Youth social behaviour and network therapy (Y-SBNT): adaptation of a family and social network intervention for young people who misuse alcohol and drugs a randomised controlled feasibility trial

Substance use among children and young people

The two most commonly consumed substances by young people, accounting for 90% of treatment admissions, are cannabis and alcohol. 1 Statistics on drinking and drug use among young people are divided between those drawn from surveys of school-age children and those drawn from surveys of adult populations, including young adults. Among children of school age, although the proportion of those drinking at all has dropped slightly since 1988, the average number of units consumed increased markedly between 1990 and 2006 and has since stabilised at this level. 2,3 On the other hand, contrary to media reporting, average alcohol consumption among young adults (aged 16–24 years) has fallen since a peak in 2000–2. Nonetheless, 15- to 16-year-olds in the UK still come third highest out of 36 European countries in the proportion self-reporting drunkenness in the past 30 days (26%). 4 Cannabis use has also shown a decline among school-age children since reaching a peak in 2000–33 and a longer decline among young adults since 1998. 5 However, the UK is among the 10 European countries with the highest proportion of 15- to 16-year-old students reporting smoking cannabis within the past 30 days. 4

Since the commissioning of the current study, there has been increasing public concern about the use of novel psychoactive substances (NPSs) or ‘legal highs’ among young people. Accordingly, new questions were introduced to the smoking, drinking and drug use school survey conducted in 2014 in participating secondary schools across England, with 2% of pupils in Years 7–11 reporting using a NPS/legal high in the past year. 3 In addition, employing a generic definition of NPS use for the first time, data from the most recent Crime Survey for England and Wales show only slightly higher NPS use among 16- to 24-year-olds (2.8%) in 2014. 5

Early onset of drug use, including alcohol, has been associated with later problematic use. 6,7 Early onset of use and early hazardous use have also been associated with a range of other problems including risky sexual behaviour, injury, antisocial behaviour, violence and changes in brain development. 8,9 Moreover, in considering the impact of substance use on the family, research has shown that substance use among young people can adversely affect relationships with parents, carers and other family members,10 but also that family involvement in interventions can influence the course of the problem in a positive way. 11

Family-based interventions

Research has highlighted the pivotal role that families play as both a risk for, but also a protection against, substance-related problems. 11 As a consequence, a range of preventative and treatment approaches has focused on the family. In the UK there has been a strong focus on preventative programmes. A systematic review by Foxcroft et al. 12 identified the Strengthening Families Programme (SFP),13 developed in the USA, as the most promising, with positive outcomes in both the short and the long term. Emerging findings from the application of this model to the UK context have also been promising. 14

Turning to the treatment field, reviews of evaluations of family-based interventions with young substance users have tended to show that the most empirically supported family interventions are multidimensional family therapy (MDFT), multisystemic therapy (MST), brief strategic family therapy (BSFT) and functional family therapy (FFT) (Carmel Bennett, School of Psychology, University of Birmingham, 2013, unpublished review). Eleven literature reviews were included in a two-part review undertaken as part of the current feasibility study of Y-SBNT (Carmel Bennett, unpublished review). Although there is growing evidence for the effectiveness of these interventions (particularly MDFT and MST), problems exist concerning initiation, engagement and retention, treatment decay and translating research into practice.

Barriers to implementation in the UK

In terms of translating research into practice, specialist training is needed for family interventions, which can be an obstacle to their implementation. 33 Public Health England statistics suggest that < 1% of interventions with the under-18s consisted of ‘family work’. 1 The large majority of young people with substance misuse problems receive psychosocial or harm reduction interventions that focus on the individual user and do not engage family members. Likewise, a survey conducted in the UK with services for adult family members showed that even those family interventions recommended by the National Institute for Health and Care Excellence (NICE),34 such as behavioural couples therapy,35 are rarely implemented in services. 36 It should also be noted that some of the components and techniques described above and in Table 1 assume a level of resources – and a level of assertiveness – that may not be achievable or acceptable in a UK context. Finally, there are problems with defining ‘the family’, a contested concept that carries implications for the delivery of family interventions. 37 Young people with substance use problems frequently come from disrupted families and may be looked after by single parents, grandparents, other relations or the state (e.g. Lloyd7 and Boys et al. 38). Traditional, systemic family approaches may be difficult to deliver in such situations.

In summary, the research evidence shows that there is a high prevalence of substance use among young people in the UK. Early onset and high levels of use are associated with a range of negative outcomes, including increased risk of later problematic use and dependence. A growing body of research has identified family interventions to be effective at treating young people’s substance use problems and a range of techniques has been identified to encourage engagement and retention. However, despite this evidence, take-up of family-based approaches, at least in the UK, has been low. A key factor appears to be the resource-intensive nature of many family interventions, making them difficult to implement and deliver in many service settings, especially in the context of funding cuts to drug and alcohol services for young people. 39 Another potential barrier may be the cultural adaptation of approaches developed in the USA to a UK setting. There is a growing awareness of the need to adapt evidence-based treatments to different cultural groups and settings to ensure successful implementation. 40–42 Finally, it should be noted that a key distinguishing feature of drug treatment with young people is the coercive context in which it frequently takes place. This may encourage physical attendance but does not necessarily encourage therapeutic engagement.

Social behaviour and network therapy

Social behaviour and network therapy is an intervention developed in the UK that has been shown to be effective with harmful drinkers43 and is recommended in recent NICE guidance. 44 The original SBNT intervention was developed and tested as part of the UK Alcohol Treatment Trial (UKATT) conducted in UK alcohol treatment services. 43 It was later developed further for use with people with other drug problems. The conceptual underpinning of the approach is the positive influence of social as well as more addiction-specific (or abstinence-related) support on the improvement and eventual resolution of addictive behaviour. General social support, alcohol-specific social support and the drinking behaviour of the social network of alcohol users, for example, have all been shown to be unique predictors of positive alcohol treatment outcomes (for examples see Beattie et al. ,45 Havassy et al. ,46 Longabaugh et al. ,47 McCrady,48 Wasserman et al. 49 and Mohr et al. 50). SBNT brings together elements of network therapy,51 social aspects of the community reinforcement approach (e.g. Meyers et al. 52), relapse prevention (e.g. Chaney et al. 53) and approaches with family and concerned others (e.g. Copello et al. 10).

Utilising cognitive and behavioural strategies, SBNT helps clients to develop family and social networks that are supportive of change. A key strength of the approach is the primary focus on addressing drug and alcohol problems by engaging with a network of positive support for lifestyle change. SBNT has additional advantages to help sustain engagement with vulnerable young people, who may be disconnected from their families, by broadening the reach of the intervention beyond the traditional family to include supportive peers. Core strategies such as motivational techniques, improving communication and coping mechanisms are used, as well as (given the nature of substance misuse) developing a network-based relapse management plan. Using such a therapeutic approach also provides scope to address client-focused elective areas such as educational requirements. 54

Involving young people in research

Involving those who are the focus of research can have a positive impact on what is researched, how research is conducted and the impact of research findings (e.g. Brett et al. 55 and Staley56). In recent years there has been increasing interest in children and young people’s involvement in research,57–60 both as sources of data and through their active involvement in the planning and process of research. Although there is less of an evidence base in relation to children and young people’s involvement in research compared with the involvement of adults, the case for this involvement has been explored in a number of publications. 57,61,62 Research that actively involves children and young people should lead to research, and ultimately services, that better reflect their priorities and concerns63,64 and enhance the opportunity for optimal health outcomes. 65 However, the voices of children and young people who are less frequently heard, for example users of mental health services66,67 and looked-after children,59 are often absent from the literature on children and young people’s involvement in health and social care research.

As well as being located within the wider traditions of public involvement in research, young people’s involvement in this study was also developed with reference to children and young people’s involvement in the development and delivery of health and social care services. 64,68,69 Article 12 of the United Nations Convention on the Rights of the Child70 states that all children have the right to express their views in all matters that affect them, and this applies to children’s involvement in both decisions about their own experience as patients and service users and the development of health services and research. 68 In addition to drawing on previous work by the research team and other published studies, the Y-SBNT study sought from the outset to actively involve young people with experience of substance misuse services throughout the study. Young people’s involvement in the study was key to ensuring that the research addressed the concerns and issues faced by young people with substance misuse problems and we hope that this will also be an important addition to the wider evidence bases on public involvement and young people’s involvement in research.

Believing that SBNT could be successfully adapted to the youth context and would have great potential as a clinically effective and cost-effective intervention that could be readily and widely implemented in services for young people, the research team adapted the current SBNT approach to produce a purpose-designed therapy manual and resource kit. The feasibility of recruiting young people to receive this specifically developed family- and network-based intervention was then investigated in a feasibility study to establish the acceptability of the intervention to young people, their therapists and network members.

Intervention development phase

• To adapt an evidence-based family and social network intervention developed and tested with adult substance misusers to the youth context.

• To undertake a review of the evaluation and implementation literature to inform this adaptation.

• To involve young people, parents and therapists in the adaptation process to improve acceptability to these groups and ensure ready implementation in routine services.

• To develop a manual, resource kit and training programme for the delivery of the adapted intervention in a feasibility trial.

Feasibility randomised controlled trial phase

• To demonstrate the feasibility of recruiting young people to a family- and network-based intervention (Y-SBNT) across two service sites.

• To test the feasibility of training staff from existing young people’s addiction services to deliver the family and social network intervention.

• To evaluate the level of treatment retention among participants randomised to the family and social network intervention.

• To explore through qualitative interviews the participants’ views, acceptability and experiences of the intervention and the study process.

• To explore through qualitative interviews the views and experiences of those attending treatment sessions as members of the young person’s network and whether the intervention was acceptable.

• To establish treatment effectiveness through 3- and 12-month quantitative outcome data.

• To explore cost-effectiveness in preparation for a large definitive randomised controlled trial.

• To explore and develop models of public involvement that support the involvement of young people in a study of this nature.

Young people with experience of services

Involvement of young people was based on a series of young advisor and consultation meetings (see Chapter 8 and Appendix 2). During these meetings, the two public involvement leads and other members of the study team worked with young people to explore the principles behind the intervention (i.e. social networks, social support, engaging with services) as well as to obtain young people’s input/comments on the content and components of the intervention itself. The core principles and aspects of the Y-SBNT approach were retained, as the young advisors considered these to be relevant to their experience, but they also suggested adaptations to aspects of the approach and associated materials to make them more relevant to young people.

Activities undertaken with young people in relation to the content of the intervention included considering the nature of social networks and how these may differ between young people and adults using substance misuse services, such as thinking about the important people in young people’s lives and why they might be helpful/supportive or unhelpful/unsupportive (examples in Box 1).

Family/network members

The involvement of family members as part of this adaptation phase consisted of a focus group discussion with two adults (woman with foster daughter and mother with adopted son). The discussions highlighted the challenges faced by these two women that would influence both their engagement and that of the young people in something like Y-SBNT. The participants felt that, to some extent, social network support was already in existence for young people, although it was not always positive as some contacts promoted continued substance use behaviour. They identified communication between themselves and the young person to be ‘poor’ and that the intervention may help with this aspect. However, they felt that the work would depend on the extent to which the young person was ready to engage, for example ‘I just feel that mine would only engage as much as she wanted to engage . . . she would only share what she wanted to share’, and being able to be open and honest about their substance use was perceived to be central, for example ‘I don’t think she’s ready or in the frame of mind where she could open up to a different network’. The theme of ‘honesty’ from the young person was prominent in the discussions. Both participants felt that they would ‘definitely’ get involved in an intervention such as Y-SBNT if invited and would find the ‘opportunity to talk to the young person helpful and useful . . . if you are hearing it from a user’s point of view I can understand it a bit more’. One of the participants stated that involvement in sessions may help her as a mother to have more ‘tolerance and understanding’ as it was not something she had been brought up with or understood. Also, the idea of working together and supporting the young person was discussed; for example, ‘it would be the unison of it, both of us working towards a common goal which we’ve both got an understanding of’. Therefore, in summary, the two participants felt that the intervention was feasible and acceptable. There were some perceived potential benefits of the intervention including the potential for better and more open communication, increased understanding and supporting agreed goals. Potential challenges were identified in terms of the young person being ready to be ‘open and honest’ about their difficulties and being able to discuss them with potentially supportive people.

Therapists

Consultation was conducted with seven therapists as an early part of the training workshops. In addition, therapists were interviewed at the end of the study (see Chapter 6) when their experiences of intervention delivery were explored in more detail using one small group interview (two therapists) and individual interviews.

As part of the adaptation work, an early session with therapists involved providing them with a detailed description of the intervention including some of the additional materials developed following the work with young people described earlier in this chapter. This was followed by a discussion about the feasibility of delivering the intervention in their services as well as their views on its acceptability to young people. The components (core and elective sessions) of the intervention presented in Box 3 were discussed one at a time. These were perceived as appropriate by therapists and contrasted to routine practice. Some of the strategies, for example talking to teachers, hostel workers or other important people as part as the development of network support, were said by the therapists to be performed occasionally in routine work with young people. But the perception was that the structure of the Y-SBNT intervention and the coherent focus on positive help and support made it easier to deliver it in a more planned, consistent and coherent way.

The overall feedback stressed that more emphasis was needed on the different experiences and social contexts of young people but that there was no need to change the structure or content of the psychological components and strategies of the original SBNT intervention. An additional map/framework of the intervention journey (illustrated in Figure 1) was produced and included with the revised materials.

FIGURE 1.

The treatment map.

Inclusion criteria

Young people were considered potentially eligible if they met all of the following criteria:

• were aged 12–18 years inclusive

• had drug and/or alcohol problems and were newly referred and accepted for treatment by one of the two agencies during the period of recruitment

• were willing and able to provide written informed consent.

Exclusion criteria

Young people were excluded if they met any of the following criteria:

• had a concurrent severe mental illness that precluded them from active participation

• suffered from severe mental or physical illness likely to preclude active participation in treatment or follow-up

• were unable or unwilling to provide written informed consent.

Demographics

Details including age, sex, sexual identity (if aged ≥ 16 years), ethnicity, nationality, religion, education, employment and current living arrangements were collected.

