Factors within the clinical encounter that impact upon risk assessment within child and adolescent mental health services: a rapid realist synthesis

Supporting evidence

This programme theory component is based on the NCISH report, which promotes an assessment process that goes beyond strict actuarial approaches. 38

Evidence base: three systematic reviews, one NICE guidance document, one feasibility study, one qualitative study, one narrative review, one survey, five commentaries, and one textbook.

Risk-assessment scales are commonly used in clinical practice to quantify the risk of suicide, with 85% of NHS mental health trusts using checklist-style approaches. 38 Currently, no standardised risk-screening tool is available for use within clinical practice in the UK. 39 Furthermore, risk-screening tools that exist possess questionable validity, reliability and acceptability (see Validity and Table 8).

In contrast, NICE guidance (CG133) recommends that risk assessment should take place within a comprehensive assessment of the patient’s needs. 14 A recent systematic review43 concludes that current evidence is not yet sufficient to recommend that structured diagnostic assessments should be universally adopted as an adjunct to clinical practice. However, the reviewers suggest that structured diagnostic assessments could be applied cautiously and mindfully pending further evaluation. A minority of users of the Davies’ structured interview for assessing adolescents in crisis expressed concern that ‘having a form to fill in’ hampers the development of rapport and a relationship between the young person and the professional.

Critics of actuarial approaches comment on the paucity of empirical evidence to support the ability of tools to predict accurately. 16,44–46 See Programme Theory 12. Many argue that tools are based on information about groups, which is of limited value in predicting the behaviours of an individual. 5,44,47 Within adult mental health care the literature consistently affirms that the focus of mental health organisations is now on risk management,48,49 quality assurance and patient safety. 49 Recent studies suggest that this may also be true for CAMHS. 50,51 Risk assessment in isolation from the development and implementation of clinical judgment frameworks becomes potentially ineffectual. Clinicians should not shelter behind the ‘fallacy’ of risk assessment, instead of acknowledging that assessment tools are likely to serve the organisation more than the patient. 52

A possible corollary to Programme Theory 1 is that development of simple assessment tools within a complete assessment process could result in higher rates of referral for risk of self-harm and suicide, thereby increasing utilisation of CAMHS services.

Supporting evidence

This programme theory component on building relationships is based on the NCISH report, which found that clinicians believed that an important focus of risk assessment involved building a rapport such that the assessment flowed smoothly. 38

Evidence base: one NICE guidance document, five qualitative studies, two surveys, one case study, and one commentary.

NICE guidance 133 states that ‘health and social care professionals working with people who self-harm should: aim to develop a trusting, supportive and engaging relationship with them’. Such a recommendation is further informed by qualitative research using interviews with nurses on wards of four psychiatric hospitals. 53 Professionals are concerned about how risk assessment may influence their relationship with service users. Often mental health nurses tend to emphasise risk avoidance to maintain safety. 5,54,55 Literature describing nurses’ perceptions of safety in acute mental health reports that nurses perceive their role as mainly risk management. 5 Most packages focus on assessment skills, risk screens and risk-factor tools but do not address tensions between divergent views of people in distress and professionals involved and how to build empathic partnerships5 in time- and resource-poor environments.

A further tension relates to working environments that privilege ‘task-based nursing over therapeutic care’ and those that create ‘conditions for open and genuine communication’. Task-based working environments, exemplified by a preoccupation with tick-box risk assessment, often prove detrimental to person-centred care. Furthermore, within a mental health service context, a focus on risk management ‘inherently erodes the formation of a therapeutic relationship, as patients who are viewed as risky are not trusted’. 56

In contrast, where ‘conditions for open and genuine communication’ exist staff members seek to focus on ‘developing an accurate and meaningful picture of patients’. 53 As a consequence staff members can enhance their capacity for compassionate and considerate contact and communication with patients experiencing suicidal ideation.

Compassionate care is particularly important – unlike their feelings for the self-harm population in general, staff typically hold positive attitudes towards self-harm specifically in adolescents and young children. 40,57 If done well in an unhurried, empathetic and non-judgmental manner, the interview can be therapeutic and encourage the patient to seek future help. By contrast, negative attitudes and a focus on the patient’s physical needs might result in the patient avoiding emergency services in the future. A healthcare professional should not give false reassurance, because patients may doubt that they are taking their situation seriously. 58 If possible, they should seek to obtain a corroborative history of the event from a third party. 58

Assessing young people requires engagement, empathy and a genuine curiosity about what has happened to bring the young person to a point of acute risk. Such an approach seeks to increase the chances of openness and honesty and a collaborative risk assessment. Otherwise, young people will keep risky thoughts and plans hidden, particularly if they think they will be judged or punished.

When presenting to their GP, young people feel that it is important that their GPs initiate the conversation about mental health, suicide and self-harm. 59 If a GP asks directly about such topics this may overcome some of the barriers to disclosure of suicidal thoughts, depressive symptoms or mental health problems more generally.

In the context of risk assessments for suicidal behaviour and/or self-harm, young people dislike labels such as ‘risk’ and ‘risk assessment’. 59 They perceive such labels to be potentially stigmatising and problematic. Young people may be especially vulnerable to labels that could increase stigma; language and terms related to suicidality or self-harm may be perceived as ‘pathologising’. Awareness of these attitudes may help in a shift away from professional-focused terms such as ‘at-risk’ and ‘risk assessment’, to patient-focused language such as ‘coping assessment’. 60 However, participants in one qualitative study disliked the term ‘assessment’, suggesting the inclusion of language relating to ‘well-being’. 59

Young people endorse the need for ‘comprehensive psychosocial-based assessments that prioritise collaboration and the therapeutic alliance, are holistic, acknowledge that risk is dynamic over time, and are needs-driven’. 59 Individualised, needs-based approaches to assessment are key for young people. 59

A collaborative dialogue facilitates empowerment and creates opportunities for young people to be involved in decision-making and to meet their growing needs for autonomy, agency and control. 59 Such a dialogue is concordant with principles of patient-centred care, shared decision-making and patient engagement. Furthermore, patient-centred care is fundamental to a biopsychosocial approach and recognises the pivotal role of the family. Young people may be particularly sensitive to power disparities and condescension. A friendly, non-judgemental attitude is critical; poor attitudes and body language and impersonal, overmedicalised approaches impede the therapeutic alliance and the disclosure of suicidal behaviour/self-harm. 59

