Understanding practitioners’ and young people’s views of a risk calculator for future psychopathology and poor functioning in young people victimised during childhood

Abstract Background Although children who are exposed to victimisation (including abuse, neglect, domestic violence and bullying) have an increased risk of later psychopathology and functional impairment, not all go on to develop these outcomes. Risk calculators that generate individualised probabilities of a victimised child developing future psychopathology and poor functioning have the potential to help practitioners identify the most vulnerable children and efficiently target preventive interventions. Aim This study explored the views of young people and practitioners regarding the acceptability and feasibility of potentially using a risk calculator to predict victimised children’s individual risk of poor outcomes. Methods Young people (n = 6) with lived experience of childhood victimisation took part in two focus groups. Health and social care practitioners (n = 13) were interviewed individually. Focus groups and interviews were audio-recorded, transcribed and thematically analysed. Results Ten themes were identified, organised according to the three main topics of discussion: (i) identifying risk (risk factors, current practice, accuracy, implementation, response); (ii) protective factors and prevention (individual, environment, preventative intervention); and (iii) communication of research (stakeholders, methods). Conclusion Risk calculators have the potential to enhance health and social care practice in the United Kingdom, but we highlight key factors that require consideration for successful implementation.


Introduction
Childhood victimisation, including abuse, neglect, exposure to domestic violence, and bullying by peers, is a pervasive and serious public health concern (Radford et al., 2013). Longitudinal studies of the consequences of victimisation suggest that exposed children have poor long-term outcomes. For example, childhood victimisation has been associated with externalising psychopathology including attention-deficit/hyperactivity disorder, conduct disorder and substance use disorders (e.g. Braga et al., 2017;Capusan et al., 2016;Lansford et al., 2010), internalising psychopathology such as depression and anxiety (Li et al., 2016;Nanni et al., 2012;Takizawa et al., 2014) and psychotic symptoms (Arseneault et al., 2011;Fisher et al., 2013). Furthermore, victimised children are more likely than their non-victimised peers to have lower educational attainment (Currie & Widom, 2010), be unemployed or "Not in Education, Training or Employment" (NEET; Brimblecombe et al., 2018;Jaffee et al., 2018), engage in criminal offending (Malvaso et al., 2018), be teenage parents (Herrenkohl et al., 1998), and experience further victimisation (Ports et al., 2016) and loneliness (Matthews et al., 2017) in adulthood. Such adverse outcomes may have negative and long-lasting implications for the individual's health, wellbeing and lifeopportunities as well as entailing significant costs for communities and public services.
Although victimisation is a key risk factor for later psychopathology and poor functioning, there is significant heterogeneity of outcomes among victimised children. Indeed, some individuals demonstrate very positive outcomes despite their negative experiences (Cicchetti, 2013;Rutter, 2013). Research seeking to understand why some victimised children are vulnerable to poor outcomes whereas others are resilient has identified a wide range of moderating factors at the level of the individual (e.g. intelligence, personality), family (e.g. socio-economic status, parental warmth) and community (e.g. peer relationships, social support) (Afifi & MacMillan, 2011;Crush et al., 2018;Meng et al., 2018). This knowledge is important to identify potentially modifiable factors to target to reduce the detrimental impact of victimisation and to help identify children who may be at greater risk and in need of support.
Crucially, these studies inform us about the factors that increase or decrease victimised children's risk of poor outcomes on average. However, it cannot be assumed that such knowledge can be directly applied to accurately forecast outcomes for individual victimised children (Danese, 2020). As such, these findings are of limited practical use to health and social care practitioners who assess individual children and therefore seek insight into that particular child's level of future vulnerability or resilience. Our recent work to translate this current knowledge into a risk calculator screening tool that makes individual-level predictions is therefore notable (Latham et al., 2019;Meehan et al., 2020). Although this tool needs to be validated in external samples before being implemented, our initial evidence demonstrates its potential to support practitioners to accurately identify which victimised children are at high risk of developing poor outcomes at the transition to adulthood. In turn, this could promote a more cost-effective allocation of resources by enabling practitioners to target preventative interventions to those children who are most in need and, therefore, most likely to benefit.
However, the implementation and use of a risk calculator by practitioners to predict adverse outcomes among children exposed to victimisation may not be straight-forward. For example, there is a need to consider who should administer the tool and in what setting. Furthermore, it may raise issues regarding the implications of a victimised child being identified as "high risk" which will need to be carefully thought through to ensure the tool is implemented sensitively. Therefore, it is vital to investigate what the key considerations are from those practitioners who might use the risk calculator (e.g. social workers, psychiatrists) as well as those with whom it would be used (e.g. children exposed to victimisation).
Thus, this qualitative study aimed to explore the views of young people with lived experience of childhood victimisation or mental health problems, and practitioners from health and social care about the potential use of a screening tool to identify those victimised children who are at highest risk of developing psychopathology and poor functional outcomes. Their opinions were gathered via focus groups (with young people) and individual interviews (with practitioners). The main aspects focused on were how to best integrate such a screening tool into health and social care services in the United Kingdom (UK), the potential benefits and issues that may arise when implementing and undertaking the screening, and how risk prediction would be best communicated to children and their families. We also explored with the young people how our risk calculator research could be communicated effectively to children, their families, practitioners and the wider public. The findings are intended to provide valuable information to inform future studies aimed at implementing and evaluating the use of such a risk calculator in practice.