Timeline Follow-Back interview

The Timeline Follow-Back (TLFB) interview77 records substance use, particularly the proportion of days on which the main problem substance was used in the preceding 90-day period. This interview is based on a retrospective calendar review of each day’s consumption and has previously been validated and widely used with adolescent populations. 78–80 It has been shown to demonstrate high comparability with the Global Appraisal of Individual Needs (GAIN) (r = 0.7–0.8) in a population of adolescents aged 12–18 years admitted to a residential treatment programme for substance abuse or dependence. 78 Applying the TLFB interview allowed the collection of detailed data on the full range of licit and illicit drugs used by participants. Participants were asked to recall on which days they had used their primary problem substance in the last 90 days. The number of days was summed and divided by 90 to obtain the proportion (from 0 to 1).

Strengths and Difficulties Questionnaire

The self-report Strengths and Difficulties Questionnaire (SDQ) is a brief behavioural screening questionnaire designed to be completed by adolescents. 81,82 The 25-item questionnaire has five separate subscales for different aspects of problems or behaviours: emotional problems, conduct/behaviour problems, inattention/hyperactivity, relationships with peers and prosocial behaviour. Scores for the first four scales can be added together to produce a score for total difficulties ranging from 0 to 40, with a lower score indicating greater emotional well-being and a score of ≥ 20 indicating an ‘abnormal’ score. The SDQ has been used extensively and has demonstrated high levels of reliability and validity. 83,84 In a factor analysis with 3983 11- to 15-year-olds, all 25 items loaded on the predicted factors relating to the five subscales (> |0.33|). 84 A few items also loaded on additional factors > 0.30, but the loadings on the predicted factors were higher than the loadings on the additional factors for all but one item. The total SDQ score demonstrated good reliability with a Cronbach’s alpha coefficient of 0.80 and test–retest reliability of no less than 0.62.

Important People Drug and Alcohol interview

Given the emphasis on family and peer support of the intervention, social network support was measured using the Important People Drug and Alcohol (IPDA) interview85 to understand the influence of social support on treatment for substance misuse, assess the degree to which social network changes had been achieved and assess the extent to which these changes still needed to be made. This interview requires the participant to provide information on up to 12 significant people who have featured in their lives in the previous 3 months. Data collected on these people include their relationship with the young person, frequency of contact, importance, supportiveness and use of drugs and alcohol. This instrument has a three-factor structure of substance involvement (α = 0.92), general/treatment support (α = 0.84) and support for substance use (α = 0.85),85 established in an adult population.

Family Environment Scale: Family Relationships Index

The 27-item relationship dimension of the Family Environment Scale (FES)86 consists of cohesion, expressiveness and conflict subscales (nine items each), for which a raw score from 0 to 9 can be calculated and converted to a standard score. 87 The standard scores range from 4 to 65 for the cohesion subscale, from 16 to 71 for the expressiveness subscale and from 33 to 80 for the conflict subscale; a higher standard score indicates greater cohesion and expressiveness and lower levels of conflict among family members, respectively.

The Family Relationships Index (FRI) is designed to look at the atmosphere in the family household with subscales that measure support, expression of opinions and angry conflict within a family. This was completed by the young person (other family members were not asked to complete this instrument). This measure has been used previously in similar populations88,89 and has undergone extensive development and validation work. It has shown acceptable levels of internal consistency in a validation study with 1298 pupils aged 11–18 years from 16 schools in Victoria, Australia (Cronbach’s alpha of 0.39 for the expressiveness scale, 0.67 for the cohesion scale and 0.72 for the conflict scale),90 test–retest reliability of each subscale of at least 0.73 and good content and construct validity. 91

Health-related quality of life

The EuroQol-5 Dimensions five-level version (EQ-5D-5L)92 is a standardised measure of health status developed by the EuroQol Group to provide a simple, generic measure of health for clinical and economic appraisal, with health characterised on five dimensions (mobility, self-care, ability to undertake usual activities, pain and anxiety/depression). Participants were asked to describe their level of health on each dimension using one of five levels: no problems, slight problems, moderate problems, severe problems or extreme problems. 93 These dimensions were accompanied by a visual analogue scale (VAS), with participants asked to assign a number between 0 and 100 to measure their perceived health status.

Health and social care resource use

The feasibility of collecting details regarding hospital and primary health-care services use, contact with the police and criminal justice system and social care services use was tested in this participant group. Participants indicated how many times in the previous 12 months they had seen a GP or practice nurse or received hospital care. They were asked about any accidents that they may have had in the past 12 months and attendance and absenteeism from school or employment was recorded. In addition, contact with the police, criminal justice system and social services was documented.

This collection of self-reported resource use data was continued until participants had been in the study for 12 months or until they withdrew from follow-up or fully withdrew from the study.

Training in the delivery of youth social behaviour and network therapy

Training was delivered by two professionals during two workshops in each site – West Midlands and the North East. Volunteers from the therapists’ pool in each of the two services were trained to deliver the Y-SBNT intervention (five in the West Midlands and two in the North East). Once allocated, therapists delivered Y-SBNT to participants only for the duration of the trial, with the non-trained therapists in each service delivering TAU within the other arm; hence, there was no overlap between therapists across the two trial arms.

The training took place from 0930 to 1500 on each day. After the first day, trainees conducted practice cases in their respective services before coming back to training approximately a month later for the second day when this work was reviewed.

As soon as recruitment commenced and throughout the trial period when treatment was delivered, the Y-SBNT therapists in each site received group supervision twice a month for 1 hour each time. In the West Midlands, supervision was face-to-face whereas in the North East it was delivered over the telephone. Most of the supervisions were delivered by one of the group trainers, with a couple of supervisions delivered jointly by both trainers. The group supervision sessions focused on discussion of each individual case that therapists were working with.

At one meeting during the trial all Y-SBNT therapists came together in the West Midlands for an afternoon supervision session. This was delivered by both trainers.

Supervision finished once the last client was seen within the Y-SBNT arm.

An early additional session was provided by the researchers covering protocol issues and research processes including the rationale for the study, patient eligibility, the randomisation process, study paperwork completion, digital recorder use and handling of participant withdrawal.

Intervention content

Participant follow-up

Appendix 7 shows a summary of the Y-SBNT trial and the timing of follow-ups.

Trial completion

Participants were deemed to have completed the trial when their final follow-up had been completed.

Participants were deemed to have fully withdrawn from the trial when:

• they wished to exit the trial fully

• their therapist withdrew them fully from the trial.

Instead of withdrawing fully from the trial, participants had the option of:

• withdrawing only from receiving trial treatment but continuing to complete follow-up data collection

• withdrawing only from follow-up data collection but continuing to receive trial treatment.

Participants electing to withdraw from both the trial treatment and the follow-up data collection were deemed to be full withdrawals.

Primary analysis

As this was a feasibility study, outcomes were primarily summarised only descriptively. The end point of main clinical interest, the TLFB interview, was analysed to obtain the effect size and CI. Analysis was conducted in Stata 13 (StataCorp LP, College Station, TX, USA).

Intervention

Intervention costs encompassed staff costs of the therapists delivering the intervention and the trainers who delivered the training and supervision to these therapists, travel costs and overheads. The therapists’ remuneration was recorded by the research team and a 12% overhead was added based on the information provided by the service providers. Hourly earnings were then multiplied by the time that participants spent in contact with the therapist, which included both appointment length and the time that the therapist spent travelling to the appointment, as recorded in the treatment log. Travel costs were estimated using the NHS reimbursement rate and we did not distinguish between vehicles (e.g. car, bus, motorcycle). 106

Training costs included staff costs for training sessions and supervision sessions and travel costs when applicable. For each site in the trial, 2 days of training was delivered by two professionals, whose hourly wage including overheads was obtained from their finance departments. Staff costs of the therapists who attended the training were also taken into account. Supervision costs included twice-monthly 1-hour group meetings for each site.

Service use

Participants’ use of health care was identified retrospectively by self-report using a service use questionnaire at baseline and 12 months’ follow-up. It covered primary care, secondary care and social services. A set of national average unit costs (see Table 23) was then applied to the usage recorded to estimate the total cost of health-care and social services during the previous 12-month period.

Similarly, contacts with the police and the criminal justice system were also recorded by self-report by completion of a service use questionnaire at baseline and 12 months’ follow-up. In this case it was limited to arrests, court appearances, probation and prison days because of the complexity of the procedures involved. A set of national average unit costs (see Table 23) was applied when applicable, as with health-care and social services use.

Health-related quality of life

During the pilot trial, HRQoL measured by the EQ-5D-5L was collected at baseline and 3 and 12 months’ follow-up. Although by conventional approach we would estimate QALYs by calculating the area under the curve, for this feasibility study we did not intend to do so. 107 As this was a pilot trial with a small number of participants, calculated QALYs could appear conclusive and misleading. Instead, the mean index scores and the VAS scores were investigated.

Future involvement

We hope to be able to support young people’s involvement in the development of a lay summary and other publications related to public involvement in the study, as well as in dissemination and planning of a possible follow-on project to establish a stand-alone young people’s group or panel.

Involving young people in the Trial Steering Committee

In early 2014 we also reviewed our initial plans to have young people as lay members at TSC meetings. The emerging model of young people’s involvement did not include the planned core group of young people engaged with the project regularly on an ongoing basis, who would not only have an understanding of the study itself but also have insights into the process and management of randomised controlled trials. It was felt by the Trial Management Group (TMG) that this would have made it more difficult to make any involvement in the meetings accessible and meaningful for lay members. The public involvement leads therefore agreed to report to TSC meetings and offer young people involved in the study the chance to contribute to this report or to attend the meetings if they wanted to do so, in which case we would seek to make the meetings as accessible as possible. However, the one young person who did express an interest in coming to a TMG and TSC meeting was unable to do so because of ill health.

Involving parents and carers

We had initially proposed that a small number of parents and carers, unrelated to the young people involved in the study either through involvement activity or as participants, would also be involved in the project through representation on the study advisory group. In July 2014 we ran a focus group with two adults (a woman with a foster daughter and a mother with an adopted son). Both focus group participants said that they would be keen to be kept updated on the project and were invited to participate further, but neither made any commitment to future involvement. It is also true that parental involvement, as a small part of the overall public involvement element of the study, perhaps got rather lost amid all of the additional work involved in the recruitment and retention of young advisors.

Trial recruitment

Young people’s services in two UK regions participated in the study: the West Midlands and the North East. Our initial aim was to recruit 60 participants between 1 May 2014 and 31 October 2014 at a rate of 10 participants a month. In practice, recruitment did not open until 19 May 2014, with the first participant recruited on 30 May and the last on 14 November 2014 (Figure 2). In total, 53 participants were randomised, 26 (49.1%) to the Y-SBNT arm and 27 (50.9%) to the TAU arm. This equates to an approximate average recruitment rate of nine young people every 4 weeks of active recruitment (Figure 3). The number of participants recruited in the West Midlands was 33 (62.3%) and in the North East was 20 (37.7%).

FIGURE 2.

Cumulative recruitment against target recruitment by month in the Y-SBNT trial.

FIGURE 3.

Monthly recruitment of young people into the Y-SBNT trial.

In Figure 4, the CONSORT flow chart shows the progress of participants through the trial.

FIGURE 4.

The CONSORT flow diagram for the Y-SBNT trial.

The majority of participants were male (n = 40, 75.5%) and the average age was 16.1 years (range 12–18 years) (Table 5). Participants were mostly of white British ethnicity (n = 40, 75.5%) but there were also participants with Caribbean, African and Asian ethnicity. Over half of the participants were still in education (n = 32, 60.4%) and most lived with their parent(s) or carer(s) (n = 32, 60.4%). A number of young people reported religious beliefs, with a higher proportion of young people with religious beliefs in the TAU group (29.6% vs. 7.7%).

For those randomised into the trial from the North East, 12 (60.0%) were referred from the community, seven (35.0%) were referred from YOTs and one (5%) self-referred from the YOT. In the West Midlands, a lower proportion was referred from the community (n = 9, 27.3%) and a higher proportion was referred from YOTs (n = 24, 72.7%). Overall, the referral routes were reasonably well balanced between the two treatment groups. In the Y-SBNT group, nine (34.6%) were referred from the community and16 (61.5%) from YOTs, with one (3.8%) self-referring whereas in the TAU group 12 (44.4%) were referred from the community and 15 (55.6%) were referred from YOTs.

Participants were followed up at 3 and 12 months post randomisation. Each follow-up consisted of a face-to-face meeting with a member of the study research team. Significant efforts were made to ensure high follow-up rates. Attempts were made to follow up all participants at each time point, except at month 12 for one participant in the TAU group who formally withdrew from follow-up after the intervention because of deteriorating mental health. In total, 41 (77.4%) participants were followed up at month 3 (Y-SBNT, n = 23; TAU, n = 18) and 39 (73.6%) were followed up at month 12 (Y-SBNT, n = 20; TAU, n = 19) (Table 6). Because of delays in the study set-up, patients recruited in the last 2 months of recruitment were followed up at a time point earlier than 12 months to enable the study to be completed within the study timeline (median length of follow-up 12.5 months, range 10.0–15.3 months). Participant follow-up ended on 6 October 2015, at which point the study ended.

Treatment sessions

Treatment sessions took place between 12 June 2014 and 26 March 2015 for the Y-SBNT group and between 4 July 2014 and 26 August 2015 for the TAU group; in total, 103 Y-SBNT sessions and 62 TAU sessions were attended (Table 7). It was intended that the Y-SBNT intervention would consist of six 50-minute SBNT sessions delivered over a maximum of 12 weeks. In total, 22 Y-SBNT participants attended at least one session and 11 participants attended all six sessions (median number of sessions attended 5, range 0–6). For the 11 participants who attended all six sessions, the time between the first and sixth session ranged from 8 to 20 weeks (median 12 weeks). The median length of the Y-SBNT sessions attended was 60 minutes (range 15–90 minutes). In the TAU group, 20 participants attended at least one session during their 12-month follow-up (median number of sessions attended 2, range 0–15).

Overall, among participants referred from the community, the median number of sessions attended was four (range 0–6), whereas among participants referred from YOTs (including self-referral), the median number of sessions attended was two (range 0–15).