Young people’s views of self-harm services have not been extensively studied. 61 A recent study has explored the views of young people in relation to the role of GPs. 59 GPs have been found, in one study, to be the most frequent healthcare practitioner source for urgent referrals of children and young people for self-harm, suicidal thoughts or following overdose. Families may prefer to access their GPs when worried about these issues. GPs can feel dependent on specialist support and feel the need for increased training in supporting children and young people with mental health issues. 62 Young people expect GPs to be skilled and knowledgeable in providing practical resources and support for presentations of suicidal behaviour and self-harm, including crisis support. 59 Assistance from the GP with accessing crisis resources or using a safety plan is viewed as highly beneficial. 59 GPs taking the time to demonstrate resources to the young person was another expression of care and connection to assist a positive relationship. 59 Young people may have little previous experience of how the healthcare system is structured, and therefore might require more ‘scaffolding’ than adults. 63

Young people are typically ambivalent when seeking help. They may isolate themselves, feeling that it is not safe, or that they are not ready to disclose their suicidal thoughts and feelings (e.g. as a consequence of feeling shame). In response, nurses describe how they try to enable patients to communicate in an open and genuine way. 53 By presenting themselves as accessible and approachable, reaching out to patients, and encouraging patients to approach them and talk to them nurses are able to work on creating an open and communicative environment. 53 Nurses highlight the need to develop a trusting relationship, respect the emotions of patients and reassure patients that they can disclose suicidal ideation. 53

All the above suggests that ‘… policy makers and hospital leaders should aim to create environments where [staff] can be involved in multifaceted and interpersonal approaches to suicide risk assessment’. 64 In such environments organisations could create relationships between children and young people and professionals that release preventive and therapeutic potential, rather than encouraging impersonal observations and ineffective checklist approaches. See Programme Theory 1.

Supporting evidence

This programme theory component is based on the NCISH report,38 which found that clinicians believed that an important element of risk assessment is the quality of the information gathered. The clinicians interviewed noted the importance of gathering a thorough history of previous incidents, and having an awareness of triggers for distress, for example significant anniversaries. They reported that a good risk history should include details of the incident and its consequences as well as the likelihood of the incident being repeated. However, some highlighted the difficulty of predicting suicide. 38

Evidence base: two quantitative studies, one multicentre study, three qualitative studies, three surveys and three commentaries.

Critics argue that tools tend to focus on historical (static) risk factors thus ignoring the dynamic or situational variables, which impact on the person. 5,8 The Functional Analysis of Care Environments–Child and Adolescent Risk-Assessment Suite (FACE–CARAS) suite of tools promotes use of schedules that enquire about both historical (static) and current (dynamic) risk factors. 65

Key to risk assessment is a collaborative dialogue, which encompasses the provision of adequate, detailed information across all aspects of a young person’s care, including treatment options and confidentiality. 59 Assessment tends to focus on risks people with mental health diagnoses pose, which marginalises consideration of other risks like living in inadequate accommodation. 66 It constructs individuals as risks who need interventions rather than identifying issues within particular communities, such as those with higher levels of poverty, substance abuse and unemployment. 48 It may also obscure risks that come from accessing mental health services, which potentially include loss of liberty, forced treatment or negative experiences. 67,68

Young people value the protection of their privacy, particularly for sensitive issues. 59 However, this should not be interpreted as a reason not to ask them about their thoughts of self-harm or suicide. Health professionals should also be aware that different types of self-harm may be viewed differently by children and young persons (CYP). For example, stigma associated with cutting may make a child or young person more secretive whereas attempted suicide frequently signals that the young person has reached a point where they are no longer able to cope at all. 61 Young people also express concerns regarding the privacy and confidentiality of their medical information relating to mental health and suicidal behaviour/self-harm. 59

Challenges exist in relation to incompleteness of information. A survey of outpatient and inpatient adolescents in the UK showed that 20% reported at least one episode of self-harm on the questionnaire that was not recorded in the clinical record. 69 The authors concluded that ‘using a combination of clinical interviews (with multiple informants), paper-and-pencil tools and comprehensive clinical records’ keeping afford the best chance of identifying adolescents who self-harm’. 69 A multicentre study of self-harm in England70 reported that psychosocial assessment occurred in only 57% of presentations in the study, even though the three centres (six hospitals) involved had well-established specialised self-harm services. The authors concluded that this ‘low rate of completion demonstrates the extent to which hospitals fall short of implementing the national guideline recommendation that all self-harm patients should receive a specialist assessment’. 70 They suggest that this low completion requires further investigation, particularly as ‘non-assessment may have several causes (e.g. self-discharge, patient refusal, unavailability of staff, emergency department policy). 70 They argue that this is particularly critical given what they claim as ‘accumulating evidence that psychosocial assessment is associated with reduction in risk of repetition of self-harm’ and the fact that ‘provision of appropriate psychiatric and social care is unlikely in the absence of an assessment’. 70

The FACE–CARAS tools are predicated on a stepped approach to completion – such that subsequent tools are only completed when indicated by the overall risk profile – but even within the context of research and evaluation completion of subscales was found to be unacceptably incomplete. 65

While advances in computerisation and clinical records have shifted the exact nature of this challenge the need for multiple and complementary approaches remains as pressing as ever. Specific challenges relate to conducting suicide risk assessment. Self-report measures of suicidality are limited by reporting biases (e.g. young people may conceal suicidality to avoid anticipated negative consequences) and high temporal variability (i.e. self-reported suicidal ideation may fluctuate from moment to moment). 71

When patients feel able to communicate in an open and genuine way, nurses are able to get to know patients, can assess suicidal ideation and also identify risk and protective factors. Strategies used to characterise the presence and severity of suicidal ideation, include listening to and observing patients, asking patients about the presence of suicidal thoughts and plans, and checking with colleagues. 53 Nurses must be alert to expressions that might be indicative of suicidal ideation (e.g. self-harm and social isolation). Nurses describe how they depend upon their intuitive senses, and that their own emotional responses, including ‘feeling anxious about the potential of a suicidal attempt’, provide cues to emerging suicidal ideation. Conversely, such emotional responses may also make nurses more likely to assess suicide risk as higher than it actually is. 53

Supporting evidence

This programme theory component is based on the NCISH report38 in which patients recommended that risk-assessment tools should incorporate a focus on suicidal thoughts, i.e. ‘to encourage staff to confidently tackle difficult questions’.

Evidence base: one meta-analysis, one quantitative study, one service improvement project, five qualitative studies, five surveys and one editorial comment.

