Introduction

Internationally, safety of care in child and adolescent mental health has received limited attention. Attempts to understand this area have mainly focused on issues of safety in relation to safeguarding [1] on the one hand, or lack of access to services [2] on the other. There is a call for clinicians, service developers and researchers to consider harm and safety more generally in child and adolescent mental health service (CAMHS) [3••]. America and others have begun to initiate discussions on the possibility of harm caused by psychotherapy [4, 5]; however, the lack of shared definitions as to what constitutes safety and harm present ongoing challenges [3••, 4]. To start to rise to these challenges, Wolpert et al. [3••] outlined a possible framework for considering harm in relation to child and adolescent mental health provision in Fig. 1.

Fig. 1
figure 1

Framework for consideration of patient safety in child and adolescent mental health.

Within the framework, safety and harm are considered in relation to terms of three aspects:

  1. 1.

    Ineffective engagement—whether through lack of services available or through lack of uptake of existing services;

  2. 2.

    Ineffective practice—examples of which include the prescription of non-evidence-based treatments or a breach within accepted guidelines such as waiting times;

  3. 3.

    Adverse events—where clear negative outcomes are associated with particular events, such as youth death, including suicide.

The latest findings and developments in each of these three areas are considered below, as well as key interventions that have been developed within the UK at each level (Tables 1, 2 and 3).

Table 1 UK interventions for harm caused by lack of engagement with appropriate help
Table 2 UK interventions for harm caused by ineffective practice
Table 3 UK interventions for harm caused by adverse events

Safety in relation to service engagement

Global prevalence rates suggest that around 25 % of young people experience mental health issues, with fewer than half receiving mental health specialist treatment [6]. There is also evidence of growing rates of mental health difficulties, particularly emotional problems in teenage girls [7] and increased rates of self-harm being reported over the last decade [8]. Within the UK, these increases are occurring alongside concurrent extensive disinvestment in specialist provision with many areas already experiencing 25 % cuts in funding [9], and less than half of CAMHS reporting being able to provide appropriate crisis access for young people [9, 10]. Thus, the primary safety issue is the need to consider safety for those young people not able to even currently access services.

For those not accessing specialist mental health services, there is some evidence that they do draw on educational support and/or primary care [11, 12]. Schools are increasingly looked to as a site for future work for mental health promotion and intervention [13], but there is a need for greater clarity on what works best in this context [14••]. In addition, there is an interest in increasing access by use of online resources, such as online counselling and support [15], though the efficacy remains unclear [16].

Harm from ineffective engagement may also occur when service provision is available, but young people fail to make use of what is on offer. For those young people who choose not to access services, reasons include fear and stigma [17, 18]. Increased education around mental health literacy may improve this [19].

In addition, a lack of knowledge about how and where to access services may also affect whether people attend [18]. There are an increasing number of websites to aid people finding help such as Find Get Give (http://www.findgetgive.com), the Youth Wellbeing Directory (http://www.youthwellbeingdirectory.co.uk), ReachOut (http://www.reachoutuk.org) and MyCAMHSChoices (http://mycamhschoices.org). A systematic review into online help suggests that young people view websites favourably, stating they were satisfactory and they would use them again or recommend to a friend. However, more research needs to be undertaken to establish whether online services increase help-seeking behaviour [20].

To tackle current issues with service provision and reduce the risk of harm, increased investment is needed, as well as new ways of maximising existing resources. Novel models of service provision include that of THRIVE, which reconceptualises CAMHS into resource-homogenous groups based on their needs, as well as drawing distinctions between treatment and support, and between self-management and intervention [21]. Aligned with this, work from payment systems in CAMHS also highlights distinct needs-based groupings for those seeking help and support [22]. Reorganising services according to such frameworks may allow for more transparent discussions between providers and users, options for more targeted performance management and potential for more targeted funding which could help with service engagement [23].

Another barrier to engagement with help is people dropping out of treatment prematurely. Attempts to increase attendance rates have focused on reminding service users, or having opt-in appointments. Drawing on data from both adult and child mental health populations, research suggests that when utilising an opt-in system, median non-attendance rates fall from around 27 % to around 4 % [24]. However, opt-in systems may not tackle underlying barriers to non-engagement [24] and may exclude those who need treatment such as those undecided or scared about attending [25•].