Intervention effectiveness: Timeline Follow-Back interview

Data for the proportion of days that the primary substance was used in the preceding 90 days were missing for 12 participants at month 3 and for 14 participants at month 12 (Table 8).

The primary problem substances for participants were alcohol (Y-SBNT, n = 2; TAU, n = 3), cannabis (Y-SBNT, n = 19; TAU, n = 20) and other substances including legal highs (Y-SBNT, n = 5; TAU, n = 4). Other substances included NPSs [n = 7; specifically mentioned were synthetic cannabinoid (n = 1) and mephedrone (n = 2)] and solvents (n = 2). The mean number of days that participants reported using their primary substance in the preceding 90 days was higher in the Y-SBNT group than in the TAU group at baseline, month 3 and month 12 (Table 9).

The main clinical outcome is summarised by treatment group and time point in Table 10. The Y-SBNT group had a higher mean proportion at all time points.

Data for the outcome at baseline were complete but nine participants were missing data at both month 3 and month 12 and so the model was based on 44/53 (83.0%) participants [Y-SBNT, n = 24 (92.3%); TAU, n = 20 (74.1%)].

The baseline characteristics of the participants included in the TLFB analysis model are compared between the treatment groups in Table 11. It does not appear that the loss of the nine participants with missing outcome data significantly impacted on the balance achieved at randomisation between the treatment groups.

An exchangeable (compound symmetric) covariance pattern was used for the analysis model. The assumptions of the model were checked visually. The normality of the standardised residuals was assessed using a Q–Q plot and the homoscedasticity of the errors was checked by plotting the residuals against the fitted values (see Appendix 11). The Q–Q plot is used as a graphical method of comparing the distribution of the residuals to a normal distribution. The quantiles of the distributions are plotted against each other. If the residuals are normally distributed, the points will lie on the straight line, which represents the normal distribution. Deviations from the straight line indicate skewness in the distribution of the residuals. These plots gave no reason to be concerned about the validity of the assumptions.

Treatment effect at 12 months

There was no evidence of a difference between the Y-SBNT group and the TAU group at month 12 in the proportion of days that the main problem substance was used in the preceding 90-day period (AMD 0.13, 95% CI –0.12 to 0.39; p = 0.30; see Table 10), that is, the Y-SBNT group had a slightly higher average proportion at 12 months but this difference was not statistically significant (Figure 5). The effect size was calculated as –0.32 (upper limit for one-sided 80% CI –0.05).

FIGURE 5.

Adjusted means for the proportion of days that the primary problem substance was used in the preceding 90-day period by treatment group over time.

Treatment effect at 3 months

There was no evidence of a difference between the Y-SBNT and TAU groups at month 3 in the proportion of days that the main problem substance was used in the preceding 90-day period (AMD 0.06, 95% CI –0.19 to 0.32; p = 0.63; see Table 10), that is, the Y-SBNT group had a slightly higher average proportion at month 3 but this difference was not statistically significant. The effect size was calculated as –0.15 (upper limit for one-sided 80% CI 0.11).

The proportion of days that the primary substance was used in the preceding 90 days at month 12 is summarised by primary therapist in Table 12.

Each therapist was associated with between one and six participants (median 3). The mean proportion of days that the primary problem substance was used in the preceding 90 days at month 12 varied across the therapists from 0.01 to 0.86.

Strengths and Difficulties Questionnaire

The five subscales and the total score for the SDQ are summarised descriptively by intervention group and time point in Table 13.

At month 3, the Y-SBNT group had a slightly lower average total SDQ score than the TAU group, with a mean difference of –0.7 points (95% CI –3.9 to 2.5), indicating that the Y-SBNT group experienced, on average, greater emotional well-being at month 3 than the TAU group. The associated treatment effect was calculated as 0.13 with an 80% one-sided confidence limit of 0.39. However, the direction of the effect reversed at month 12 to –0.15 in favour of the TAU group (Figure 6).

FIGURE 6.

Mean scores by treatment group for the subscales and the total score of the SDQ: (a) emotional symptoms scale; (b) conduct problems scale; (c) hyperactivity scale; (d) peer problems scale; (e) prosocial scale; and (f) total SDQ score. BL, baseline; M3, month 3; M12, month 12.

Among the young people who provided data at month 12, 25% in the Y-SBNT group (n = 5) and 15.8% (n = 3) in the TAU group had a total SDQ score that fell into the abnormal banding, which can be used to identify likely cases with mental health disorders (see Appendix 12).

The Y-SBNT group also had a higher mean impact score than the TAU group at month 12 (mean difference 0.5, 95% CI –0.8 to 1.9) and a larger proportion of participants fell into the abnormal banding for the impact score subscale (35.0% vs. 21.1%) (Table 14). In a similar trend to that seen for the total SDQ score, the direction of the effect reversed from being in favour of the Y-SBNT group at month 3 to being in favour of the TAU group at month 12.

A greater proportion of participants in the TAU group (78.9%) than in the Y-SBNT group (65.0%) said at month 12 that their problems were a bit or much better since coming to the clinic. Overall, the TAU group appeared to perform better than the Y-SBNT group in the constructs measured by the SDQ at 12 months post randomisation; those in the TAU group tended to experience slightly greater emotional well-being and the problems that they experienced had a lesser impact on their lives.

Important People Drug and Alcohol interview

At month 12, participants in the Y-SBNT group discussed a median of nine people (range 4–12, total 174) and those in the TAU group discussed a median of seven people (range 3–12, total 132) in the IDPA (Table 15). There were at most seven missing responses to a particular item across all participant responses at any one time point, indicating that missing data are not a significant problem when using the IPDA interview with this population.

The proportion of the network with whom participants have daily contact, who are family members and who use cannabis, alcohol or other substances is also summarised in Table 15 by treatment group and time point. On average at month 12, participants in the TAU group had daily contact with a greater proportion of their support network than those in the Y-SBNT group and family members tended to make up a greater proportion of their network; however, there was a slightly higher prevalence of cannabis and alcohol use among network members of participants in the TAU group, although the differences were marginal. z-Scores for the three components (network substance involvement, general/treatment support and network support for substance use) were calculated (Table 16 and Figure 7).

FIGURE 7.

Mean scores by treatment group for the component z-scores of the IPDA interview: (a) component 1 version 1: network substance involvement (primary substance); (b) component 1 version 2: network substance involvement (all drugs and alcohol); (c) component 2: general/treatment support; and (d) component 3: network support for substance use. BL, baseline; M3, month 3; M12, month 12.

A lower standard score is favourable for the two versions of component 1 and component 3 and indicates lower than average levels of primary substance (version 1) and all drugs and alcohol (version 2) involvement within the support networks and less encouragement of substance misuse, respectively. At month 12, participants in the Y-SNBT group had, on average, scores for both versions of component 1 and for component 3 that were higher than the population average. A higher score for component 2 indicates greater support for treatment and thus is favourable; members of the Y-SBNT group reported, on average, higher levels of support than members of the TAU group.

Family Environment Scale: Family Relationships Index

Each participant provided some data for this instrument at baseline; individual item data were complete for all participants except one (TAU group), who did not provide a response to two items (items 13 and 25). When participants were followed up at months 3 and 12, a response to each item of the FRI was received. There were several cases in which the participant either ticked both boxes or ticked in between the two boxes (baseline: seven items among four participants; month 3: nine items among four participants; month 12: two items among two participants). Standard scores for the cohesion, expressiveness and conflict subscales are summarised by treatment group and time point in Table 17. The Y-SBNT group had lower average scores for all three subscales at month 3 and month 12, indicating that members of the Y-SBNT group tended to experience higher levels of conflict and lower levels of cohesion and expressiveness in their family environment than the TAU group throughout their participation in the trial (Figure 8). The difference for the expressiveness scale was statistically significant. Floor and ceiling effects were explored using bar charts of individual items (see Appendix 13).

FIGURE 8.

Mean scores by treatment group for the subscales of the FRI: (a) conflict subscale; (b) cohesion subscale; and (c) expressiveness subscale. BL, baseline; M3, month 3; M12, month 12.

Working Alliance Inventory

Participants and therapists were asked to complete a WAI at the end of sessions 1 and 3. The number received as a percentage of the number of participants and therapists attending these sessions is presented by treatment group, session number and WAI type in Tables 18 and 19, respectively. Response rates in the Y-SBNT group were significantly higher than in the TAU group. The Y-SBNT group had a higher mean score than the TAU group for all three WAI subscales of both the patient and therapist instruments for each session, although the magnitude of the differences was notably larger for the therapist WAI (see Tables 18 and 19). This indicates that, on average, participants and therapists in the Y-SBNT group had a greater perceived strength of working alliance than those in the TAU group. Graphs of mean scores by session and treatment group are provided in Figure 9.

FIGURE 9.

Mean patient and therapist WAI subscale scores by treatment group and session number: (a) patient tasks scale; (b) patient bond scale; (c) patient goal scale; (d) therapist tasks scale; (e) therapist bond scale; and (f) therapist goal scale.

School attendance and engagement

At month 12, 10 (38.5%) participants in the Y-SBNT group and eight (29.6%) in the TAU group reported being in education in the previous 12 months. Among these, participants were absent from school within this time for a median of 4.5 days (range 0–60 days) in the Y-SBNT group and 20.5 days (range 1–140 days) in the TAU group. A median of 0 absent days (range 0–4 days) in the Y-SBNT group and 0 absent days (range 0–50 days) in the TAU group were the result of drug or alcohol misuse. In total, 11 participants (Y-SBNT, n = 4; TAU, n = 7) believed that their performance at school or college had been affected as a result of drug or alcohol misuse over the last 12 months: four slightly (Y-SBNT, n = 1; TAU, n = 3), six moderately (Y-SBNT, n = 3; TAU, n = 3) and one extremely (TAU).

Adverse events

No adverse events were reported.

Missing data

Participants’ responses were considered missing if left blank or if answers were deemed unrealistic. Unrealistic answers included the number of days spent in a secure children’s home/young offender institution exceeding 365 days (in the last 12 months). When the option of ‘not answered’ was possible and selected, questions were considered answered and responses were therefore not considered missing.

Overall, the level of missing data appeared to be low (Table 20). There appeared to be a general missing data pattern, with missing data not more prevalent in the TAU group than in the Y-SBNT group or vice versa. Three questions showed a relatively higher level of missingness: (1) the number of times answered bail, (2) the number of contacts with a probation officer and (3) the total number of appearances at Crown Court. The ‘not answered’ responses were treated as missing in the complete case analysis. Overall, 17 questions had the option of ‘not answered’, of which seven received ‘not answered’ responses. The proportion of participants choosing ‘not answered’ ranged from 1.9% to 43.6%. The question on travelling in an ambulance had the highest percentage of ‘not answered’ responses (41.5% at baseline and 43.6% at 12 months). A slightly higher percentage of ‘not answered’ answers were seen at 12 months than at baseline.

Costs

Health-care and social services costs

Table 22 shows the health service use at each time point by group. At baseline there appeared to be no great differences between the two groups in the numbers of participants reporting usage. Similarly, at 12 months, there was little indication of differences between the two groups in the numbers of participants reporting health service usage. In terms of differences between baseline and 12 months, the absolute numbers tended to show a decline in service use for both groups. However, because of loss to follow-up, the proportions of participants who used services varied. In the Y-SBNT group there was an increase of 18.2% in the proportion of participants receiving prescriptions and in the TAU group there was an increase of 6.6% between baseline and 12 months. The number of nights stayed in hospital as an inpatient in the Y-SBNT group remained the same (5 nights) between baseline and 12 months, but this represented a proportional increase of 30%. On the other hand, there was a decrease in accident and emergency (A&E) visits for both groups. This might indicate no specific trend for either group when loss to follow-up is considered.

TABLE 22 - Comparison of health service use between groupsa

A&E, accident and emergency.

Total number of participants who used the service (the total number of times used is reported in parentheses unless stated otherwise).

Number of times stayed in hospital as inpatient asked only at 12 months, not at baseline.

Total number of participants visited, with the total number of times visited in parentheses.

Health service use	Y-SBNT		TAU
	Baseline (n = 26)	12 months (n = 20)	Baseline (n = 27)	12 months (n = 19)
GP visits at surgery	17 (52)	12 (75)	16 (78)	12 (78)
Practice nurse visits at surgery	8 (14)	3 (9)	6 (21)	3 (11)
GP visits at home	0 (0)	0 (0)	1 (1)	0 (0)
Practice nurse visits at home	0 (0)	0 (0)	0 (0)	0 (0)
Contacted NHS Direct (‘yes’)	3	0	3	2
Visited NHS walk-in centre	9 (20)	3 (4)	9 (21)	8 (10)
Received any prescriptions (‘yes’)	11	10	16	12
Visited hospital A&E department	15 (44)	10 (46)	14 (33)	9 (15)
Visited hospital A&E department (number of nights)	4 (8)	8 (43)	0 (0)	3 (6)
Travelled in a 999 ambulance	9 (17)	7 (37)	7 (16)	4 (5)
Stayed in hospital as an inpatient (number of nights)	5 (21)	5 (111)	5 (10)	4 (10)
Stayed in hospital as an inpatient (number of times)b	–	5 (30)	–	4 (4)
Visited hospital as outpatient	8 (18)	4 (26)	9 (18)	3 (9)
Motor vehicle accident	2 (2)	1 (1)	2 (2)	3 (6)
Motor vehicle accident and attended hospital	1 (1)	0 (0)	1(1)	1(1)
Personal injury	9 (34)	10 (12)	9 (23)	10 (19)
Personal injury and attended hospital	7 (11)	8 (9)	6 (17)	6 (7)
Injury to another person	7 (31)	5 (20)	8 (26)	7 (24)
Injury to another person and attended hospital (number of times attended hospital)	3 (5)	2 (2)	2 (6)	3 (6)
Involvement with social services (‘yes’)	13	5	15	10
Number of times visited by social worker at homec	8 (133)	3 (169)	9 (149)	6 (45)
Number of times visited social worker at their officec	6 (26)	3 (13)	1 (6)	1 (2)

For the health service use questions some outliers were found. At 12 months, one participant in the TAU group reported 30 GP visits at surgery and one participant in the Y-SBNT group reported 48 visits. The same participant in the Y-SBNT group was an outlier for three different questions, all at 12 months (30 visits to A&E, 30 nights in A&E and 30 times used an emergency ambulance). Two participants in the Y-SBNT group also reported considerably longer hospital stays than other participants (40 and 45 nights).