Mental health nurses who are confident can make responsible decisions related to risk management. 5 Some nurses seem to have the interpersonal qualities and skills to move beyond checking and controlling suicide risk and instead make efforts to acknowledge and connect (with) the patient as a person, even during standardised assessments and observations. 53 These nurses adopt a focus that transcends a reductionist focus on static risk and protective factors and seems to open doors to a holistic picture of patients by being attentive to their needs and hopes and trying to understand the nature of their suicidal expressions. 20,72

One possible source of discomfort for staff members, particularly those who do not specialise in mental health, is the fear that asking patients about suicide might induce suicidal ideation. In general, nurses favour ‘daring to discuss’ suicidal ideation to support the patient’s communication. However, they also felt that they must not ‘force the conversation’. 53 Thirteen studies (2001–13) have examined whether asking about suicide induces suicidal ideation. 73 With samples including both adolescents and adults and both general and at-risk populations, none of the identified studies found a statistically significant increase in suicidal ideation in participants as a result of being asked about their suicidal thoughts. Findings suggest that acknowledging and talking about suicide with adolescent populations may in fact reduce, rather than increase suicidal ideation, with a suggestion that repeat questioning may benefit long-term mental health. 73 Studies in treatment-seeking populations suggest that asking people who are or have been suicidal about suicidality can lead to improvements in mental health. 73 Review findings suggest that recurring ethical concerns about enquiring about suicidality could be relaxed.

The fear that asking about suicide itself precipitates action (so-called iatrogenic risk) persists, especially among clinicians with a non-psychiatric background. A meta-analysis quantitatively synthesised 13 studies that explicitly evaluated the iatrogenic effects of assessing suicidality via prospective research designs. When pooled the overall effect of assessing suicidality did not demonstrate significant iatrogenic effects in terms of negative outcomes. A key strength of this study is that the review authors stratified studies according to the timing of their follow-up assessments, concluding that assessing suicidality did not result in any significant negative effects on immediate, short-term, or long-term follow-up assessments. The authors conclude that their findings support the appropriateness of universal screening for suicidality, and state that this should allay the fears of clinicians that assessing suicidality is harmful. 74

Clinicians’ anxieties may increase the reliance on undertaking an assessment based upon a checklist of phenomenological or epidemiologically valid items that provide few opportunities to account for individual differences that may provide a more accurate and richer suicide risk assessment. 75 Use of risk-assessment tools may provide false reassurance, assuaging the clinician burden and sense of dyscontrol, while giving the impression of effective working and so mediating corporate risk.

Losing a patient by suicide can impact on professional practices, including issues around objective clinical decision-making. It may lead to behaviours likened to learned helplessness, such as increased vigilance when dealing with future suicidal patients and avoidance of treating suicidal patients. 76,77 These in turn may lead to an ongoing reliance on the same systems for assessment and treatment. 75

One feature that might influence staff’s comfort and willingness to ask young people about suicidal ideation relates to whether young people themselves feel comfortable with such questioning. Increasing numbers of qualitative studies have found that, contrary to the beliefs of many, young people do not mind being asked about the presence or absence of suicidal thoughts. 78–81 Several tools utilise self-report approaches. For example, the developers of the Risk-Taking (RT) and Self-Harm (SH) Inventory for Adolescents (RTSHIA) point out how the quality of data produced by self-report measures is comparable to those obtained through clinical interviews. 82 They state that people may feel more comfortable admitting to sensitive thoughts and acts when they are asked to circle a response or write a brief explanation instead of providing a verbal report, which may be influenced by interpersonal reactions to interviewers. Reassurance of the confidentiality and anonymity of self-reports is also important for young people. Pragmatically, few alternatives to self-report data exist when requesting personal and sensitive information from young people.

Supporting evidence

Programme Theory 5 is based on the NCISH report,38 which found ‘little consistency in the length, content or use of risk tools, although there was greater consistency in some places than others’. Risk assessment also needs to be consistent across mental health services. 38

Evidence base: one systematic review, one mixed-methods study, one interrupted time series, one case series, one service improvement project and two surveys.

As articulated the programme theory relates to inconsistencies in the role and personal characteristics of the staff member making the contact and to inconsistencies resulting from contact with multiple, uncoordinated individuals. Patients who were critical of the assessment process felt that there was inconsistency between teams. 38 It is noteworthy that one of the strengths of the Wales Applied Risk Research Network (WARRN) initiative, as identified by clinicians, is the development of a consistent approach, within and between organisations. 83 Clinicians acknowledged that different agencies had created a common language and understanding that improved communication both across and between agencies. 83 These benefits have been similarly realised by a consistent two-step risk assessment and management process (Comp RA) within Northern Ireland. 38 Benefits can also extend to the development of standardised training and supervision procedures and processes, seen in the WARRN83 and the Collaborative Assessment and Management of Suicidality (CAMS) training programmes. 75

Programme Theory 5 is further supported by a mixed-methods study,84 which examined which risk-assessment tools were currently in use in the UK, and collected views from clinicians, service users and carers on the use of these tools. Findings showed little consistency in the use of these instruments. 84 Clinicians, patients and carers expressed both positive and negative views of the featured instruments. Findings attest to the need for assessment processes to be consistent across mental health services. Many professionals using the Davies’ structured assessment for adolescents in a crisis thought that it was good for a professional to have some structure and framework within which to operate so that ‘nothing would be missed’. Significantly, this view was not shared universally. Ongoing supervision is another provision to support consistency of approach. Care for self-harm within emergency departments appears to be particularly variable, with research showing it to be ineffective and delayed. 85

Areas where compliance needs to be improved include appropriate completion of the risk assessment. 86 A recent study extracted anonymised data from CAMHS at two time points. Data were compared with prevalence and population data and then a subsample was evaluated against NICE guidelines. Between time points there was a significant decrease in the number of cases that had a risk assessment completed appropriately and the number that had a full risk screen completed. 86 It is unlikely that this result was due to either a genuine reduction in the level of risk seen in CAMHS87 or that it represents a change in reporting practices. Even where a risk screen is completed somewhere in their notes, consistency needs to be improved to standardise risk monitoring and communication between services. For example, if a young person transitions to adult services having readily accessible information on risk is crucial.

Further variation relates to the experience of the clinician; experienced clinicians tend to use a positive risk-taking approach, whereas recently qualified clinicians do not feel as confident with suicidality cases unless they are routinely confronted with such cases (such as those working in a crisis team). 75

Supporting evidence

This programme theory component is based on the NCISH report,38 which found that healthcare professionals do not feel confident in being able to implement care plans, within which immediate risks can be mitigated, if they lack appropriate support and guidance to inform their assessment. 39

Evidence base: one systematic review, one RCT, three quantitative studies, one mixed-methods study, one qualitative study, one pre–post study, one service-improvement project, three surveys and one commentary.