Engaging parents prior to a CAMHS appointment may be useful as it can provide the opportunity for clinicians and parents to discuss opportunities and barriers to attendance. In one study, in which a key worker contacted the parent/guardian to discuss items such as why the referral was made and logistics in getting to CAMHS, appointment non-attendance was significantly less likely than compared to the two control groups [25•]. Moreover, this intervention was found to have similar outcomes regardless of ethnicity and deprivation level [25•].

For children, dropping out prematurely from treatment may be considered a safeguarding issue, as attendance may be considered to be in the best interests of the child [26]. Moreover, non-attendance may be the result of the parent or carer rather than child choice [27•]. Missed appointments have been highlighted as a warning sign in a number of serious case reviews and child death reviews [27•].

Developing safe care in mental health for children and young people thus requires the understanding and reinforcement of guidelines around missed appointments and safeguarding. There is evidence of inconsistencies and variability as to what should and does happen following a missed appointment in NHS organisations [28], as well as staff being unsure what the guidelines are [29]. Some academics have proposed a change in guidelines suggesting the ‘catch all’ category of did not attend (DNA) distorts the nature of non-attendance [27•]. Instead, the term ‘was not brought’ (WNB) has been suggested, in order to highlight the lack of autonomy and dependence of the child or young person, as well as potentially leading healthcare professionals to consider why the child was not brought [27•].

Young people who drop out of treatment early are more likely to experience continuing mental health difficulties and functional impairment when compared to those who stay in and finish treatment [3033]. Satisfaction with the service, both for young people and for parents/guardians, is an important predictor of engagement and treatment completion [34], yet young people and parents do not feel listened to [35, 36] and state that they find a lack of continuity of care unhelpful [18].

One approach that may address disengagement with services is an emphasis on shared decision making [37]. This is an interactive process during which patients and practitioners collaborate in choosing healthcare, and sees the the patient as an expert about themselves and a clinician as an expert in the treatment [38]. Its application to CAMHS is still in its infancy, yet tools and approaches implementing shared decision making may increase knowledge, improve clinical outcomes and may be useful in facilitating conversations about values and treatment options [37, 39, 40]. Common engagement approaches such as assessment, accessibility promotion, psycho-education about services, homework assignment and appointment reminders may also be a way of keeping young people and families in services [41]. A review of common elements of successful treatment engagement within CAMHS suggests that assessment and accessibility (defined as ‘measuring a client’s strengths/needs through a variety of methods’ and ‘any strategy used to make services convenient and accessible in order to increase participation in treatment’, respectively) were consistently associated with positive outcomes, such as attendance and adherence [41].

Another way of addressing some of the concerns of young people’s lack of satisfaction with services could be through using frameworks such as the common assessment framework (CAF). An investigation by the Local Authorities Research Committee (LARC) found that CAF improved outcomes for young people and their families by treating them as equal partners, as well as providing a consistent lead professional, which facilitated better working relationships between professionals and service users [40]. Local authorities also emphasised the need to engage young people and their families, and reported positive benefits in doing so, including young people and their families better understanding their own and other’s needs, improved parenting skills and techniques, improved relationships between families and professionals and new opportunities for multi-agency professionals to work with parents [40].

Safety in relation to ineffective practice

Worldwide, there is a shortage of trained professionals working in child and adolescent mental health; this is particularly true in developing countries where there are not enough professionals trained in basic child mental health treatments [42]. In the UK, CAMHS staff traditionally provided a wide variety of interventions with little standardisation. Concerns have also been raised about a range of practices that may be limiting effectiveness of intervention in the UK [43, 44]. Up until recently, most practitioners had not received training in some interventions with the strongest evidence base [44, 45]. Moreover, there is evidence that the majority of people accessing services are being seen for a shorter amount of contact time than recommended in best practice guidance [46], though it is not yet clear how far this relates to patient choice. The setting up of practice research networks (PRN) and collaborations such as the learning collaboration of CAMHS providers led by two of the authors of this paper—Child Outcomes Research Consortium (CORC)—may be a way of bridging the gap between researcher and clinicians, focusing attention on potential for harm from ineffective practice, and developing effective practice by involving clinicians and increasing generalisability of findings [47]. However, for clinicians to be engaged in research, it must be clinically helpful to them [48]. This could be through the collection of data for monitoring the progress of a patient or using the outcome measures in supervision to reflect on practice [49, 50].