A set of national average unit costs was extracted from several sources (Table 23). After excluding those with missing cost or EQ-5D-5L data, the set of unit costs was applied to the usage data presented in Table 22. The resultant health-care and social services costs are presented in Table 24 for the complete case and available case analysis by time point. The complete case analysis shows that both groups appear to have similar level of health-care and social services costs at baseline. However, at 12 months the Y-SBNT group had higher costs, which is partly attributable to one participant with a health-care and social services cost of > £20,000. If this outlier is excluded, the cost per person reduces to £1388. The available case results are slightly lower than the complete case results; however, the differences between the two groups are similar in both analyses.

TABLE 23 - Unit costs of health-care and social services use

Service use	Unit cost (£)	Source
GP consultation (in surgery)	38	Curtis106
GP consultation (home visit)	62	Curtis106
Practice nurse consultation (in surgery)	11	Curtis106
Practice nurse consultation (home visit)	18	Curtis106
NHS walk-in clinic	56	Curtis106
Inpatient admission	562 (per night)	Department of Health113
Emergency medicine	124	Department of Health113
Ambulance conveyance	231	Department of Health113
Outpatient attendance	114	Department of Health113
Prescription cost per year	163 (average per head)	Health and Social Care Information Centre114
Social worker (home visit)	66	Curtis106
Social worker (at office)	55	Curtis106

TABLE 24 - Health-care and social services costs (not including intervention costs) at baseline and 12 months

Cost (£)	Y-SBNT				TAU
	Complete case		Available cases		Complete case		Available cases
	Baseline (n = 16)	12 months (n = 16)	Baseline (n = 24)	12 months (n = 19)	Baseline (n = 11)	12 months (n = 11)	Baseline (n = 21)	12 months (n = 13)
Mean (SD) health-care and social services cost per participant	1933 (2916)	2150 (5517)	1719 (2460)	1899 (5073)	1564 (2166)	724 (405)	1402 (1756)	683 (384)
Total group health-care and social services cost	30,930	34,403	41,244	36,081	17,205	7964	29,427	8876

Figure 10 demonstrates the change in cost per person from baseline to 12 months. Whereas costs in the TAU group fall, in the Y-SBNT group there is an increase in costs.

FIGURE 10.

Comparison of health-care and social services costs per person between groups at baseline and 12 months (complete case).

In this section, an open question in the questionnaire regarding social services use should be noted because of its ambiguous phrasing. First, this question did not specify the recall period for involvement with social services as the last 12 months. As a result, some participants recalled contact with social services over their entire life at both baseline and 12 months, which could have resulted in double counting at the 12-month follow-up. Second, the term ‘involvement’ posed confusion with regard to whether this question referred to the participant personally or included other family members as well, with a few answers referring to siblings/parents/children who had received social care. Because of this the costs reported above did not include this question, but did include two other questions regarding use of social services. A summary of the responses for this question can be found in Appendix 14.

Health-related quality of life

EuroQol-5 Dimensions data were collected with all other components in the questionnaire. Apart from those lost to follow-up there were no missing data in this section.

Figures 11 and 12 show the results for the complete cases, defined as those participants who answered the EQ-5D-5L at baseline, 3 months and 12 months. Figure 11 shows the proportion of severity category in each domain for Y-SBNT. At baseline, the anxiety/depression domain was the only domain in which extreme problems were reported, whereas the self-care domain was the domain with the highest percentage of participants reporting no problems.

FIGURE 11.

EuroQol-5 Dimensions complete case results by domain in the Y-SBNT group at (a) baseline; (b) 3 months; and (c) 12 months.

FIGURE 12.

EuroQol-5 Dimensions complete case results by domain in the TAU group at (a) baseline; (b) 3 months; and (c) 12 months.

The results at 3 months reveal an increase in those reporting severe problems within the anxiety/depression, pain/discomfort and usual activities domains. In the anxiety/depression domain there was a fall in the percentage reporting no problems, but now no participants experienced extreme problems in any domain. Overall, Figure 11 demonstrates how there was an increase in the percentage of participants in the Y-SBNT group reporting no problems in the anxiety and depression domain from baseline to 12 months. There was also a general decrease in the percentage reporting slight problems and moderate problems for this domain, but reports of extreme problems were again seen at 12 months.

Figure 12 presents the EQ-5D-5L results at all three time points for the TAU group. A lower percentage of participants in the TAU group than in the Y-SBNT group reported no problems in the mobility, pain/discomfort and anxiety/depressions domains at baseline. Moreover, a larger percentage reported slight problems across the domains, which was particularly prevalent in the anxiety/depression domain. There were no reports of severe or extreme problems across any domain or at any time point. There was an increase in the percentage reporting no problems in the anxiety/depression domain from baseline to 12 months in the TAU group. Overall, the TAU group saw a greater improvement across all domains than the Y-SBNT group.

Figure 13 reports the mean index score for each group at each time point. Although both groups show signs of improvement within the complete case analysis, this improvement is greater for the TAU group because of the group’s lower baseline score but higher 12-month score. The mean imputation results show the TAU score to be much lower than the Y-SBNT score across all three time points. This suggests that, for the TAU group, those with missing data and those lost-to-follow-up are the participants with the worst index scores. This difference is not nearly as pronounced for Y-SBNT, with the mean imputation analysis giving very similar results to the complete case analysis. Adjusting for baseline had little effect on the results for Y-SBNT; however, the overall improvement for TAU was reduced.

FIGURE 13.

EuroQol-5 Dimensions mean index score at each time point. a, Mean imputation; b, complete case analysis adjusted for baseline.

Table 29 presents the mean VAS score at each time point for each group. Three participants were missing at 3 months but were then seen again at 12 months (TAU, n = 2; Y-SBNT, n = 1). The complete case analysis for the VAS includes those participants who had no missing VAS data, without taking into account the missing data for costs and the EQ-5D-5L index, as both are unrelated to the VAS results. The complete case results reveal that the mean VAS scores for the Y-SBNT group are greater at all three time points than the equivalent VAS scores for the TAU group. On the other hand, the available case results show that the TAU group has a greater mean score than the Y-SBNT group at 12 months. This result is more in line with the complete case results for the EQ-5D-5L index score analysis.

Figure 14 shows that the median VAS score for the TAU group increases from baseline to 3 months, but then remains unchanged at 12 months. In contrast, the median VAS score for the Y-SBNT group increases from baseline to 3 months and then falls back to the baseline score at 12 months. Again, this contrasts with the EQ-5D-5L index score results, which showed that those in the Y-SBNT group experience worse scores at 3 months. However, different definitions of ‘complete case’ have been used, which may partly explain this contradiction. The size of the box plot at 12 months shows greater dispersion for Y-SBNT than for TAU. However, more outliers are seen in the TAU group than in the Y-SBNT group across all three time points. The skewness of the data is mostly symmetrical across all box plots except for the TAU group at 12 months, for which the median is lower.

FIGURE 14.

Comparison of VAS scores between groups and across time points (complete case).

Young people’s lives before treatment

An important theme identified was that of seeing young people in the context of their whole lives and not just focusing on their referral to a drug and alcohol support service. There was a mix of young people who lived with family members or relatives and young people who were in social care or living in a hostel or with foster carers. A few young people aged > 16 years had independent living arrangements with social services involvement. For those involved in the care system this could bring uncertainty to their lives:

Not being able to settle and that, not knowing what’s going to happen and not knowing where I’m going, puts more stress on us.

Young person 1032, female, 14 years, mephedrone

The majority of the young people attended school (both mainstream and pupil referral units)/college or training courses. For those not in training or education this had implications for their well-being. If in care and also not attending school/college or training, this could have a serious impact on how they spent their time:

Because the thing is, when you’re in the [hostel], I was basically independent, so I could go out when I wanted and I could sleep when I wanted.

What were you doing, if you don’t mind me asking?

Begging.

Young person 1031, male, 17 years, NPS

Therefore, any potential intervention should be viewed in the light of the challenging nature of some young people’s lives, in that their substance misuse may be symptomatic of other more challenging circumstances that they are experiencing.

Entry to treatment

For the vast majority of young people it was their first experience of drug and alcohol treatment. Referral sources included school/college, social workers and the youth offending system. Only a few young people perceived their drug or alcohol use as a problem for them:

Yeah, I went into the YOT when I went to collect my money and I had a word with someone there and I said that I think I’ve got a problem.

Young person 1020, female, 17 years, mephedrone

In the main, the young people did not view their substance use as a problem themselves before attending treatment services:

It was strong use, it was nothing major.

Young person 1000, male, 17 years, cannabis

Why did you think it wasn’t a problem for you?

Because it was making me happy.

Young person 1032, female, 14 years, mephedrone

They also said that their friends did not perceive their drug and alcohol use as an issue, as many of them were also using substances:

How about your friends, did they see it as a problem?

No. They see it as fun.

Is it because they’d been taking it with you?

Yeah.

Young person 1001, male, 16 years, cannabis

Nevertheless, they did recognise that their behaviour was causing concern to those close to them, particularly close family members:

They [mum] were really worried ‘cause the way I was dramatically losing weight; the way I looked.

Young person 1020, female, 17 years, mephedrone

None of the young people specifically mentioned coercion or pressure to enter drug and alcohol treatment but for some of them referred from the youth offending system they perceived treatment as part of an order that they were obliged to complete:

What was the reason why you attended the young people’s service?

It was part of my probation order.

Young person 1008, male, 18 years, alcohol

Involving family/social network

The majority of young people recalled identifying positive and negative influences in their social network as part of their treatment and generally said that they found this exercise useful:

Yeah it was useful because you sit back and you analyse the people around you like . . . It makes you actually sit there and think about what I was getting involved in . . . it’s just a wake-up call really.

Young person 1028, male, 17 years, cannabis

Was that [drawing a network diagram] useful for you?

Yeah, that’s why I’ve stopped hanging around with certain people.

Young person 1002, male, 15 years, cannabis

In a number of cases in the West Midlands the young people did not recall being asked to invite supportive others to their treatment sessions:

Did you ever discuss inviting people into your sessions?

No.

Did he [therapist] ever ask you, do you want to bring a friend or a family member?

No, he never said anything like that. I just wouldn’t bring any of my friends into probation with me, you know. I wouldn’t do it.

OK, so things like that you wouldn’t really like to discuss with family?

No, I wouldn’t. I don’t know, I just don’t involve other people in like what are my things.

Young person 1010, male, 17 years, cannabis

Among those young people with whom the therapist discussed inviting supportive others to their treatment, some were very clear that they wanted to receive treatment on a one-to-one basis and that their substance use was something that they preferred to deal with on their own:

No, I would never take my mum.

No, why’s that?

She’d feel embarrassed, I don’t want that.

Young person 1001, male, 16 years, cannabis

I just don’t like talking to people I’m close to. I keep everything in, I don’t like talking.

Young person 1013, female, 14 years, alcohol

It’s about me, no one else [it’s] private.

Young person 1019, male, 17 years, cannabis

The young people mentioned that the involvement of others in their treatment was an individual choice and should be discussed with them before it took place. Young people having an impact on decision-making and being kept informed of what was happening was perceived as important:

If they wanted to ask and you could make it an option, like at the first session you could say, you can have people, like family and friends if you want them here.

Young person 1033, female, 12 years, solvents

Those young people who agreed to have others attend their sessions generally found this to be a positive experience:

Did you find their [teacher] involvement supportive?

Yeah because he was saying like when . . . because stuff that he’s seen me do . . . and I don’t really notice it and he was telling me and I was just laughing because I didn’t even know what I do.

Young person 1002, male, 15 years, cannabis

In addition to attending sessions, young people with network members involved in their treatment were able to provide examples of how supportive others, mostly family members, were better able to help them:

My sister used to just stay in with me and just play on the PlayStation which keeps me out of trouble.

Young person 1021, male, 15 years, cannabis

Treatment goals

For the young people who did not receive their later treatment sessions, discussion of treatment goals did not feature strongly in their accounts. In setting and agreeing treatment goals, some young people really appreciated that the goals did not exclusively focus on their drug and alcohol use. Examples included doing alternative activities and improving family relationships:

I remember that we spoke about positive things taking over the smoking like replacing it with something, like gym or things and I have done that and that’s worked out.

Young person 1014, male, 17 years, cannabis

You said you didn’t really talk about the drug use so much, it was more about getting back with your family or getting a better relationship.

Aye, which . . . well, it’s kind of happened, so . . .

Young person 1031, male, 17 years, NPS

A number of young people also mentioned that it was helpful that there was enough flexibility within the goal-setting component to talk about life objectives:

It was good because . . . it wasn’t always talking about the drugs . . . we were talking about my life and in general . . . so it was like a mixture.

Young person 1032, female, 14 years, mephedrone

What I want to do with my life. Like my future and what I would do and how much better it would be.

Young person 1033, female, 12 years, solvents

For the majority of young people who set goals relating to their substance use this was about cutting down as opposed to abstinence:

Do you remember any of your goals?

Like cutting down and that. Smoking less a week and then certain like specific days when I would and wouldn’t.

Right, yeah, and that’s worked do you think?

Aye, aye, it helped.

Young person 1041, male, 15 years, cannabis

Treatment outcomes

A number of young people reported that treatment had a minimal or no impact on their substance use and/or social relationships. Two main issues appeared to influence the treatment journey of these young people: lack of exposure to the active components of treatment, especially not engaging a network of support and/or setting agreed goals, and/or referral from the youth offending system, with young people tending not to perceive their substance use as problematic and being resistant to involving network members in their treatment:

It hasn’t changed like nothing; you know what I’m saying? Like it hasn’t changed me in any other way like . . .

Young person 1010, male, 17 years, cannabis

For some young people their substance use continued or increased because of external factors such as relationship breakdowns or moving to hostel accommodation with peers who were also using substances:

I was going out with my girlfriend and then I broke up with her. That’s why I started smoking again.