Nurses with good confidence can make responsible decisions related to risk management. 5 Continuing education about the use of risk-assessment tools is needed to demonstrate that their use is compatible with therapy. 5 Staff need training if they are to use risk assessments in such a way that ensures their reliability. 88 A mixed-methods study in the UK reported little consistency in the use of instruments and highlighted a need for adequate training. 84 Nearly a third of clinicians surveyed in UK mental health services reported poor levels of training, highlighting practical issues in the use of tools and the poor quality of documented information. 84 Noticeably, training is a substantive component of both the Davies’ structured assessment for adolescents in a crisis and for the WARRN formulation-based approach83 – both indicating that familiarity with a structured process and how it integrates within clinical judgement should be considered more important than technical mastery of a tool or checklist. Over 2 days, the WARRN training modules cover basic clinical skills, such as how to conduct a clinical interview and what should be covered, techniques for asking difficult questions, how to formulate, and how to produce risk-management plans. The essential need for documentation and communication of presenting risks and the reasons underpinning these risks are highlighted. The value of co-production with the service user and family/carer is also covered. Standardised paperwork and forms to record the WARRN assessment and formulation are provided for use by clinicians following training. 83

Typically, healthcare professionals within emergency department environments have limited mental health training, and as such, feel ill equipped to assess and manage the associated risks apparent for children or adolescents presenting following an episode of self-harm or attempted suicide. 89 The limitations of these prediction methodologies likely impact on clinicians’ confidence when assessing suicide risk. Dealing with patients who self-harm and/or are suicidal is perhaps one of the most difficult challenges faced by clinicians. 75 One study estimated that 88% of mental health professionals have at least some level of fear relating to a patient dying by suicide, as well as discomfort around working with suicidal patients. 90 More than two-thirds of doctors practising emergency medicine believe that they are insufficiently trained at assessing those attempting self-harm. The limited training that health professionals receive relating to the assessment and management of suicidality may contribute to the burden felt by clinicians working in healthcare settings. Learned helplessness may result as suicide rates remain unaffected and predictive data have little impact on reversing this rate. The checklist-style structure of risk assessment within many NHS mental health services forms an ‘aide memoire’ of items characteristic of many suicide risk prediction tools.

Evidence highlights that training focusing specifically on the management of the suicidal drivers, or factors mediating the cognitions, emotions and behaviours augmenting suicidal risk, resulting in suicidal behaviours, can have a positive effect on clinicians’ confidence, clinical skills and implementation of evidence-based practices. 91,92 Greater awareness and accurate knowledge can de-stigmatise self-harm behaviour by staff enabling them to develop a greater understanding of contextual issues. 93 Additionally, education and attitude awareness may equip professionals with alternative explanations for self-harm behaviour that can help them to become more empathic and, subsequently, to alter their behaviours. 89,94

Training was a major component of a service-improvement initiative aimed at improving suicide prevention in North East Lincolnshire. 75 Three phases of training were delivered across the organisation: ‘suicide risk triage’ training, CAMS training, and CAMS concordance. All qualified staff were required to attend a mandatory 1-day training course entitled ‘risk triage training’ in groups of approximately 12 staff. Besides providing an overview of how the ‘suicide risk triage’ model was to be implemented within services, training collected data on factors that clinicians felt impacted on their confidence during the suicide risk assessment. The mandatory training also ensured all clinicians met a baseline level of ability and knowledge and was delivered to all new and newly qualified clinicians. Anecdotal feedback from the training highlighted the positive impact of a clear, structured approach to clinical risk decision-making to help clarify the most appropriate pathways to care for suicide risk presentations and the benefit of having support available for decision-making around challenging risk cases. The authors highlight evidence that CAMS training can significantly decrease clinician’s anxiety about working with suicidal risk and increase confidence, with results sustained at 3-month follow-up. 95 However, they acknowledge that the CAMS approach has yet to be evaluated in the UK.

Evidence from another study suggests that while training may help in ensuring staff can engage with the theoretical aspects of the situation they need additional provision for practical implementation. 96 Reflective peer review is suggested as one mechanism by which to help staff to reflect on their risk assessments, consider the knowledge and information that has informed their risk-management plans and discuss this with their peers in a supportive environment. 96 The authors claim that such a programme improved staff skills, confidence and documentation. 96

Similar findings are reported from a joint Australia–Switzerland initiative to investigate whether a training intervention increases GPs’ detection sensitivity for probable mental disorders in young people. 97 While improvements in detection were demonstrated these related only to more clearly detected cases and not to a more pragmatic clinical definition. The authors concluded that improving recognition of mental disorder among young people attending primary care is likely to require a multifaceted approach targeting young people and GPs. 97 Training is a necessary, but not sufficient condition for improved detection.

Supporting evidence

NICE guidelines specify that ‘Child and adolescent mental health service practitioners involved in the assessment and treatment of children and young people who have self-harmed should… have regular supervision’. 14

Evidence base: one clinical guideline, one service improvement project and one survey.

Successful suicide prevention requires that clinicians are confident when faced with suicide risk is a pertinent issue for suicide prevention. 75 There is conflicting evidence on whether clinicians focus on predicting the probability of suicide, despite little evidence supporting the utility of this approach. 75 A recent survey suggests that attitudes and behaviours towards the predictive ability of tools may differ between doctors and other health professionals. 38 However, the survey did not specify whether doctors were specialists or non-specialists or whether their experience was based on specialist training in mental health or rotations.

One NHS mental health provider implemented a service-wide, systems-level approach to suicidal risk (known as ‘suicide risk triage’),75 with supervision as a key component. This sought to address issues around clinicians’ confidence when assessing suicide risk, identified through training sessions, which highlighted the value of shared responsibility with senior supervising colleagues when considering more challenging suicide risk assessments. By addressing the concerns of all clinical staff, through a formal supervision hierarchy this system-level approach sought to ‘minimise confounders of objective, person-specific clinical risk decision’. 75 The supervision hierarchy was provided to support clinicians if they were unsure about the level of suicidal risk a service user presented with, the treatment plan they would develop for them, or if they felt that the risk was potentially life-threatening and therefore needed escalation for assessment and intervention. 75 Supervision arrangements included additional training for nominated clinicians within each team who were available to support/advise their colleagues when making difficult decisions around assessment and management of suicide risk. This support could be extended further up the hierarchy to trained clinicians and senior staff with extensive experience of managing clinical risk. 75 Clinicians affirmed the benefits of having a supervision structure in place, together with an organisation-wide approach for handling suicide risk cases.

