the agent and their interpersonal skills, as well as some comments about education and training.

Qualities of the self-care support agent

Our case study data suggested that the personal qualities of the staff providing the self-care support were important, with staff generally seen to have positive attitudes such as being non-judgemental, welcoming and an active listener. There were also some hints in the perceptions review that the quality of the relationship with the agent and the expertise of the agent may be as important as, if not more important than, the intervention itself. This is a common theme in the literature: participants in our previous NIHR study15 remarked on the approachability of the self-care support agents and their abilities to listen and be non-judgemental; a recent Health Foundation review on self-management205 found that the attitudes and skills of staff can have a significant effect on perceived levels of engagement and support; and in CYP’s mental health, the ‘therapeutic alliance’ has been identified as a significant but neglected factor in treatment. 230

Leadership in self-care support services may also be an important personal and organisational attribute, and we saw evidence of strong leadership in our case study sites. Effective and consistent leadership was identified as a key element of ‘what works’ in CAMHS provision,203 and the recent NIHR study on self-care support in adult mental health12 noted that charismatic leadership was a feature of many of the projects examined.

Another key personal and organisational attribute is child-centredness. As in our previous NIHR study,15 being child-centred was a feature of most of the case study sites. The general literature indicates that effective patient-centred care requires two specific elements: the straightforward eliciting and discussion of the patient’s perspective and ‘activation’, i.e. a deeper understanding of the individual’s motivation and readiness to become involved in a service, change their behaviour or self-care. 231 This suggests that for the self-care support agent to be truly child-centred, they need to understand not only the CYP’s problems and needs but also recognise the extent to which they are ready. This raises the question of whether or not a readiness measure of some sort, whether standardised and formal like the Patient Activation Measure213 or something more informal like simply asking the question ‘do you feel you are ready?’, should be routinely used in health-care practice.

From an organisational perspective, child-centredness requires – as we saw in the case study sites in this study and in our previous study15 – a friendly, fun, welcoming atmosphere where the CYP’s views are listened to and respected. Above all, child-centredness requires a partnership between the provider and family, one that demands that CYP and their families be seen as part of the care team or, to use a phrase currently gathering momentum in adult mental heath care, as ‘coproducers’ of their care. 232

A consideration of the attributes and skills of the self-care support agent inevitably raises the issue of education and training. None of the attributes and skills required of the support agent necessitate mental health training, and it was interesting to note that our effectiveness data suggested that there is no specific advantage in the agent being mental health trained, although there is some evidence82 in the literature to counter this, and it was notable that around three-quarters of those delivering interventions in the effectiveness studies were specifically trained for the role. What seems to be essential is the ability to understand the CYP’s perspective, and to focus on their needs and the needs of their families, rather than on the needs of the service or the people leading the service. This may require attitude changes among CAMHS staff and those managing services and, certainly, if an eclectic range of interventions is to be made available within CAMHS, then it will require opportunities for staff to retrain and develop new skills. 204 In terms of attitude, in England and Wales at least, our findings suggest that the voluntary sector seems to be ahead of the NHS and other statutory services. However, as our case study data indicate, there is also excellent service provision in the NHS. Thus, it may not matter which sector delivers the service, so long as the staff are child-centred, have opportunities for training and development and are supported by strong leaders. Moreover, there could be considerable benefits for CYP with mental health needs and their families if opportunities are made available for the NHS and the voluntary sector to learn from each other.