Young person 1002, male, 15 years, cannabis

Yeah it’s got more frequent [drinking] purely on the fact of who I hang around with and who I talk to and that.

Yeah, because you live in a hostel now . . .

Yeah.

Young person 1008, male, 18 years, alcohol

Having close friends who continued to use substances made it challenging for young people to maintain the changes that they had made to their substance use intake:

It’s hard to keep away from it. When you smell it; when [boyfriend’s] got that thing in his mouth and you smell it and you still want it. I mean I wanted to stop it. But you still have that urge because of how happy it makes you.

Young person 1020, female, 17 years, mephedrone

However, other external factors such as having a supportive family could have protective effects on continued use:

It was just trying to like think about my family and that . . . like keeping them happy and that.

Young person 1041, male, 15 years, cannabis

Regardless of positive or negative environmental factors, some of the young people who on entry to treatment did not perceive their substance use as problematic admitted their ambivalence to living a life without continued drug or alcohol use:

Like I said earlier, I would happily cut down on my cannabis, but as long as I can have it when I’m annoyed or if I’m on a weekend and I should be free to have it, as long as I’m in control of it.

Young person 1031, male, 17 years, NPS

The majority of young people who successfully engaged with treatment reported experiencing positive changes in their substance use. One young person in the West Midlands said that he had cut down from up to 20 cannabis joints a week to one:

Before you started seeing [therapist] how much were you smoking then?

10 or 20 a week.

10 or 20 a week, that’s quite a lot because now you’re only smoking . . .?

One joint a week.

About one joint a week and that’s shared?

Yeah.

Young person 1000, male, 17 years, cannabis

Another young person mentioned that he changed his perceptions and drinking habits as he learned from the therapist that it was better to drink occasionally to avoid the consequences of binge drinking. Having the opportunity to reflect on their substance use and realise the impact that it was having on their lives gave some of the young people the motivation to enact change.

Most of the young people who made positive changes in their substance use also reported better relationships with the network supporters who attended treatment sessions and with their family members:

I just speak to him [teacher who took part in sessions] like a normal person because normally I don’t speak to teachers.

You speak to him. You feel that you could talk to him about things?

Yeah, like a mate.

Young person 1002, male, 15 years, cannabis

How did things change with your family?

Just more talking; well me demanding talking and they’d listen.

Young person 1020, female, 17 years, mephedrone

However, for some young people from difficult backgrounds, although they engaged successfully in treatment, any positive changes had to be considered within the wider context of their lives:

Things will never get better with me and my family. I think it’s because I was in care and they feel sorry for us, so that’s what I eventually think it is, because to be fair I’ve never had so much attention within my family since I’ve gone into care. I think it’s just trying to look good on the local authority’s behalf.

Obviously you are not using or anything like that, so you do more fun things?

Well, still like doing things that aren’t allowed by the law.

Young person 1032, female, 14 years, mephedrone

Perhaps an unintended consequence of the Y-SBNT approach of appraising social networks was that, although some young people were able to successfully distance themselves from friends who were a negative influence on their substance use, this could lead to isolation if other supportive friends or peers were not readily accessible:

So why was the relationship with your friends worse do you think?

Because they are different, they had all done the same thing as me, but I didn’t want to do it no more, but they did.

Young person 1013, female, 14 years, alcohol

In what way did it [relationships] change?

My family got closer but I had to push away all my friends. I’ve got no friends now.

You’ve got no friends now? Well who do you hang out with normally?

No-one!

Young person 1028, male, 17 years, cannabis

In terms of wider outcomes, some young people reported other positive changes in their lives. Examples included being able to budget money better, having more energy to perform other pleasant activities to increase well-being and feeling less angry and more content with themselves.

A few young people with more complex needs mentioned that they were not entirely satisfied with the outcomes from the treatment that they received. One young person with co-occurring mental health difficulties said that the treatment that he received was too short and that he was reluctant to accept further help from Child and Adolescent Mental Health Services (CAMHS):

I don’t want to talk to someone because I don’t want to be opening up to them and telling them my problems again and going over it. What’s the point like? He [therapist] hasn’t helped me. I’m still at the same stage when he opened up my case.

Young person 1019, male, 17 years, cannabis

Another young person expressed general satisfaction with his treatment but said that he was still experiencing problems with sleeping.

Treatment experience

It was apparent for the young people who engaged in treatment that their relationship with the therapist was a key factor in maintaining that engagement and the positive changes that they experienced through treatment.

This included having initial trust in the therapist and feeling that the therapist was providing relevant information:

What did you expect when you first sat down with [therapist] then?

I probably thought I was going to get grassed up. That’s what my first thought was but then when he told me it was just between us it was all right.

Young person 1002, male, 15 years, cannabis

Also important was helping the young people gain insight into their behaviour and being supported throughout the process:

It’s the way she [therapist] makes me understand everything like; I’ve still got to understand myself and [therapist] just pushes me to try to understand a little bit more . . . I don’t like talking about my feelings and the way [therapist] puts it, it makes it easier to talk about it . . . Yeah. Even if I didn’t attend them [sessions] I’d come the next day and she’d be there.

Young person 1020, female, 17 years, mephedrone

Finally, feeling able to talk openly about their thoughts and feelings, not just confined to their substance use, and having a therapist who understood their situation and responded appropriately was key:

I liked the fact that we could talk on a level, you know, I didn’t have to worry about saying anything. I could just blurt it out basically and tell him anything that I was worrying about or that I wanted to do and he’d help me with it.

Young person 1014, male, 17 years, cannabis

They [therapist] understand your circumstances and that. There are some people that pretend they know where you are coming from and I’m just, like, you are just lying really. Then you know when someone actually knows where you are coming from because they respond in a way I know that they understand me.

Young person 1032, female, 14 years, mephedrone

However, some young people, particularly those ambivalent about their substance use and referred from the youth offending system, perceived their treatment as part of probation and hence had a more negative view about the treatment that they received, although one young person mentioned that much more support was provided in contrast to their experience of adult probation services, which they were now involved with:

No, it’s just a meeting, isn’t it? So it’s just probation. I just see it as probation really.

Young person 1010, male, 17 years, cannabis

How did you find working with [therapist]?

It’s all right but it was just boring being there.

Was there anything in particular that you found boring?

Just the work!

Was there anything that you thought would improve this?

No. It’s not supposed to be good; same as probation.

And why do you think that didn’t work for you?

I don’t know, because I didn’t want to do anything.

Young person 1043, male, 15 years, cannabis

I got a lot more help from the YOT than what I do now with adult probation. There was more support and meetings every week.

Young person 1028, male, 18 years, cannabis

Treatment suggestions

There were mixed views from the young people about treatment length. A few young people referred from the youth offending system felt that the treatment that they received was too long and that six sessions should be the maximum amount. For some young people, six sessions felt like the right amount for them:

Yeah I think the six sessions were helpful. They were spread out as well. It weren’t like they were every week. Sometimes you’d say, oh I can’t come this week, you know what I mean. It would give me a little break, time to think and stuff like that was helpful. It was right, it was perfect. It was never hurried.

Young person 1014, male, 17 years, cannabis

Even those who were happy with six sessions said that the appropriate amount of sessions should depend on the individual and their perceived level of need. The majority of young people said that they would have preferred more sessions, with responses ranging from eight to twelve 1-hour sessions in a time frame of 3–6 months. A couple of young people mentioned that they would prefer ongoing support depending on their needs and for the treatment not to be time limited:

I would just have preferred for her [therapist] to just always be there . . . I don’t think there should be a limited time because if you are going into depth about something.

Young person 1032, female, 14 years, mephedrone

One young person found early morning sessions problematic and another was sceptical that a treatment based solely on talking was the best approach; a health scare would be required to motivate them to change and they would prefer integrating activities as a better treatment option:

I think if people had more activity, something to occupy their mind, then they might see it a bit different thinking these people took us out to go and do this, we could do that and we’re going to stop smoking this, do you know what I mean?

Young person 1028, male, 17 years, cannabis

Experience of the research process

Despite the fact that several questionnaires had to be completed at baseline and 3 and 12 months’ follow-up the young people tended to find this process acceptable:

Aye. If it were where you were like, oh, write this, it would have been totally different, probably would have got ripped up, but if it’s something where it’s tick or draw, it’s fine.

Young person 1031, male, 17 years, NPS

Aye, I didn’t really mind it. The questions all made like sense, to be honest.

Young person 1041, male, 15 years, cannabis

A number of young people even commented that they found completing the questions quite helpful. Reasons included observing how their substance use and/or social network changed and that young people’s experiences/lives were considered in the questions:

It changed quite a lot. The people that are around have changed. You don’t expect, oh I’ve been doing this much now, this much now, and so on. It actually makes you think about how much the intake actually is.

Young person 1008, male, 18 years, alcohol

Yes, that was interesting, that was, like, interesting to know that someone is actually considering their life, their personal life, if you know what I mean. They were all relevant, yes.

Young person 1032, female, 14 years, mephedrone

One young person suggested making the process more fun to involve some creativity; another found that one of the questions on the FES was not relevant to his situation:

Was there anything that was really difficult or you thought, oh, why are you asking this?

The one about family members hitting each other.

You thought that wasn’t relevant.

No.

So it’s a bit extreme.

Yes.

Young person 1021, male, 15 years, cannabis

In general the young people said that they would be happy to take part in a similar study in the future if they needed help with their substance use and found it a useful experience:

Yes, I wouldn’t feel that I needed it if I was older, but being a kid and under 18 then yes probably.

Why? What’s the difference do you think?

Only because when you’re 18 you’re like a responsible adult I think, so therefore, you make your own choices.

Young person 1032, female, 14 years, mephedrone

Do you think that it was a worthwhile experience taking part?

Yes. You’re going to ask why?

Why?

Because of the gift vouchers and helping other people.

Young person 1033, female, 12 years, solvents

Training and supervision for treatment delivery

The two North East therapists talked in detail about their experiences of the training and supervision received to deliver the new treatment.

Overall, training was a positive experience. However, because of the fact that they were the first group of therapists to deliver the new approach, they felt that there was a lack of case material relating specifically to young people, which would have been helpful to ground the treatment with likely scenarios that they might face. They also mentioned that it would have been useful to have more guidance on session content while dealing with a young person in crisis:

Just a bit more clarity about sort of crisis sessions and what they should look like. In future training should have examples of what’s the best way to go about that without it impacting on what we’re supposed to be doing.

Nevertheless, they felt that when issues emerged during treatment implementation that were not addressed comprehensively in training, the trainers were very responsive in providing supplementary information. Having expert trainers who they could contact as the need arose was also found to be very beneficial.

The two North East therapists held joint supervision sessions with the trainers and both found this arrangement very helpful in terms of increasing their learning and understanding of the new treatment. As an extension of this they appreciated having the opportunity to share practice with the West Midlands therapists at a group supervision session held in the West Midlands:

Like we could see how the other group [West Midlands] were working, we could see how we were doing it differently [engaging more with young person’s network] and that kind of spurred us on a little bit.

Although both therapists had a generally positive view of the training and supervision provided, in addition they reported performing reflective practice and peer supervision throughout the duration of the study, which really helped to enhance their understanding and application of the new treatment approach:

It [reflective practice] worked really well and we were able to kind of constantly feed back and reflect on our own practices and how we could improve it. We wanted it to be the best it could be for those young people.

Treatment approach

All five therapists were positive about the new treatment and could see value in the approach. They particularly highlighted the structure that it provides as a key strength, with a logical sequence of sessions that provided clarity both to the therapist and to the young person. Although the new treatment was much more structured than their usual practice, the workers appreciated that there was flexibility within the approach, enabling them to respond to the needs of the young person while still achieving therapeutic progress:

What I liked about Y-SBNT, although there may be some leeway about how many sessions that it may take to move through some of these different stages, it’s logical, it’s flexible enough but it is quite an ergonomic sort of approach.

West Midlands therapist

Experience of delivering treatment

In general, the therapists had positive experiences of treatment delivery. They mentioned that it was a learning experience for them and they got a lot out of implementing the new techniques. As they became more familiar with the techniques and materials they were able to see the benefits of the approach.

However, for one of the therapists in the West Midlands service, the study period coincided with a change in his role brought on by organisational change because of a retendering process. This had ramifications for how much personal resource they could devote to the new treatment, particularly in terms of performing outreach work for their cases:

So my experience of delivering the intervention, I really liked it. It made sense to me, I think it’s a really useful way to explore with young people and their families and their wider network change. I have felt throughout and reflecting on it since I don’t feel like I’ve put enough into it to make it work and it sits really uncomfortably with me.

Other therapists in the West Midlands service found the structure and paperwork associated with the study quite challenging. Having to record sessions and being aware of session number and the associated materials and forms could be difficult when having a number of young people in active treatment on their caseload.

Treatment engagement

The North East therapists found that the trial procedures interfered with their normal way of engaging young people. A young person is normally allocated to a therapist and will be with him or her throughout the assessment and treatment journey. However, because of randomisation the therapist might be seeing the young person for the first time and would have to commence the first treatment session of Y-SBNT. Initially, the therapists found this an uneasy process but they became used to it as the study progressed and overall they did not feel that it hindered building a relationship with the young people:

I think it would still be nice to have a little bit of time at the beginning to at least build on a little bit of a relationship because it still felt a little bit disjointed because you’re sort of straight in there, whereas it would be good to try and get to know that young person a bit more first, but again I don’t know this is maybe coming from my side of things.

Randomisation was not highlighted as a problem for the West Midlands therapists, although they said that they had mixed results in terms of treatment retention. This was in part because of ambivalence by some young people, particularly those referred from the youth offending system, towards changing their substance use. For one therapist, a role change within the organisation impacted negatively on treatment retention.

Involving family/social network

Therapists in both the North East and the West Midlands services advocated the benefits of taking a wider approach and engaging a network of support around the young person. Perceived benefits included being able to support the young person beyond treatment sessions; providing the network members with an understanding of what the young person was going through; having a co-ordinated approach to help both the young person and the supporters; and having shared responsibility for change:

It was useful for the young person because they’re getting support from the social network and at the same time for the network members they’re getting to understand what the young person is going through.