Supporting evidence

This programme theory component is based on the NCISH report,38 which found that clinicians considered that closer contact with a patient’s family is the second most important risk-reduction factor in preventing suicide, after closer supervision of the patient. NICE guidance [NG 225] states that the professional should ‘discuss with the person and their families or carers (as appropriate), their current support network, any safety plan or coping strategies’. 14

Evidence base: one clinical guideline, one mixed-methods study, two qualitative studies, two service improvement projects and one survey.

One mixed-methods study from the UK reviewed risk-assessment tools currently in use and concluded that personalised management plans should be collaboratively developed with patients and their families and carers. 84 Engaging carers in discussions on risk has been shown to improve carer satisfaction. 98 Davies’ structured interview for assessing adolescents in crisis17 is one tool that includes structured interview/checklist assessment with ‘parenters’. The rationale cited is the need to guard against any omissions, for whatever reason, from the young person. Specifically, its developers point to how mood troughs are more critical than averages when conducting the risk assessment and that talking to significant others (e.g. parents) can serve to elicit such information.

More broadly, involvement of service users and their carers in mental healthcare planning is largely welcomed by mental health professionals. 99 However, tensions between user and carer involvement and professional accountability remain to be resolved. Conventional staff training programmes are commonly viewed as deficient, requiring that user involvement depends for its success on individual, relational skills. 99 Notwithstanding a generally favourable professional view of user involvement, challenges remain in relation to a lack of effective implementation support. 99

Supporting evidence

This programme theory component is based on the NCISH report,38 which identified a pressing need to improve access to and collaboration between primary care and mental healthcare services.

Evidence base: Three qualitative studies and one commentary.

Many UK studies describe issues that GPs have with mental healthcare services. Mental healthcare professionals are thought to tend to minimise GPs’ assessments of patients’ suicidal state. Adolescent contact with primary care presents an opportunity to conduct suicide screening and intervention. However, most primary care providers do not screen adolescents for suicide risk, perhaps because of suicide being a low base rate event. 100 Providers may feel that they lack formal psychiatric training or they may experience a general discomfort about screening adolescent patients for suicide risk. 100 Cumulatively, as many as 83% of adolescent suicide attempters are not identified as such by their primary care providers. 100

GPs report feeling stuck with patients, because they rarely meet the criteria for review and, therefore, remain in primary care. 101,102 A recent British study described how GPs feel professional isolation; being ‘lost in a referral maze’. 103 British GPs have also expressed the need for mental health staff based in GP practices. 103

Supporting evidence

This programme theory component is based on the NCISH report,38 which found that patients expressed a wish for ‘a personalised approach, not based on the completion of a checklist’. The team highlighted their previous research to suggest that risk is often individual and risk management should be personalised. 104,105

Evidence base: two qualitative studies, one mixed-method study, one pilot study, one service improvement project, one confidential inquiry, one survey and one commentary.

Young people dislike assessment approaches that are inflexible or binary; perceiving them as failing to capture nuance in their mental states and potentially impacting negatively on access to health care. 59 Instead, young people want to be treated in a holistic and individualised manner. 59 This finding reinforces recommendations that methods that categorise patients into ‘risk-level’ groups should not be used to determine treatment outcomes, as they can miss key opportunities for intervention. 18 In particular, methods that feel impersonal to young people, such as ‘tick-box’ or checklist-style approaches, are unwelcome. 59

A UK-based mixed-methods study, which examined risk-assessment tools in current use, highlighted the need for management plans that are personalised and collaboratively developed with patients and their families and carers. 84 These findings are substantiated by contemporary qualitative research exploring conversations of self-harm in the emergency department. 106 Patients identified two main types of approach:

• A therapeutic interaction made people feel their life mattered and instilled hope for the future.

• A formulaic assessment focusing on risk made people feel their life did not matter and hopeless about the future (see Table 4).

Even though patients interviewed were 18 years and older (eligibility was 16+ years) the data explores staff approaches rather than patient-specific factors so is likely to be transferable conceptually to the experience of children and adolescents presenting to the emergency department.

In the wider context of safeguarding it has been observed that health and social care professionals are in constant tension, accountable for promoting individual autonomy while seeking to predict accurately the level of risk resulting from subsequent action. 107 This tension is equally present within risk assessment for children and adolescents and reflects a wider literature, which contends that the focus of mental health organisations is now on risk management and quality assurance and patient safety. 108 Even where this imperative is not explicit it is revealed in how the purpose of tools is explained. For example, Davies’ structured assessment for adolescents in a crisis concludes by stating that ‘a structured interview or checklist of questions offers a fail-safe for clinicians to make sure that all important factors are considered when making an assessment’. 17

Supporting evidence

This programme theory component is based on the NCISH report, which stated that risk assessment should form ‘one part of a whole-system approach that should aim to strengthen the standards of care for everyone, ensuring that supervision, delegation and onward referral are all managed safely’. 38

Evidence base: one systematic review, one qualitative study, one survey and two commentaries.

The need for a whole-system approach is noted by the most recent systematic review considered within this realist synthesis. 109 It noted the policy direction of numerous best practice and policy guidelines for the assessment of risk from the UK, USA and Australia among others, all of which identified the need for ‘a whole system, multiagency, and collaborative approach’. 109 The same systematic review drew attention to ‘a clear lack of specificity as to how to implement the recommendations in practice’. 109 Furthermore, the review pointed out how ‘no single model of risk assessment was discussed in more than one document’,109 substantiating the conclusions of variability and fragmentation. Research suggests that paternalistic professional attitudes, homogenisation of service users, and organisational structures prevent the cultural change required to shift to a strengths-based approach to risk. 5,49,107,110

As a complementary activity the review team identified three ‘counter programme theories’, which relate to how the risk-assessment process might result in unintended consequences.

Supporting evidence

This programme theory component is based on the NCISH report,38 which reported that risk assessment has traditionally focused on prediction; patients being categorised into low, medium or high risk of a particular outcome.

Evidence base: two clinical guidelines, three systematic reviews – one with meta-analysis, one narrative review, two cohort studies, one observational study, one mixed-methods study, two surveys and four commentaries.

NICE guidelines state that risk-assessment tools and scales should not be used to predict future suicide or repetition of self-harm (risk screening), or to determine who should or should not be offered treatment. 14 The NCISH report38 found that scores on checklists also determined management decisions; contrary to national guidelines for self-harm assessment.