Participants
Young people were recruited to take part in two focus groups. For inclusion, they had to be aged 18-25 years, reside in the UK, and self-identify as either having mental health problems and/or childhood exposure to abuse, neglect, domestic violence or bullying. Recruitment was undertaken in collaboration with the McPin Foundationa London-based mental health charitywho shared advertisements via Twitter and via email with their network of young people.
Practitioners aged over 18 years who resided in the UK and worked in health or social care services supporting children or young people who have been exposed to victimisation were also recruited. Advertisements were circulated via email to local Child and Adolescent Mental Health Services (CAMHS) and National Society for Prevention of Cruelty to Children (NSPCC) Service Centres, shared through Twitter and via the research recruitment websites of King's College London and the South London and Maudsley NHS Foundation Trust.

Procedure
Full ethical approval for the study was granted by the Psychiatry, Nursing, and Midwifery Research Ethics Subcommittee at King's College London (ref.: HR-18/19-10547) and all participants provided informed written consent. Young people were invited to attend two semi-structured focus groups held at the McPin Foundation offices in London. Participants completed a brief demographic questionnaire at the beginning of the sessions. The focus groups were co-facilitated by a member of the McPin Foundation (RKT) and a King's College London researcher (RML). Each group lasted no longer than three hours, was audio-recorded and then transcribed. Young people received a £90 voucher for their participation in a focus group and were reimbursed for their travel expenses.
Initially, we invited practitioners to take part in a focus group. However, due to difficulties identifying a mutually convenient time that a sufficient number of practitioners could attend, we instead conducted semi-structured individual interviews with each participant to ascertain their views. These one-to-one interviews typically lasted for approximately one hour and were audio-recorded and then transcribed. Practitioners were offered a £10 voucher in return for their participation.

Materials
Semi-structured topic guides (see Supplementary Materials) were developed to guide the focus group discussions and interviews. These outlined open-ended questions related to identifying risk, protective factors and prevention, andfor the young people's focus groupscommunication of the risk calculator research. Example questions included: "What factors in a child's life do you think might increase their risk of having mental health problems and poor functioning following exposure to victimisation?", "How receptive would you be to implementing a screening tool to identify victimised children at high risk of poor mental health and functioning in your role?"; and "Who do you think would be interested in hearing about the screening tool?". Questions for the young people were tailored to ensure the language was accessible.
We provided the young people with an outline agenda for the focus group session, so they were aware of what topics to expect and scheduled breaks. To facilitate their discussions, we included some activities in the focus groups. For example, participants wrote their ideas on post-it notes which were then used to create visual mind maps. We also used cards with different emotions written on them (e.g. "hopeful", "embarrassed", "worried") to help them consider how a victimised child may feel if they were identified as being at high risk of having poor outcomes. Photographs were taken of the work produced through these activities to accompany the focus group audio recordings.

Analysis
Transcripts and photographs were coded and analysed using the software package NVivo 12. Thematic analysis was used with the aim of understanding young peoples' and practitioners' views regarding the acceptability and feasibility of using a risk calculator to identify those victimised children who are most vulnerable to psychopathology and poor functioning. The six phases of thematic analysis (Braun & Clarke, 2006) were followed to guide this analytic process. Transcripts were read and re-read to establish familiarity, initial codes were then generated, and broader themes identified. These themes were then refined and defined. One researcher (RML) independently coded all transcripts.
A second researcher (HLF) independently coded one third of the transcripts and any differences were resolved by discussion and repeated inspection of the data until a consensus was reached.