West Midlands therapist

It takes the pressure off the young person as well because making changes especially around substance misuse is massive isn’t it? And it’s not something that, you know, you have to do on your own.

North East therapist

However, in practice in both sites the therapists mentioned that for some young people there were practical and logistical restraints that made it difficult to have network members attend sessions. These included co-ordinating diaries, locations and meetings at the youth offending service, particularly in the West Midlands, where it was not normal practice for families or friends to attend appointments.

For some young people, particularly those referred from the criminal justice system and ambivalent about changing their substance use, the therapists reported facing a challenge to convince them of the benefits of including network members in their treatment. Some of this work centred on assuring the young people that they would be in control of the information discussed and that they did not necessarily have to talk about their substance use but their chosen goals.

When it was not possible to include family members as part of the network, one therapist from the West Midlands said that he then focused on supportive others such as residential care staff in the young person’s network. A North East therapist emphasised that it was important to be accommodating to who the young person wanted to be involved in his or her support network:

It’s the young person who decides who’s in their social network and who’s going to be helpful to them . . . you do have to go with what that young person wants to put on there because it’s not who we think should be supporting, it’s about what they want.

Treatment goals

Both sets of therapists felt that the new treatment worked best when they were able to facilitate alignment of the young person’s treatment goals with those of their support network. Ideally, this would involve having network members physically attend sessions. However, this was not always possible and, even in these circumstances, good work could be carried out if network members were on board with the treatment.

The West Midlands therapists tended to speak in general terms about network member involvement supporting the young person in goal setting and did not provide particular examples of this. However, the therapists in the North East spoke at length about the perceived strength of engaging a network of support around the young people to help them work towards their goals. For example, they mentioned that a foster carer had a plan of action all worked out with the young person before the end of treatment and a father had a productive discussion with his daughter by the end of which they understood each other’s position and then were able to set realistic goals:

She had the plan all set up. I just like consolidated it by writing it up and she knew what she needed to do for him and that worked well and it kind of made me feel good because I was thinking, this is good. I love this model. But I do because sometimes walking away, you just think what is going to happen to this young person.

In both sites, the therapists did not go into specifics about what the treatment goals entailed.

The North East therapists stated that the key characteristic of successful network member engagement was establishing understanding and agreement between the young person and the network members on the way forwards in terms of supporting the young person’s chosen goals:

By including network members, like everyone has got a clearer understanding about what the young person is wanting to achieve and work towards and maybe that is something that the young person might not have felt able to discuss with network members before that point.

They also emphasised the utility of the goal-setting materials in helping to plot out appropriate goals for the young person and network members. One therapist felt that the goal-setting sheet was particularly helpful:

That was the most purposeful sheet I think because it was something that could be used in any situation not just substance use. It could be used for anything and that’s why it was so good.

If goal setting facilitated by the materials was successfully carried out by the young people this could increase self-efficacy and motivation for further action.

Comparison with treatment as usual

In both sites therapists said that the new treatment differed from their usual practice. The most common difference mentioned in the West Midlands was establishing a social network of support as part of the young person’s treatment:

We don’t usually go into the details of what that family member brings which I think was a really important aspect because you get the best out of what they want to get for their goal and who could support them.

In addition, because of a new provider in the West Midlands, TAU was not well defined during part of the study period as this was before the launch of their model.

Adopting a more structured approach as part of the new treatment was identified as a key area in both services. In the North East the therapists highlighted the length of treatment time as a departure from their usual practice. In normal practice sessions continue and are as frequent as the young person finds helpful:

It [treatment] will always continue as long as the young person is finding it helpful and wants that support. Whereas obviously with Y-SBNT it was like, right you‘ve got six sessions, and within 3 months . . . We couldn’t equate the 3 months could we? We were like six sessions we’ll be done in about a week and a half [laugh].

However, initial fears about treatment time were not realised as the therapists were able to change the way that they approached providing treatment to the young people in terms of session frequency and ongoing support:

It worked and actually some of our work can be done like that I think and it can be short, sharp, and out.

One therapist from the West Midlands also emphasised the motivational component of the new treatment and associated materials. In both sites, care planning was identified as a core element of both Y-SBNT and usual practice.

Integration with existing practice

Therapists in both services saw merit in the new treatment and had integrated some elements into their usual practice. Examples included being proactive about what can be done now for the young person and not dwelling on past events and giving therapeutic work with family members and others a clear focus and purpose:

What I did like was that actually the model itself is very present. So in that sense it’s like you’re working with the immediate scene.

The therapists in the North East felt that the active components of the Y-SBNT even had applicability in other settings:

I actually feel the model is really good. I think it can be used in any situation and I think in any setting as well as in services.

However, they emphasised that training is required to ensure that the new treatment is understood and utilised to its full potential. The North East therapists felt that they had already incorporated key elements of the new approach into their routine practice.

Me and [other therapist] actually embedded it into our practice and we believed in what we were doing and because of that it kind of spurred us on to make it right in all of the situations. Because we do, if we’re honest, I want it to go to trial. I want it to get to the next stage because it’s worth it. It’s got value to it.

Experience of the research process

In general, the therapists found being involved in the research process acceptable to them. There were some initial issues with study recruitment in the North East, with therapists finding the randomisation process after triage difficult as in usual practice a young person is allocated to a therapist throughout his or her treatment journey:

What I found difficult was how we were doing the assessments at the beginning and I spent time with these young people and then they were not my young people. So it meant that was hard because they’d got to know us and then we were handing them over to someone else.

At both sites there were initial concerns about the structure of the new treatment package and how it could fit in with existing processes within services. In the West Midlands therapists mentioned that having a more structured approach had implications for service-level agreements between the different organisations involved in the care of the young people and internal processes within the agency. However, in practice, with some negotiation the Y-SBNT approach was able to accommodate multiagency action plans. In the North East the therapists mentioned that with only six sessions the new approach would not be recorded as treatment within their existing databases:

Logistically you can’t just put six sessions over a 3-month period because they would never touch the database because it wouldn’t be seen as treatment. I think it ended up being over the period of time. But really why should we have to manipulate it to suit the service rather than the young person but without it, it doesn’t get the value that it deserves.

As the North East service was a small service with some staff turnover, one therapist was allocated young people who were in TAU after the 3-month treatment period. The therapist found this difficult as she had to be careful not to expose the young people to any of the techniques from the Y-SBNT approach.

Recruitment

The North East manager reflected on the initial challenges of under-recruitment to the study in his service and how after discussion the triage system was modified so that young people who previously may not have received structured treatment, for example those in the criminal justice system who may have received an intervention from their YOT worker supported by a substance misuse specialist, were now eligible for the study. The manager reported having mixed feelings about this, as screening is an important part of the successful operation of the service:

I’m not always sure that was dead right either because I do think it’s important to do proper screening to establish whether they have the need for treatment.

He perceived an important place for unstructured education or support for young people with lower levels of substance use.

Treatment engagement

The interviewee thought that the young people should be perceived in the context of their whole lives, as those requiring a structured treatment for their substance use would have other difficulties in their lives (e.g. with regard to their families, school, the police) and should have an existing support team of professionals around them. Furthermore, the role of the substance misuse service should be to provide advice and support to those with an existing relationship with the young person. Supporting this key practitioner to deliver an intervention by having regular meetings with them and ensuring that the young person is kept informed about how the substance misuse service is involved in the treatment process is important:

Rather than having this culture where you’re just constantly looking to refer somebody else on to the next service provider, the next specialist, which in my view leads to disengagement rather than actual engagement. It’s about identifying who’s already got a reasonable relationship and how do you support them to deliver some intervention.

The manager perceived a role for the new treatment approach in facilitating communication and collaboration within the existing referral support network.

Treatment approach

There was thought to be great merit in the new treatment approach, especially the ethos of taking a wider focus on the young person’s life and involving a wider network of support in his or her treatment:

I think the big advantage is it highlights or focuses on recognising that a young person has numerous aspects to their life. Their substance misuse isn’t just the only thing in their life.

The manager highlighted that it is important for therapists to take a directive approach to empowering and facilitating the social network in a way that is helpful for the young person. He strongly believed in engaging a social network as support can be provided on an ongoing basis and is not just limited to treatment sessions. Also, having the needs of network members acknowledged and being able to review ongoing progress without involving a substance misuse specialist were seen as particular strengths of adopting a network approach.

Treatment implementation

In practice, the North East manager felt that, although some good work was carried out with the young people, some of the active components of the Y-SBNT approach were not fully implemented. His comments centred on social network involvement and goal setting. He felt that the social network was not engaged to its full extent in a number of cases. Also, when social networks were engaged there appeared to be an over-reliance on formal or professional sources of support within the network, as these people (e.g. school teachers) were usually time and role limited within the young person’s life:

If you’re sort of concentrating on the outside professionals, they have a take on it I suppose in terms of what they can and can’t do. There’s not unconditional support within that.

Some suggestions were made to address the issue of network composition. These included for the therapists to be proactive in helping the young person identify appropriate supporters; that the supporters understand the nature of their involvement, which is not time or role limited, and the responsibility and commitment required; and the potential benefit of involving positive peers in the young person’s network and not just relying on adult role models.

The North East manager also spoke at length about how he thought that the young people’s social network could have been more fully involved in goal setting. The manager’s feeling was that goals were mostly set individually with limited involvement of the young person’s social network and not having collective goals meant that the young person potentially lacked support for goal attainment from his or her network:

My interpretation of what the staff were doing was spending several sessions setting goals or agreeing stuff with the young people, that wasn’t necessarily understood or much less time was taken about actually involving a network.

Despite the perceived issues of involving networks in goal setting within the current study, the manager saw great benefit in the approach of having a shared understanding and agreement on goals if successfully implemented:

Setting reasonable goals, I suppose when you talk about young people, there is trying to get that as a shared experience almost. A shared agreement that says, so we’re all quite clear what we’re aiming towards.

Integration with existing practice

The North East manager was very positive about the new treatment approach and was keen for the other members of staff who delivered TAU to receive training:

I think understanding the philosophy in what you’re trying to achieve, the idea that actually this whole network effectively becomes the client . . . For me it’s an approach that ought to be fundamental to the way in which we work.

As also mentioned by the North East therapists, the manager believed that the new treatment approach had applicability in other settings beyond substance misuse:

I’m probably very conscious that for me I don’t just see it in terms of being a substance misuse intervention. I think it’s got direct applicability as a model right across the board and we should be utilising it.

As a treatment approach the manager felt that potentially Y-SBNT would work even better as an early intervention (e.g. in school health settings) when young people were using at lower levels and more natural networks existed in which appropriate supporters including positive peers could be identified:

Personally I think it’s got a lot of mileage and I would try and push it further. Bring it into the system earlier as an intervention.

Young people’s views

From the young people’s accounts it was apparent that their substance-related problems were symptomatic of other more challenging circumstances in their lives. The majority of young people did not perceive their substance use as a problem on entering treatment nor did their friends, who also tended to use. They were usually referred because of concerns from other agencies or their family when still living together.

The young people reported a range of responses to the new treatment. Some of the young people referred from the youth offending system were ambivalent about changing their substance use and perceived Y-SBNT as compulsory probation meetings that they wished to attend alone. Others who engaged with and were exposed to the active components of Y-SBNT reported positive benefits such as strengthening relationships and network member involvement in non-using activities. These young people appreciated that treatment goals did not have to be specific to their substance use or abstinence focused.

Network members’ views

Only two network members were interviewed, both from the North East, which perhaps reflects a combination of a lack of network member engagement and researcher focus on following up the young people. From both accounts there was a sense that their involvement was more peripheral and supportive in nature rather than carrying out agreed actions discussed in treatment. When actions were mentioned these tended to be more practical such as ensuring that the young person was available and happy to attend his or her sessions. Despite this, network members perceived positive changes for the young people in terms of better communication and strengthening relationships.

Therapists’ views

The therapists from both the West Midlands service and the North East service said that the structure that Y-SBNT provides is a key strength. Having a logical sequence of sessions with flexibility if required gave clarity to both the therapist and the young person. This was a departure from routine practice in which sessions continue and are as frequent as the young person finds helpful. The therapists saw great merit in taking a wider network approach in helping to support the young people beyond their treatment sessions. Therapists in the North East felt that the new treatment worked best when they were able to facilitate alignment of the young person’s treatment goals with those of their support network. However, there were issues with treatment implementation, particularly in the West Midlands service. This was largely because of a service retendering process causing staff upheaval during the study period. In addition, therapists in both services mentioned practical and logistical restraints that made it difficult to have network members attend sessions.

Service manager’s views

Only the views of the service manager in the North East are reported as the retendering process in the West Midlands service meant that there were several managers in post throughout the duration of the study. The North East service manager believed in the ethos behind the Y-SBNT approach of taking a wider view of the young person’s life, as his or her substance use is symptomatic of another malaise. He felt that the new treatment could be helpful for facilitating communication and collaboration within the existing referral networks. As a treatment approach the service manager spoke in glowing terms about Y-SBNT; he thought that having the whole network effectively becoming the client should be fundamental to how services work and had wider applicability. However, he mentioned that, in the current study, social networks were not fully engaged with goal setting and there was an over-reliance on professional supporters. He suggested that Y-SBNT could work better as an early intervention when more natural networks existed.

The rating scale

All available Y-SBNT and TAU sessions were rated using the adapted UKATT PRS, which included three sections: session management (maintaining structure, agenda setting, explanation of philosophy of treatment or treatment session, reviewing intersession change, consistency of problem focus, end-of-session summary); specific Y-SBNT tasks (homework, alternative activities to drinking, social support for change – general, involvement of others in behaviour change, identification of sources of support for change); and therapist style (therapist as task orientated, therapist as active agent for change, collaboration, interpersonal focus). In addition to these three sections, ‘goal setting’ was included as a task within Y-SBNT and so was also rated for all taped sessions.