The NCISH report38 highlights research that suggests that categorising risk in such a way is unhelpful in guiding the treatment and management of a patient,111,112 and has poor predictive value. 16,113–116 It is supported by a mixed-methods study in the UK, which collected views from clinicians, service-users and carers on the use of risk-assessment tools. 84 Graney et al. highlight how most patients who died by suicide in the UK had been assessed as low risk in their last contact with mental health services. 84 They concluded that, in line with national guidance, risk assessment should not be seen as a way to predict future behaviour and risk screening should not be used as a means of allocating treatment. 84 The NICE guidelines suggest risk assessments might be used as prompts or measures of change. 18 Evidence suggests that risk-screening tools are no more accurate at predicting risk than expert specialist mental health professional clinical judgement in non-acute psychiatric outpatients. 114 A later review suggested the pooled positive predictive value for suicide was 5%: for every 100 people rated at high risk, five would go on to die by suicide. 117 More importantly, risk scales would miss suicide deaths in the large ‘low-risk’ group. 117

Most tools identified in the NCISH survey38 encouraged staff to make predictions of future behaviours and stratify risk, for example, into high, medium, and low or numeric risk categories. Overall, 80 (94%) tools used risk categorisation to inform care. In mental health services risk assessment has traditionally focused on prediction (risk screening). Around a third of nurses (n = 15, 32%) and managers (n = 11, 38%), but none of the doctors, thought tools had predictive value, compared to around two-thirds of psychologists (n = 20, 70%). 38

Notwithstanding acknowledged risk factors for such harmful acts as suicide and violence, no evidence has substantiated that identifying and responding to risk factors is useful in predicting, preventing or reducing risk of harm. 20 Even where risk assessment is believed to be useful, incorrect interpretation of individual need for care118 can lead to restrictive mental health practices, such as involuntary hospitalisation, restraint, sedation and seclusion. 20,119 Unintended consequences of such practices themselves present competing risks, placing both patients and staff at risk of harm. 20,120

In summary, then, no widely accepted tools exist for clinically assessing a patient’s risk of subsequent self-harm or suicide. 121 Specifically, within a child and adolescent population context, many promising measures for use in child and adolescent populations have insufficient psychometric data, and require further research. 121

Supporting evidence

This programme theory component is based on the NCISH report,38 which found that, contrary to national guidance, scores on locally used tools determined management decisions. Indeed, one of the clinical messages of the report is that ‘Risk is not a number, and risk assessment is not a checklist. (Risk screening) Tools if they are used …should be considered part of a wider assessment process. Treatment decisions should not be determined by a score’. 38

Evidence base: one systematic review, three qualitative studies, one service improvement project, two surveys and six commentaries.

A recent systematic review surveyed the available tools and concluded that limitations in the use of risk-screening tools in isolation as a predictor needs to be recognised. 109 The review concluded that no one risk-screening scale was supported by sufficient evidence to sustain its use in clinical practice. The review authors argued that this lack of empirical evidence should be used to engineer a radical shift in the contemporary discourse in the patient safety literature on risk assessment. Furthermore, they claim that the focus on risk-screening tools may deter the development of sound clinical judgement frameworks. Clinical judgement is considered essential for the use of Davies’ structured interview for assessing adolescents in crisis. 17 Its originators reason that sometimes a person only ticks a few boxes when they are at significant risk of harming themselves. Weighting within the items of a tool may mean that these important indicators in isolation may fail to trigger a clinical threshold. Furthermore, the development team reason that ‘mood troughs’, not averages, are more critical within the risk assessment. 17

Numerous writers have discussed the advantages and disadvantages of using risk assessment tools to assess risk. 8,48 Survey research suggests that nurses favour ‘interpretative’ approaches to assessment, relying on their own ‘instinct’ to guide assessment. In contrast, other studies suggest that most respondents believe that risk-assessment tools facilitate professional decision-making. 5 Critics point out that despite claims that risk-assessment tools help to manage risk, there is little evidence to support such assertions. 48 For instance, 87% of respondents in a survey of 1937 psychiatrists endorsed the view that ‘tools provide a false sense of security, as there is little direct evidence that tools help to reduce adverse events’. 19

The literature available suggests that intuition or the unstructured approach continues to form a key part of how nurses determine decisions about risk,122,123 with some studies suggesting that nurses see risk assessment as the doctor’s responsibility,122 and try to offset clinical responsibility when practising risk assessment and management by referring decisions to a psychiatrist or the team. 124 In relation to risk-assessment tools, only one study was located that reported nurses using validated tools or derivatives to guide their practice;66 while some of the community mental health nurses (CMHNs) in this study reported using tools they still favoured clinical judgement and ‘interpretative’ approaches. Nurses in other studies also report some ambivalence towards using tools, viewing them as a technology of psychiatry designed to erode clinical expertise or as bureaucratic instruments without value or purpose. 122,125 Conversely, others propose that they facilitate discussion between practitioners about risk and enhance care documentation. 122 They are also viewed as providing a measure of legal protection from liability and an important way of documenting and justifying decisions. 123 However, nurses also view risk-assessment approaches as little more than strategies to protect organisations, should an adverse event occur,124 contributing to defensive antitherapeutic practices. 126

Supporting evidence

This programme theory component is based on the NCISH report,38 which concluded that ‘there is little place for locally developed tools’. Approximately two-thirds of NHS mental health organisations use locally devised adaptations that lack formal psychometric validation. 115

Evidence base: one systematic review, three qualitative studies, two quantitative studies, one mixed-methods study, one observational study, one cohort study, one service-improvement project, two surveys and two commentaries.

Recent research confirms the limited clinical utility for predicting suicide and self-harm using risk-screening scales. 113,114,116 Furthermore, the use of such scales may result in unnecessarily restrictive treatment options for those categorised as ‘high-risk’. 127 Evidence reviewing the predictive value of widely used risk-screening scales in the UK has highlighted the low specificity of such scales for suicide and self-harm, which may result in individuals remaining within mental health services for longer than necessary. 113 In such cases, where staff inappropriately identify suicide risk, targeted treatment to assist suicidality may be superseded by restrictive care planning, such as compulsory detainment and hospitalisation. 127

Individuals with suicidality present with needs that are not exclusively mental-health-based, including societal, community, relationship and individual risk factors. 128 Assuming that suicidality is the result of a mental health diagnosis may place an unnecessary burden on mental health professionals to prevent suicide, as well as increased blame if an individual who does not seek help completes suicide. 129 Previous research estimates that, for those individuals who do have contact with healthcare services, only between 3% and 22% of individuals had reported suicidal intent at their final appointment with a health-care professional before their suicide,130–132 suggesting suicide risk identification is more complex than a simple dyadic relationship between suicide expression and psychiatric disorder. Unsurprisingly, UK suicide rates remain high, given the limited utility of suicide risk prediction methodologies that remain routine practice across mental health providers.