Sample characteristics
Two focus groups with young people were held. The first group comprised four participants and the second group included an additional two participants, all of whom had lived experience of mental health problems and childhood victimisation. Participants were predominantly female and aged 20-21 years old (see Table 1).
Thirteen individual interviews were conducted with practitioners. The majority of whom were female, white British, and most commonly aged 30-39 years old. Approximately, half of the practitioners worked in health care and half were social workers (see Table 2).

Thematic analysis
From the two focus groups and 13 interviews, 10 main themes were identified. These themes were organised according to the three topics of discussion: (i) identifying risk (risk factors, current practice, accuracy, implementation, response); (ii) protective factors and prevention (individual, environment, preventative intervention); and (iii) communication of research (stakeholders, methods). Several of these themes also included subthemes (see Figure 1 for an overview of the thematic structure).
Full details and explanations of the themes and subthemes are provided in Table 3 with illustrative quotes from the young people and practitioners. We found that themes were largely similar across the two stakeholder groups.

Discussion
This study explored the views of young people and practitioners regarding the acceptability and feasibility of implementing a risk calculator to identify children who are most at risk of developing psychopathology and poor functioning following victimisation. Thematic analysis revealed significant commonality between the considerations of these two key stakeholder groups. We discuss below the implications of our findings for future work aimed at implementing the screening tool.
Our study suggests that the risk calculator would be a relevant and novel addition to practitioners' toolkits when working with victimised children (theme 1.2. "current practice" and subtheme 1.4.2. "setting"). Even though children's current presentation was often the primary focus of practitioners' assessments, many also considered what their future needs may be. We found that practitioners typically relied on their professional experience and knowledge to do this. Some people did also identify toolsincluding the Adverse Childhood Experiences (ACEs) questionnaire (Felitti et al., 1998) and the Risk and Resilience Matrix (Daniel & Wassell, 2002)that they believed could be used to inform their assessment. However, the suitability of these tools for forecasting individual outcomes is limited. For example, although research on ACEs is important for highlighting the association between childhood adversity and poor outcomes, not everyone follows the average pattern. That is, many individuals who score high on the ACE checklist do not show poor outcomes therefore it cannot be assumed that screening for ACEs will give accurate information of individual risk prediction within groups of victimised children (Danese, 2020). Similarly, although the Risk and Resilience Matrix encourages practitioners to consider the presence of protective factors alongside the child's experience of adversity, it does not translate this information into a prediction of vulnerability or resilience related to a specific future outcome. Using the matrix to forecast a victimised child's future mental health and functioning thus requires subjective interpretation by the practitioner. Our findings therefore suggest that an individualised risk calculator for future psychopathology and poor functioning fills an important gap that could help these practitioners to assess and support the future wellbeing of children who have experienced victimisation.
Our study also showed that young people and practitioners were positive about the notion of early intervention to try and prevent psychopathology and functional problems from developing among individuals exposed to childhood victimisation (subthemes 1.2.3. "improving practice" and 1.2.2. "consideration of the future"). Although this approach was endorsed by young people as well as practitioners in both social work and CAMHS, most CAMH services are not currently set up to provide preventative mental health intervention as their limited resources are concentrated on those with existing and severe psychopathology instead. This has important implications for how effectively the risk calculator could be implemented into UK health and social care services. It is vital that a victimised child who is screened and identified as being at high risk of developing mental health problems has access to appropriate support regardless of whether they have already developed psychiatric disorders. Our findings cast some doubt on whether this would currently be the case (subtheme 1.2.1. "emphasis on current need"), suggesting that a shift towards preventative intervention may be required (together with allocation of enough resources) for successful implementation and utilisation of other sectors, such as schools, primary care or charities (subtheme 1.4.2. "setting"). However, we posit that it is also possible that use of the tool could contribute to this shift as it has the potential to provide a more effective means of allocating preventative resources to those who are most in need and therefore most likely to benefit (subtheme 1.2.3. "improving practice").