The items within each section were rated in terms of frequency (the extent to which the therapist carried out each specific item) and quality (of the therapist’s behaviour) on a 5-point Likert scale ranging from 0 to 4. In line with Tober et al. ,95 for frequency, the question ‘To what extent did the therapist . . .?’ was scored as 0 = ‘not at all’; 1 = ‘a little’; 2 = ‘somewhat’; 3 = ‘considerably’; and 4 = ‘extensively’. For quality, the question ‘How well did the therapist perform the behaviour within each item?’ was scored from 0 = ‘not at all well’ to 4 = ‘very well’.

The rating manual

A rating manual was specifically adapted from the original UKATT PRS. 95 Guidelines were available for the process rating. Item definitions with examples and guidelines for rating were adapted and used for rater training.

Rater training and supervision

All ratings were conducted by an independent rater who had previous experience of rating therapy sessions utilising the adult UKATT process rating manual from a previous study. 72 Initially, the rater and supervisor (one of the developers of the treatment approach) scored the same three sessions (early, middle and end) independently; they then came together to discuss scores, referring to the manual and any notes made to ensure calibration and rater agreement. Once satisfactory agreement was achieved, the rater proceeded to rate all treatment sessions available. The main rater received weekly supervision to discuss independently rated sessions.

The sample

A flow diagram illustrating the numbers of participants who provided consent to have sessions recorded as part of the whole trial sample is provided in Figure 15. In total, 51 sessions were available for rating (Y-SBNT, n = 42; TAU, n = 9). The length of the sessions varied, with the shortest session lasting 14.4 minutes and the longest session lasting 116.1 minutes (Y-SBNT, range 17.5–116.1 minutes; TAU, range 14.4–68.5 minutes).

Considering the rates of consent is important in terms of feasibility for future work. About half of the young people (or more in the case of TAU) who consented to trial participation refused consent for session recording. Further challenges were experienced in the TAU arm even when consent had been given by the participant. Tables 30 and 31 display the number of sessions received, the number of sessions recorded and the number of sessions available for rating for each consenting young person in the Y-SBNT and TAU trial arms, respectively. In addition to the sessions with young people, there were four recordings of what were termed ‘unilateral sessions’. These were sessions delivered to members of the young person’s social network in his or her absence (but having had his or her consent). These sessions were an important component of the intervention but were not rated as the available rating scale was not designed for this purpose.

Analyses

In line with our stated aims, the main analyses were exploratory in nature and aimed to address four key questions:

1. Was the adapted PRS feasible to use with the pilot trial recorded sessions?

2. Were the therapists able to deliver key components of Y-SBNT within the sessions recorded?

3. Were the ratings for the Y-SBNT sessions and the TAU sessions different in relation to Y-SBNT components?

4. What were the components of the recorded TAU sessions?

The questions and the methods used to explore each are outlined in the following sections.

Question 4: A qualitative analysis of the components of the treatment as usual sessions recorded – ‘what were the components of the recorded treatment as usual sessions?’

In addition to the use of the PRS, for all TAU sessions a qualitative exploration of the session recording was undertaken.

The nine TAU sessions available were delivered by two therapists. One therapist completed four TAU sessions with one young person and one session with another young person. The other therapist completed four sessions with one young person. Characterising the nine TAU sessions involved listening to each session carefully and noting the presence of topics/themes under specific categories. Although it was not intended that the sessions would be compared, this occurred naturally as a consequence of the small number of TAU sessions that the researcher had to work with. The results of this exploration are very preliminary, partly because few sessions were available compared with the expected number of sessions. There were a number of reasons for this ranging from lack of consent to problems with equipment and therapists not recording sessions on some occasions when consent had been given.

The session content varied across all nine sessions with limited similarities. Some sessions were informal, brief and less structured whereas others were more formal, detailed and structured.

The sessions included a variety of topics such as information gathering, travel, exploration of drug/alcohol use and impact, completion of paperwork, education, alternatives to drug use and the future, learning difficulties, activities, place of living, harm reduction, relationships/family, goal setting, court orders, services, medication, sharing of information, communication, care plans, alcohol, techniques to aid recovery, physical, intellectual, emotional and social health, alternative therapies and general talk.

The sessions appeared to focus on information gathering about past events, recent events and the current circumstances of the young person. Information gathering at one session included finding out the circumstances of the young person leaving home, the length of time stayed at different places and any future plans related to leaving home. At another session the therapist gathered details of the events that led to a court order, the length of time of the order and what triggered the offence.

Both agenda setting and the planning of future sessions took place briefly in a few sessions. In two of the sessions, the therapist emphasised that the young person had ‘ownership’ of the sessions because they were about ‘him’ and therefore it was up to ‘him’ to make decisions about session content and change. In addition, the therapist explained that his role was a supportive one, to enable change. Generally, there was good co-operation between the therapists and the young people but during one session one of the therapists found it challenging to engage the young person because he was not communicating and remained almost silent throughout the session.

Drug and alcohol use and its impact was explored in some detail, including the length of time in treatment, the drug of choice, the frequency of drug use, the quantity used, the reasons for using, whether the drug was used alone or with others, reducing/stopping drug use or alcohol consumption, asking about a timeline for drug or alcohol reduction, triggers, cravings and thoughts and feelings after drug use.

In nearly all of the sessions, a variety of paperwork was either referred to or completed by the therapist and the young person. Paperwork included drug diaries, rating scales tapping into how things were for the young person, the therapeutic alliance questionnaire for research purposes, case closure forms and copies of the young person’s care plan and strengths. In one session consent was gained from the young person for the sharing of information to enable a letter received from a doctor to be responded to and to ascertain who the young person would like to share her care plan with. Another therapist asked one young person if he had any learning difficulties and, on discovering that he was colour-blind, offered to fill out the paperwork on his behalf. The paperwork listed earlier was not used systematically across all sessions but was used as and when it was appropriate for individual sessions. A drug diary was completed in three sessions, reference to and discussion about a care plan occurred in one session, rating scales including the therapeutic alliance questionnaire were completed in two sessions and discharge forms were completed in one session. In almost all of the sessions there was mention of a network member, usually a family member such as a mother, father, sister or brother but sometimes a partner or friends and, a couple of times, a professional, including a social worker. The therapists briefly explored the relationships that the young person had with these individuals and gauged whether there had been any recent contact. However, despite some overlap this work was not conducted in the same level of depth and detail as in the Y-SBNT sessions.

During most of the available sessions health was a topic that at times was initiated by the therapist and at times was initiated by the young person, but which was not always explored fully. In response to one young person mentioning spending more time with her boyfriend, the therapist enquired whether she was using contraception. In the main, emotional health was discussed in most of the sessions in terms of the mood of the young person, tearfulness, anger, energy levels, not eating and feeling like dying. This discussion led to one young person expressing her thoughts, linking her increased consumption of alcohol to her depression. In response, an alternative therapy (acupuncture) was explained and it was suggested that the young person consider this for the sleep problems and depression she was experiencing to complement her current prescribed medication. In addition to this, harm reduction was also a topic that was discussed and explored in some sessions, with the positive and negative effects on the body mentioned, as well as the impact on mental health and the dangers of interactions between prescribed medication and non-prescribed drugs (and therefore the importance of sharing information with the prescribing doctor). On one occasion an information leaflet about the effects of cannabis was given to the young person. In two sessions the techniques used to aid recovery were described (not implemented) to the young person: motivational interviewing/intervention, solution-focused techniques, roots to recovery, maps, the use of tools, solution-focused questions, looking forward to the future, identification and brief advice (IBA), harm reduction, coping strategies, the importance of identifying internal/external triggers in developing coping strategies and self-talk.

Fewer than half of the sessions included asking the young people to think about the future and, although goals were set in two sessions, a detailed plan was not devised to determine how educational goals would be achieved or how the young person would get involved in alternative activities to drug or alcohol use. In one session there was a review of goals set previously. There was a brief mention of three other services during two sessions to check whether the young person was still a patient and whether the young person had accessed other services.

Examples of general talk were limited to two sessions: on one occasion there was a discussion about a website that the young person had asked the therapist to visit and on another occasion there was a discussion about mobile phones. In terms of travel, one young person requested a day saver bus ticket on three occasions.

In summary, the sessions were focused on past and current circumstances. There was a high level of variation in terms of content and strategies used between sessions, with limited planning or formal agenda setting. However, this interpretation of the evidence needs to be treated with caution given the small number of sessions available and the potential for bias. The tentative results offer an initial exploration of the delivery of TAU for young people in substance misuse services that can contribute to further work. A summary of the identified components and definitions and examples, along with the frequency with which each component was identified, is provided in Table 35.

TABLE 35 - Components observed in TAU sessions

GCSE, General Certificate of Secondary Education.

(Frequency) Components	Definitions and examples
(8) Completion of paperwork	For TAU: drug diary; rating scale of ‘how things are’; case closure paperwork; copies of care plan and strengths
	For research: therapeutic alliance questionnaire; copies of paperwork
(7) Physical, intellectual, emotional and social health	Immune system low; emotionally low; not eating; spending more time with boyfriend – contraception; tearful; felt depressed – does not like taking medication for depression; feels like she has no energy; feels angry; feels like dying; started drinking more when became depressed
(7) Relationships/family/professionals	Therapist emphasises here to support; checks everything is OK with mum, sister; ‘ups and downs’ with mum and brother; father does not respond to messages; friends are users; boyfriend clingy – feel claustrophobic; unsure about relationship with boyfriend; boyfriend has apprenticeship; withdrawn from friends; social worker
(6) Sessions and support	Your sessions; plan for next session; case closure; therapist highlights not very talkative; agenda set briefly; attendance because of court order; difficult to engage; brief mention of what the next session will be about
(5) Information gathering	Includes gathering information about past events, recent events and current situations in terms of drug/alcohol use, relationships, education and bullying
(5) Alcohol	Reduction of alcohol consumption; consumption of alcohol; frequency of alcohol consumption; corner shop temptation; thoughts about consequences of drinking prior to alcohol consumption; triggers
(5) Exploration of drug use and impact	Length of time in treatment; drug of choice; frequency of drug use; quantity used; most enjoyed time of the day to use drugs; explores whether client can identify gains from using; checks whether client is using alone or with others; reducing/stopping drug use; probes client about when he thinks he can reduce his drug use; triggers; cravings; thoughts and feelings after drug use; therapist makes client aware that he can smell beer on him
(4) Education, alternatives to drug use and the future	Encourages young person to plan ahead for when school comes to an end; poses question about where young person sees himself or herself in 5 years’ time; interests; courses; commencing college course; GCSEs; sent home from school when falls asleep; explores how young person spends his or her spare time
(3) Medication	Risperidone for hearing voices; antidepressant for depression
(3) Harm reduction	Positive and negative effects on body; mental health; dangers of interaction between prescribed and non-prescribed drugs; provision of information leaflet about effects of cannabis to young person
(3) Travel	Client wants a day saver bus ticket
(2) Others identified on only two occasions	Alternative therapies; techniques discussed to aid recovery; goal setting; general talk; court orders; other services
(1) Others identified on only one occasion	Learning difficulties; discussion of activities; sharing of information; place of living; communication; care plan

One further analysis was conducted to ascertain how key components of Y-SBNT were delivered across the early, middle and end stages of the recorded intervention sessions. This was important to explore whether there was potential ‘therapist drift’ resulting in aspects of the intervention being applied at early stages, with therapists reverting back to usual practice at later stages when the Y-SBNT components would not be so evident. To explore this, all available recorded Y-SBNT sessions were categorised into ‘early’ (consisting of session 1), ‘middle’ (consisting of sessions 2–5) and ‘end’ (consisting of session 6, the final session; when session 6 did not take place but session 5 did, the latter was placed in the ‘end’ category). All 10 components involving Y-SBNT-specific tasks as well as therapist styles were then represented in a series of graphs (Figures 16 and 17), which illustrate the mean frequency ratings across the three stages (the full data table can be seen in Appendix 20). As shown in the graphs, there was no marked drop in frequency of key rated Y-SBNT-specific components during the middle part of the intervention, suggesting little evidence of therapists moving away from these components as therapy progressed.

FIGURE 16.

Mean frequency scores for Y-SBNT: specific tasks. (a) Homework; (b) alternative activities to drinking; (c) social support for change: general; (d) involvement of others in behaviour change; and (e) identify sources of support for change.

FIGURE 17.

Mean frequency scores for Y-SBNT: therapist style. (a) Therapist as task orientated; (b) therapist as active agent for change; (c) collaboration; (d) interpersonal focus; and (e) goal setting.

Overview of young people’s involvement in the study

As discussed in Chapter 2, during phase 1 young people were supported to work alongside the research team to ensure that the intervention was acceptable and relevant to our target groups and reflected the views of service users and their families. During phases 2 and 3 (see Chapter 3) young people were involved in tasks associated with the feasibility trial, including the design of key trial documents such as the recruitment leaflet and information sheet, the production of training materials, advising on data collection tools, data analysis and interpretation, reporting and dissemination. Further details on activity, outputs and outcomes are available in Appendices 2 and 3.

Summary of young people’s involvement in the study

• In total, 17 young people were involved as young advisors in some capacity.

• The young people came from five areas of England (the West Midlands, the north-east and the west of England, London and Oxfordshire).

• The 17 young people included 12 females and five males ranging in age from 16 to 21 years, with an average age of 18 years.

• We met 10 young people once, five twice and two on five or more occasions.

• Over the course of the study we held a total of 20 face-to-face meetings with young people. In between meetings, contact was maintained by post, text message, e-mail or telephone call, depending on the preferences of the young people involved.

• Seven newsletters were sent to young people during the course of the study with updates on the project, information about how young people had been involved and upcoming opportunities.

Young people’s involvement in the study was led by two members of the research study team (LM-B and LT) with expertise in public involvement and research with young people affected by substance misuse, respectively. They contributed to the development of the initial proposals and were then employed as public involvement leads to recruit young people and organise meetings with young people and family members, develop methods for the involvement process in collaboration with other co-investigators and produce reports of the results of this work. Various research team members contributed to the meetings with young people through presentations, planning and facilitating sessions and feeding back on how this involvement had informed the study (see Appendix 2). Public involvement was also a standing item on all TMG and TSC meetings, at which the public involvement leads reported back on public involvement activity and sought feedback and ideas from the TMG and TSC. As the project progressed and the need to adapt initial plans for public involvement became apparent, TMG and TSC meetings were key to the development of new ideas on how best to involve young people in the study.