Children and Young People - Mental Health Safety Assessment Tool

Contemporaneously (2018) with a scoping review of assessment tools of immediate risk of self-harm and suicide in children and young people by Carter et al.,39 the authors published a protocol for the development and psychometric evaluation of such an assessment tool; the Children and Young People - Mental Health Safety Assessment Tool (CYP-MH SAT). 147 The authors claimed this as the first UK-based study to develop an assessment tool to ascertain immediate risk of suicide and self-harm in children and young people presenting to acute paediatric hospital settings in mental health crisis. As a strength for the UK context, the protocol was tailored towards an English-speaking population, while recognising that further national and international testing and adaptations are required for generalisability. A corresponding weakness is its untried status with regard to ethnic minority populations within the UK population. The protocol used an opportunistic sample of self-selected experts to inform development of the assessment tool, recognising that such experts might be more motivated to take part and/or exhibit allegiances that might lead them to respond in a particular way.

The resulting instrument underwent rebranding as the CYP-MH SAPhE instrument when it appeared in 2021. 50 The published study sought to psychometrically assess the CYP-MH SAPhE instrument for identification of immediate risk of self-harm in children and adolescents, aged 10–19 years, in acute paediatric wards or emergency departments. Through the scoping review and subsequent collaboration with expert academics and clinicians, an instrument was developed to assess immediate risk of suicide and/or self-harm in children and adolescents in mental health crisis to acute paediatric hospital settings and emergency departments. Testing of the instrument across three acute hospital sites (Paediatric Emergency Departments and Acute Paediatric Wards) within the UK resulted in an eight-item instrument, weighted within two constructs (self-harm and suicidality). The authors justify development of separate self-harm and suicide facets based on extant literature that defines them as separate constructs. The authors claim that the strength of CYP-MH SAPhE lies in ‘its co-development by those in clinical practice’ for use within a clinical practice setting. 50

Despite recognised limitations (see Validity), the authors claim that the CYP-MH SAPhE instrument is a rapid and sensitive instrument to identify immediate risk of self-harm and suicidality in children and adolescents aged 10–19 years presenting to acute paediatric care. The authors conclude that the CYP-MH SAPhE tool has ‘potential utility as a screener by the paediatric health professional in the inpatient ward or emergency department as part of a holistic assessment’. Key to the authors’ claim is the phrase ‘as part of a holistic assessment’ – the role of stand-alone risk screening tools for risk prediction is no longer championed. They recommend that the CYP-MH SAPhE instrument requires further evaluation ‘to confirm its suitability and effectiveness in clinical practice’. 50

Davies’ structured interview for assessing adolescents in crisis

Having previously produced the Describe the risk; identify the options; choose your preferred option(s); explain your choice; share your thinking (DICES) System for Risk Assessment in Mental Health and Risk Management in Mental Health, the team at the Association of Psychological Therapies decided to extend their work with a tool for children. The team devised a structured interview/checklist assessment in three parts for parenters [‘parenter’ is the term used for the person(s) doing the parenting], the young person themselves and the clinician. The questions address both fixed factors (e.g. age, gender etc.) and fluid factors (e.g. current level of hopelessness at any specific point in time). Clinicians can then share information with colleagues and make an informed clinical judgement, recognising that a young person may only tick a few boxes even when they are at significant risk of harming themselves. Once significant risks are identified then a risk-management plan is put into place.

At the time of information (July 2013) only about 30 people had received the structured interviews. Much of the claim for its utility is indirectly attributed from the DICES Series of checklists to which more than 6000 people had subscribed to use. Although most professionals welcomed a tool to ensure they ‘covered all the bases’, a minority thought that ‘having a form to fill in’ hampered the development of rapport and a relationship between the young person and the professional. Many (though not all) agreed that ‘it was good that the professional had some framework and structure to operate within’, and that ‘nothing would be missed’.

Clinical judgement is vital when assessing suicide risk. Fluid factors, such as hopelessness, are not easily weighted in statistical models. Mood troughs, not averages, are more critical to an assessment; such information is only elicited by talking to the person and their ‘significant others’ (usually parents) and taking into account fluid factors, such as hopelessness and how they say they feel. The team claim that a structured interview or checklist of questions offers a fail-safe for clinicians to make sure that all important factors are considered when making an assessment.

Functional Analysis of Care Environments (FACE-CARAS) suite of tools

The FACE-CARAS toolkit has been developed to support practitioners in CAMHS in performing a structured risk assessment. 150 It covers multiple risk domains including violence, suicide, self-harm, experienced abuse and exploitation. The FACE-CARAS involves comprehensive risk assessment, including a risk scale – to produce a FACE Risk Profile and a clinical management plan (paper/electronic). Schedules enquire about both historical (static) and current (dynamic) risk factors.

Functional Analysis of Care Environments–Child and Adolescent Risk-Assessment Suite comprises a suite of tools – the practitioner completes an overall risk profile before selecting other tools for a more in-depth assessment where indicated. The three-step process comprises completing a Young Person’s Risk Profile as a ‘screening’ step, then one or more Focused Schedules as indicated at the screening stage, and then the ‘Formulation and Management Plan’ section of the Young Person’s Risk Profile as appropriate to the Focused Schedule(s). Items are either coded as absent or as representing a perceived low, moderate or high level of risk in that domain. The resulting information is then used to develop a risk formulation and management plan, as well as to assign global scores to each of a number of risk domains.

The tool was designed for a youth mental health setting; the mean age of original sample was 15.94 (range 12.23–18.71) with 36 males/69 participants. 157 Predictive ability was tested with a sample of 123 young people with clinician-completed FACE-CARAS ratings. These were examined in a retrospective file review to extract data on a relevant list of adverse outcomes at 3 and at 6 months following the assessment. 65 Although this was not a prospective longitudinal study, researchers were blind to the clinicians’ ratings, allowing valid testing of predictive power. The FACE-CARAS profile score was considered a good potential predictor of risks of self-harm, suicidal behaviours, serious self-neglect, abuse or exploitation by others, and violence to others at both 3 and 6 months. It was weakly ‘predictive’ of accidental self-harm and no better than chance at signalling physical ill health.