Over the past two years, best practice guidance has coalesced around the Children and Young People’s Improving Access to Psychological Therapy programme (CYP IAPT) which has brought together a common model of best practice [51] through an emphasis on outcomes (highlighting session-by-session feedback and review), patient preferences (shared decision making) and the use of evidence-based therapies as recommended by the National Institute of Clinical Excellence (NICE) [14••, 5153]. Markers of value and quality have been identified and are quality assured, processed and accredited through a number of organisations and programmes including the Quality Network for Community CAMHS (QNCC), Youth Wellbeing Directory with ACE-V (ACE-Value) Quality Standards, Choice and Partnership Approach (CAPA) and Child Outcome Research Consortium (CORC) [51]. Similar work addressing evidence-based programmes and practice is also being undertaken elsewhere such as by the Substance Abuse and Mental Health Services Administration (SAMHSA) in the United States.

However, what constitutes reasonable expectations of effective intervention and the possible harm from ineffective treatment (whether supported by the research evidence-base or not) remain under-discussed. There is no literature that we are aware of that considers the potential harm of loss of attendance at school, impact on self-image or self-esteem, or impact on relationships against the positive impact of any treatment, let alone the harm of being seen when treatment does not effect change [3••].

Research suggests that evidenced-based youth psychotherapies demonstrate only a modest treatment effect and are moderated by youth location and assessment characteristics [54]. There is mixed evidence as to how far current service provision constitutes effective practice. International research into the effectiveness of routine CAMHS finds the majority of children receiving community-based standard care do not show clinical improvement [55, 56] and generally show less improvement than in RCT studies. Yet, there is little literature to guide clinicians about how to discuss treatment failures with children and families or how to decide when continuing to see someone is counter indicated.

A key focus in recent years has been on the use of routine outcome monitoring and patient feedback throughout therapy to consider how treatment is progressing and make changes as relevant. Whilst there is some work from Lambert, Bickman, Miller and others [5759] tracking outcomes and considering off-track cases, there is still little developed literature on when to determine that treatment is failing and how to decide if another option would be more beneficial. Greater work is urgently needed in this regard.

Safety in relation to adverse events

The lack of a clear definition on what constitutes an adverse event in mental health, as well as how to adequately capture this in research and clinical practice, remains a challenge [4]. What we do know about adverse events is often drawn from serious case reviews, which have consistently show us that there is a need for stronger interagency working and collaboration around the sharing of information [27•].

Rather than outcomes from case reviews, wider learning mechanisms such as multi-agency training have been developed to try and clarify accountability and learning. Particular praise has been given to the Multi-Agency Critical Incident Experience (MACIE) approach which aims to foster characteristics such as shared responsibility and strong leadership, challenge partner professionals and develop a common language through simulated cases [60]. Failure Modes and Effects Analysis (FMEA) has also been highlighted as a learning mechanism [60] as it can generate collective learning including identifying remedial actions that can prevent adverse events within complex and high-risk processes. FMEA has already been applied to domestic abuse and child protection, with it being found that it was useful in generating a map and applying corrective actions where possible, as well as reported enhanced multidisciplinary, multi-agency learning around domestic abuse, though it was time and resource intensive [61]. It has been suggested that paediatric nurses may want to use such an approach for child protection [62] but this could also be used in CAMHS.

A scoping review into the development of models and approaches to address the difficulties of inter-professional working between NHS and local authority social services identified a number of initiatives being trialled [63]. For multi-agency team work, it was found that the common assessment framework (CAF) formed an integral role, as it can be used to create an inter-professional group of ‘virtual’ professionals around the individual, including principal professionals to organise and facilitate care [63]. Other approaches identified included the co-location of professionals through Multi-Agency Safeguarding Hubs (MASH) to facilitate information sharing [64]. Whilst there are some initial positive findings for MASH, such as better information sharing [63], it was concluded that for all new initiatives identified, there were too few studies to conclude their effectiveness, particularly for outcomes relating to young people and families [63]. The implementation of new multi-agency inspection frameworks in April 2015 [65] may help provide such evidence when data has been collected and analysed.