Initial plans

The original plan for young people’s involvement in the study, as outlined in Chapter 3, was to have a YPAG of 10–12 young advisors aged 12–18 years with previous experience of accessing substance misuse treatment services who would be actively engaged throughout the project. This is a common model for young people’s involvement in health and social care research used by, among others, the Medicines for Children Research Network (now the Clinical Research Network Children,126 the National Young People’s Mental Health Advisory Group, the DECIPHer (Development and Evaluation of Complex Interventions for Public Health Improvement) ALPHA (Advice Leading to Public Health Advancement) young people’s group127 and the National Children’s Bureau’s public health advisory group known as PEAR (public health, education, awareness, research).

However, recruiting young people to an advisory group proved problematic. In the first instance there were delays that arose from a request by the NHS host institution for NHS ethical review and approval for this phase of the project if recruiting young people through trust NHS services. Normally, ethical approval is not needed for the active involvement element of NIHR research, even when people are recruited through the NHS128 but, as young people’s involvement in the study was also part of a Doctor of Philosophy (PhD) research study by one of the public involvement leads, the trust requested NHS ethical review. The National Research Ethics Service (NRES) was consulted during February 2013 and confirmed that no NHS ethics review was required for public involvement in the study, with most aspects in the early phase of adaptation of the intervention seen as a service-related evaluation. However, ethical review was needed for the aspects of this work that were being used for the PhD research study. This was obtained from the University of the West of England Life Sciences Faculty Research Ethics Committee (reference number HLS/13/06/88) following which full approval to proceed with plans for public involvement and the associated PhD work was given by the NHS host institution.

These discussions and processes led to some delays and also a review and widening of recruitment sites for young advisors. Initially, we worked with a national drug and alcohol treatment charity, as it expressed an interest in the project and thought that it could be an exciting opportunity for young people who had used its young people’s services. However, initial recruitment proved a lot slower than we, or the charity, had anticipated and the first three meetings (in September, October and December 2013) were poorly attended. At each meeting four to six young people were expected but only two attended on the day and the young people who did attend varied from meeting to meeting.

Revising plans

By early 2014 we realised that the traditional format of a single-location advisory group would not on its own be the right model for working with the young people whom we wished to engage. We therefore moved to a more flexible model that focused on fitting involvement activity around what worked for the young people whom we were trying to involve. Young people’s involvement was no longer centred on regular meetings in London, but started with smaller consultations based at services or other locations that were familiar to them and also included e-mails, text messages, telephone conversations and correspondence by post when this was more convenient for the young people or timely for the project (e.g. if young people’s input was needed in between scheduled meetings). If interested, young people involved in these consultations were then given the option of ongoing involvement and invited to become a young advisor, but there was no obligation to do so. We hoped that this more flexible approach would give young people, and services, a chance to find out more about the study and potential involvement without making an ongoing commitment at the outset. We also sent out regular newsletters to all young people with whom we were in contact so that, even if they were unable to be involved for a period of time, they were kept informed:

It’s been good to get the newsletters and read about what we did and how it’s helped the project.

Young advisor at November 2015 meeting

Another factor that affected how and when young people were involved in the project was that many of them were also working, often in low-paid jobs for which they would not be paid if they did not work or from which it was difficult to take time off. In many YPAGs young people are given a gift voucher for their time and this was the model that we initially adopted: all young people involved received reimbursement of travel expenses and, initially, a £20 voucher for each meeting attended (payment was later revised; see below). However, when we realised that we were going to have fewer young people involved in the project and that loss of income was a barrier to involvement for some, we decided to adopt a different approach. After considering INVOLVE and other guidance on payment for public involvement,129 the TMG agreed in September 2014 to offer young people involved in the study payment at a rate of £10 per hour on an ongoing basis for involvement activity, plus travel and subsistence expenses as before. The time needed for any work undertaken remotely (such as commenting on draft documents or, writing sections of articles and reports) was negotiated in advance. Payments could be in the form of Love to Shop vouchers as before or could be paid directly into a bank account via a claim form. Young people could also choose not to accept payment. If young people were working they were advised that this income would need to be declared for the purposes of tax and national insurance and that it was their responsibility to do this. If they were claiming benefits they were advised about the necessary rules on earnings and signposted to relevant guidance and sources of information on these matters. One young person was supported by the NIHR Benefits Advice Service for involvement130 to manage issues around payment for involvement and benefits.

This new approach led to the following engagement with young people (further details are provided in Chapter 3 and Appendix 2):

• January 2014 – Consultation meeting in the West of England attended by three young people, one of whom continued to be involved for the remainder of the project, mainly on a one-to-one basis, and is one of the young contributors to this chapter.

• March 2014 – Consultation meeting with two young people at a service in Oxfordshire. This led to two completed young advisor applications and both young people saying that they wanted ongoing involvement, but this did not materialise despite following up with both the young people and the service.

• April 2014 – Piloting questionnaires with two young people in the West Midlands led to one young person expressing an interest in ongoing involvement. She has been involved on an ongoing basis ever since, mainly one-to-one but also attending one meeting with other young advisors, and is one of the young contributors to this chapter.

• Contact with a service in the North East (also one of the clinical sites for the feasibility trial) where Y-SBNT therapists were really enthusiastic about supporting young people’s involvement. This resulted in two therapists accompanying young people to three Y-SBNT young advisor meetings during 2015: four to a meeting at Birmingham University in April, three to a meeting in York in September and two to a meeting in York in November.

Evaluation and feedback from young advisor meetings

Young people’s views on the meetings were captured through a range of evaluation activities, as well as through summary notes produced by the public involvement leads and circulated to the research team after every meeting. These evaluation data informed both future meetings during the study (Table 36) and also plans for a new model of involvement.

Models of involvement

Much of the limited but growing literature on young people’s involvement in health research focuses on the benefits, impact and outcomes of that involvement (e.g. Boeck and Fleming,131 Moules132). As outlined at the start of this chapter, young people’s involvement in health research is also dominated by the model of the YPAG. But there has as yet been relatively little discussion in the literature about who is, or is not, involved in research and how different models of involvement might work for different young people.

‘Children and young people’ are not a homogeneous group: age and other aspects of social background such as race and ethnicity, disability, social class, family background and use of services ‘intersect as aspects of who [young people] are, their social position, and what researchers need to consider’133 in designing research approaches appropriate to the young people they wish to involve. As outlined earlier, it became apparent fairly early on that our original plan for an advisory group of 10–12 young people, who would be actively engaged throughout the project, was unlikely to be successful with a group of young people who had complex lives and little or no experience of either involvement or research. We found that input from young people to this research study needed instead to be dynamic and flexible; this enabled us to create new opportunities and ways of working more suited to those we wished to involve (p. 221),131 in this case young people who had experience of using substance misuse services:

Perhaps that sort of [YPAG] group structure isn’t going to work with these sorts of young people in a project like this. And I think it is good to have a group and people fire off each other . . . but I still think the individual-level stuff that we’ve had has been really good and that’s worked well and in a more natural way and perhaps this group need that level of support . . . because they’re pretty chaotic, [with] multiple problems so you have to go to them, they’re not going to come and travel to some place and so this individual level and trying to hold them over the course of the project might be an alternative model.

Researcher, TMG public involvement review focus group

My mental health has sometimes made it difficult for me to be involved [in the project] . . . for example that [recent young advisors] meeting, I would have liked to have been part of it but due to my mental health I couldn’t even come out to meet in a coffee shop, let alone travel to a meeting . . . But I like how, if I can’t come to a meeting, we can have a phone call instead or you’re just a text [message] away, or we can do stuff by post. I’ve not always been well enough to come and see people face-to-face but that doesn’t mean I couldn’t be involved in things, whereas some organisations I’ve been involved with would have said ‘you’re ill or you couldn’t come to a meeting so you can’t be involved’.

Young advisor A

What I’ve liked is the flexibility. I’ve never had to make my own way to one of these meetings . . . of course it is a research project so there is a certain amount of formality about it but [individual participation] keeps it relevant and convenient for to the young person which is always good . . . . I have spent the majority of my time with [public involvement lead] in this project and I like to think we have also built a friendship as well.

Young advisor B

However, the individual model of one-to-one work with one of the public involvement leads could lead to young people being more isolated. One young person, who was mainly involved in this way because of availability and geography, said that, although she appreciated the flexibility and convenience of meeting at a time and place that suited her, she would have liked to meet other young people involved in the study and that working in a group might help young people feel ‘less of a guinea pig’:

The individual model, while it’s worked very well in some ways, I wonder because [young person] has said to us it would be great to meet the more of the [research] team and other young people, and we’ve not been able to get to a stage where we’ve been able to do that.

Public involvement lead, TMG public involvement review focus group

Hanley et al. 134 identify the different levels of public involvement in research as consultation, collaboration and user controlled. Although we had started with the intention of this being a collaborative project, what we ended up with was something that actually started more as consultation (i.e. one-off meetings) and built up to more collaborative working if this was something that young people were interested in. This echoes Kirby et al. ’s135 non-hierarchical model of young people’s participation, in which the appropriate level of involvement is determined by the circumstances and young people involved:

Being on their journey with them is more important than getting results or directing them. We have to do this [involvement] at their pace.

Therapist supporting young people’s involvement

Franks136 talks about the need when ‘doing participatory research with children and young people at the margins of society’ to consider ‘creating pockets of participation’ (p. 15). This is often what happens by default in many projects but, by making it transparent and intentional, young people can better understand their potential role and discuss with researchers when, how and in what ways they would like to be involved in the research process. 136

A young advisor’s perspective

You need to go out to more substance misuse services and say ‘can you help us . . . can you provide us with [access to] young people if they are interested?’ and explore different avenues. Put the word out: maybe get a day with a drug and alcohol service, ask them how many young people they’ve got on their caseloads, just introduce the project and give them more information about what we’re doing and what we want to do, and seeing if they are interested. If you don’t get the results you want from one place I’m sure there will be a different service to go to.

To get young people interested in a project like this I think you need another young person, someone who’s walked the walk sort of thing, so they can understand on a deeper level. Personally, if I was to see something on the internet or on a piece of paper, I probably wouldn’t be that interested, I like to see people and be able to find out about things face-to-face if I can. But if you can arrange a day at a service you could see if you can talk to some young people while you’re there. Then you can give them the options of just having a chat or getting more involved, but with no rush and no pressure so more like: ‘what do you lot think about the project? Would you like to get involved or not like to get involved? If so how would you like to be involved?’ Just get them to talk to you.

Young advisor A

Facilitators’ summary of feedback from young advisors at the April 2015 meeting on what future involvement could look like

• The idea of involvement being linked closely to a few services was seen as important. For example, some of us would probably have been unlikely to attend this first meeting without support from staff.

• Some form of group with ongoing involvement based around face-to-face meetings is still the best way to do involvement, but this should be flexible as people might want or need to move in and out of the study and one-to-one meetings might work better for some young people.

• There should also be the opportunity for us to be involved between group meetings. This could be by e-mail or post and/or face-to-face, for example one-to-one meetings and/or service-based local groups combined with longer meetings where everyone comes together from different areas.

• Ideally you should start setting up a group of young advisors a year or so before the study begins. Young people could then be involved in proposal development and possibly developing films and other materials about their experiences, which could be used in training and recruitment. We would also like to be involved in a few different projects, not just one.

The young people, research team and TSC felt that, ultimately, some form of ongoing group was the best way to involve young people, but that this needed to be more flexible and young people centred than a fixed-location YPAG with a largely static membership. As outlined by the young advisors above, any model for involvement should ideally be developed in consultation with young people – recruiting those with an interest and then working out with them what they would like to be involved in and how best to do this. Young people and researchers both said that they preferred face-to-face meetings, either in groups or individually, if possible. This could be combined with telephone calls and communication by post when input was needed in between meetings.

Proposed involvement model

• Hub and spoke model with a core ongoing group of young people, who might change over time, alongside one-to-one and small group work and one-off consultations. Depending on the availability and interests of the young people involved, this group could meet regionally (e.g. with support from substance misuse services working with young people) as well as coming together nationally (e.g. through day meetings or longer residential meetings) and possibly provide advice and support to other studies in the field.

• The group could be a combination of older young people (e.g. mid- to late 20s) who are able to look back on their experiences and possibly more short-term and flexible involvement with young people currently using services. Some of these may be young people in adult services.

• Involvement would be closely linked to a few services, with an element of a public involvement budget set aside for these services to recruit and provide ongoing support to young advisors as well as hosting young advisor and/or consultation meetings.

• Young people would be involved in recruitment for public involvement, both initially during visits to services and through ‘snowballing’ through young people recruited as advisors.

• Ideally, a group would be established before a project started, so that young people are able to be involved in the development of the proposal and support the recruitment of a new group of young advisors, as well as possibly being able to be involved in other studies in the field:

Before you even get to a full trial, if you are working with a group of young people you can build up trust with them, build up a relationship with them in the hope that they might go on to be a kind of core network.

TSC public involvement review focus group

• Young people would be involved at all stages of the research, from proposal development to dissemination and, when possible, would be given opportunities to be actively involved as part of the research team as well as being consulted by them.

• The main role of the public involvement lead(s) in this model would be to plan involvement activity, facilitate links between young people and researchers through the services and evaluate this involvement, with services taking the lead on recruitment and retention as above. All members of the research team would ideally also be allocated time to feed into, and plan for, public involvement (particularly those who, unlike the public involvement leads, are directly involved in the study on a day-to-day basis). Management support is also essential to embedding young people’s involvement in the study rather than it being seen as an ‘add-on’:

What embedding participation in a project like this would mean is that we don’t think of it as special or anything different it’s just normal to think about it as part of the project. To have the young people there in a reciprocal relationship.

Researcher, TMG public involvement review focus group

• Linked to wider discussions within the NIHR and elsewhere about making young people’s involvement in research more inclusive, diverse and accessible.

• Parental involvement should also be considered but, given the complicated family relationships of many young people involved in the study, it may be necessary to broaden this out to other people who have played an important role in young people’s lives and recovery.