Evaluation indicated the usefulness of the ‘profile summary’ section of the tool as likely to generate clinically useful risk predictions, notwithstanding that guidance recommends that tools for risk assessment should not be used for risk screening in a predictive way. 14 In practice, clinicians often did not complete the subscales – the authors therefore recommend further work. 157

Risk-Taking (RT) and Self-Harm (SH) Inventory for Adolescents82

Its originators claim that the RTSHIA, a self-report measure designed to assess adolescent RT and SH in community and clinical settings, offers an improvement over existing measures by providing information about the full spectrum of potentially self-destructive behaviours alongside other significant information. As a self-report measure, they claim the benefits of standardised administration, wording and scoring and faster, more economical administration and scoring. Furthermore, they state that the quality of data produced by self-report measures compares to that from clinical interviews. People may be more comfortable admitting to sensitive thoughts and acts when asked to circle a response or write a brief explanation instead of providing a verbal report. Moreover, assuring participants of the confidentiality and anonymity of self-reports seems to be easier. In any case, few alternatives to self-report data exist when requesting personal and sensitive information from young people.

Furthermore, the RTSHIA was developed for, and validated in a clinical population of, adolescents because ‘behaviours defined by adults as risky or self-destructive do not have the same function in adolescents’. The study included a large and highly diverse sample of participants drawn from a wide range of age groups.

Another claimed advantage is that the RTSHIA assesses the frequency of self-harm behaviour, as opposed to simply recoding its presence/absence.

Several limitations must be acknowledged. First, the RTSHIA is primarily a self-report measure, and therefore dependent upon respondents’ comprehension of items, concentration ability and openness. Despite reassurance with regard to anonymity and confidentiality, participants may hesitate to give personal information to authority figures. In addition, younger adolescents, may feel that certain items are not applicable to them. Second, the wording of the questions does not discriminate between current and past history behaviours. Finally, concern has been expressed over whether the two scales are sufficiently comprehensive. Although both scales have been shown to work, have good psychometric properties and appear reliable and valid, results are preliminary and need replication with different samples.

In summary, the RTSHIA responds to a need to supplement in-depth interview-based instruments and captures wider presentations of self-harm. It supports the need to rely on multiple assessment methods. The authors acknowledge the likely added value of obtaining interview data beyond self-report questionnaires. They suggest that the RTSHIA can be used as a primary screening measure to be supplemented by interviews or focused measures, especially in clinical settings, where in-depth information is required. For the present, the RTSHIA appears to offer potential for use as a multifocused screening tool for identifying diverse problem behaviours/thoughts in adolescence and as a tool for assessing young people who self-harm. Although it is premature to draw conclusions about the utility of the scale, the RTSHIA currently represents a psychometrically sound, comprehensive tool with the potential for further empirical investigation.

Wales Applied Risk Research Network

WARRN is a formulation-based technique for the assessment and management of serious risk for users of mental health services, adopted across Wales. The developers of WARRN recognised that structured professional judgments were impractical to use in many NHS settings due to time constraints, the need for training on each instrument, and the multiplicity of possible risks faced by any service user. The development team, therefore, aimed to develop a ‘formulation-based’ approach to risk assessment to equip clinical staff with skills for implementation.

WARRN was previously in use in adult mental health services across Wales. The ‘youth’ version was modified in consultation with senior CAMHS staff from across Wales; the only changes required were to take a developmental framework to the assessment and to use age appropriate training vignettes. Training was implemented via a training the trainer cascade programme.

A service evaluation was conducted to evaluate WARRN and its impact across CAMHS in Wales. An online survey was disseminated to 88 NHS clinicians in CAMHS to evaluate their perceptions of the use and effectiveness of WARRN. Clinicians reported increased clinical skills, increased confidence in their assessment and management of risk and in safety planning, the increased safety of service users and the general public, and a belief that WARRN had saved lives. Qualitative data showed that clinicians thought a common risk-evaluation instrument across Wales and different agencies had created a common language and understanding that improved communication both across and between agencies. WARRN appears well accepted in CAMHS services, exerting positive effects on service-user well-being and safeguarding with potential implementation in other services. However, this favourable view of the WARRN tool is based mainly on internal evaluations and remains to be replicated by independent evaluation teams.

Table 5 presents summary information on the risk-assessment scales and tools included in the review. Data for these tables were extracted from included primary studies and from the seminal scoping review (2019). 39 This scoping review from Carter et al. (2019) originally reviewed 22 different tools. Our review team has not referred to the original papers reporting on the development of these tools, some of which date to the 1960s, or to the extensive literature on validation in different populations and settings. Our approach reflects a focus on use of the tools in clinical practice together with the resource constraints of this review. References to other studies are cited in the papers included in the tables.

The scoping review published in 2019 identified 26 risk-assessment tools reported in 22 full-text articles. 39 However, a high percentage (59%) of included studies were developed over 20 years ago. Concerns about psychometric properties are typically raised in connection with older tools and instruments.

Our analysis extended the comprehensive list of assessment tools included in the scoping review;39 adding two recent tools82,147 and expanding beyond formal tools to include overall approaches. The original scoping review39 started with 22 tools. This review added ten tools to those tools used for self-harm and suicide in a general population. We excluded tools previously included in the scoping review39 to reflect only tools used in a UK context and to capture the primary focus on suicide. As reported in the scoping review, tools varied in length, response and scoring format, age ranges and degree of psychometric testing. 39 Most assessments were tested across broad age ranges, and so lack sensitivity to the age groups of particular interest to this review. The relative lack of tools for children, as opposed to adolescents, is noticeable although this imbalance does seem to follow the self-harm and suicide age trajectory. Some tools, such as the SIQ and the SIQ-JR have undergone age-based revisions/adaptations. 39

The scoping review concluded that many tools were subject to limited psychometric testing, and no single tool was valid or reliable for use with children presenting in mental health crisis to non-mental health settings. 39 It recommended development of a ‘clinically appropriate, valid and reliable tool that assesses immediate risk of self-harm and suicide in paediatric settings’. 39

Tool development continues to see different rationales in terms of whether to focus only on suicide risk or whether to incorporate risk items relating to self-harm. No measure assessed risk of self-harm in isolation. 39 As with much psychological tool development most assessment tools were tested only in the USA and primarily with inpatients. Where studies report psychometric testing in UK populations this is indicated in the accompanying tables. 39 UK guidelines remain unable to promote the use of any one assessment tool to safely manage immediate risk of self-harm or suicide to inform clinical decisions in acute paediatric settings. 192 The accompanying analysis indicates that the ongoing preoccupation to identify or develop such a tool has proved something of a distraction when attention should focus on a holistic biopsychosocial assessment conducted within a whole-system approach to assessment.