Approaches currently being used within CAMHS such as Adolescent Mentalization-Based Integrative Therapy (AMBIT) may also be useful to foster interagency collaboration and work with young people who have severe, multiple needs and who may not actively engage with professionals or access mainstream services [66, 67]. Initial evaluation of two team’s that implemented AMBIT suggested that one teams’ learning focused on an improved understanding of the client group, a change in perception of the effectiveness of the team and a change in attitude towards outcomes monitoring itself. The other team's learning highlighted improved explicit knowledge of core outcomes, making sense of different perspectives about outcomes and changing practices around what was not working [68].

Adapting existing approaches used in physical health may also improve safety through addressing a lack of information sharing and collaborative practice. One such approach is the huddle, which brings together staff to disseminate and share information in a fast, focused and collaborative way [69]. This brief and rapid exchange of information allows staff to prioritise the day’s work, share information regarding patients, plan for changes to patient flow and deal with staff adjustments [70]. Within paediatrics, results have been positive, with huddling being attributed to improved efficiencies and quality of information, sharing, increased levels of accountability, empowerment and sense of community, a lowering in ‘unsafe transfers’ and an increase of days between incidence [71, 72]. Huddles are currently being trialled across a number of paediatric departments in English hospitals [73], and if successful, CAMHS teams may wish to adapt and trial it.

Understanding specific adverse events caused by psychological therapy is still under researched [5]. If psychotherapies have the potential to alleviate mental health issues, they must also have the potential to do the opposite as well [74]. Indeed, many CBT clinicians have observed adverse events from people they were treating [75]. An investigation into clinical trials involving psychological treatments found that reporting of adverse events caused by psychotherapy did not make it into any final reports, with the authors concluding that the National Research Ethics Service guidelines did not adequately capture this area [76]. Ideas about how to define and address such adverse events have been outlined [4, 77••].

Work being undertaken within this area includes the University of Sheffield’s Adverse Effects of Psychological Therapies (AdEPT) project. Results have yet to be published, though findings will be limited in their application to CAMHS as AdEPT focuses on those aged over 18 and receiving psychological therapies. Moreover, with increasing emphasis of the use of online and electronic treatments in CAMHS [78], attention should also focus on harm caused by these modalities. Findings from one study have demonstrated negative effect effects brought on by internet interventions [79]. This may be particularly relevant to young people who have grown up with and are confident users of technology [80].

Countries collecting national data, such as the US through the Substance Abuse and Mental Health Services Administration (SAMHSA) and National Mental Health Services Survey (N-MHSS), may be able to use this data to better understand services treatment offered and forecast future requirements [81]. Within the UK, child and adolescent mental health data is planned to be collected by the Health and Social Care Information Centre (HSCIC) from January 2016 as part of the Mental Health Services Data Set (MHSDS). It is not yet clear when these data will be available for researchers; however, the lack of nationally available data has been highlighted in several key reports [82, 83] and is seen as critical to underpinning long-term sustained quality improvement across the health sector, education and social care [84].

Conclusions/considerations

Internationally, safe care in CAMHS has been until recently an under-developed concept. Within the UK, a number of interventions are being trialled which children and adolescents may benefit from. For example, new models for conceiving of potential harm and initiatives for embedding a safety culture in CAMHS are emerging. These include initiatives designed to address barriers to engagement with services, including greater use of online support, schools-based and online education about sources of help (e.g. MyCAMHSChoices, the Youth Wellbeing Directory and Find Get Give); initiatives designed to promote effective practice (conceived as requiring a combination of evidence-based practice, service user preference and regular review of outcomes and feedback) from UK sources (e.g. CYP IAPT, CORC, QNCC and CAPA) and abroad (e.g. the Substance Abuse and Mental Health Services Administration N-MHSS). Lastly, initiatives designed to review and address adverse events include the emerging research agenda considering the potential iatrogenic impact of talking therapies, such as that explored in the AdEPT study, and the more meaningful use of data across systems.

However, it should be noted that due to present levels of challenge and funding services, both in the UK and abroad, the risk of harm remains high. The emergence of new models and approaches to CAMHS, as exemplified by the THRIVE model [21] and emerging payment system work [22], provides opportunities for embedding these principles, supporting both values-based and value-based service delivery [9, 51] to ensure that consideration of safe care is at the heart of the future development of child mental health services. Whilst this initial work in the UK is encouraging, how this may translate to healthcare systems abroad remains unknown. As such, such initiatives should be piloted, and process and outcome measures observed before any formal implementation.