The availability of preventative intervention was found to be important not only for the effectiveness of the risk calculator, but it was also fundamental to practitioners' acceptance of it (subtheme 1.4.1. "lack of resources"). In the absence of appropriate intervention, screening children to see who is most vulnerable for developing poor outcomes was viewed as unacceptable; practitioners deemed it to be unethical and potentially detrimental to the child due to the risk of fuelling feelings of hopelessness and negative self-fulfilling prophecies (Merton, 1948). There are a wide range of  interventions and support currently available to promote the mental health and functioning of children who have experienced victimisation (subtheme 2.3.1. "existing interventions"). These include initiatives that are delivered in schools and communities (e.g. mentoring), targeted support for specific victimisation experiences (e.g. the NSPCC's "Letting the Future In" for sexual abuse survivors) and therapeutic support through CAMHS (e.g. cognitive behavioural therapy). More work will be needed to establish a solid evidence base for early interventions in victimised children (National Institute for Health Care and Excellence, 2018). Importantly though, we note the context of limited resources in which these services operate resulting in long waiting lists, restrictions on the duration of support, and variability in what services are available in different geographic areas, all of which could impact the effectiveness and acceptability of the risk calculator.
Another key consideration that we identified for the potential use of the risk calculator was the transparency with which the tool would be used with victimised children and their caregivers (subtheme 2.4.3. "transparency"). Our findings revealed a tension between a desire to work openly with families by involving the child in the screening and sharing the resulting risk score information, and a desire to protect them from feeling overburdened by numerous assessments and frightening information. Young people expressed similarly mixed feelings about whether or not risk scores should necessarily always be shared with children and their caregivers. There is likely not a "one size fits all" solution to this; instead we suggest that a decision may be best made on a case-by-case basis taking account of factors such as the child's age and level of understanding, any legal obligation to share information with the child's parents, and the potential for increasing risk further. Making such sensitive judgements about what and how information should be shared with vulnerable children and their caregivers is certainly not unique to the use of this risk calculator but is a core part of health and social care practitioners' training and day-to-day work. Thus, they are likely well-placed to determine what level of transparency and involvement is  Table 3. Thematic analysis of young people's (n ¼ 6) and practitioners' (n ¼ 13) views regarding the potential use of a risk calculator to identify victimised children most at risk of developing psychopathology and poor functioning in young adulthood. "Feeling there's always an expectation on me, I can always do better. So, pushing myself too hard, getting to crisis point, and then feeling even worse because there's still the expectation on me." (FG2) Young people and practitioners identified low socioeconomic status and financial pressures as increasing the risk of victimised children developing poor mental health and functioning.
"Sometimes … parents don't have time for their kids because they're working … I had to take care of my younger brother … my parents will leave at four am in the morning and they will come back at ten, and they would be shattered and they would just go back to sleep " … if someone is either working multiple jobs or they have to work very long shifts or they have to work away, it's quite difficult even arranging joint meetings sometimes, or sessions with the child and carer, and then you see well that clearly impacts the relationship [between parent and child]." (P4) Young people and practitioners felt that a family history of mental health problems could increase children's risk of poor outcomes following victimisation.
"If there's pre-existing mental health issues within your family, that might make you more likely to go on and suffer with mental health issues as well." (FG2) "Yeah, I think definitely parental mental health is the number one … ." (P9) In addition to parental mental health, practitioners also identified that parents' own unresolved trauma could increase children's risk.
"And actually that is something we see a lot: parents who have a lot of unresolved traumas of their own and I think that makes it much harder to be able to support their child and it's probably going to be quite triggering for them what their child is experiencing." (P8) Practitioners suggested that exclusion or absence from school increased victimised children's risk of developing poor outcomes. "I think young people who kind of struggle in school, academically or in terms of attendance, um because then they're not enjoying school or not going to school then there's so much more opportunity for them to slip through the net and [it] doesn't give them opportunities to be involved in like positive, enriching activities." (P8) Practitioners also recognised the co-existence of multiple stressors in a victimised child's life as increasing their risk for poor mental health and functioning.
"There's a sort of co-existence of a certain number of factors so whether that be a learning disability, or a conduct disorder, and mental health issues, and financial issues, and exploitation … ." (P4) 1.1.2. Lack of support 10 (77) Young people identified that victimised children who did not have a trusted adult who they could talk to were at greater risk of having poor future outcomes.
"One factor I feel could put children more at risk is probably the lack of a trusted adult. I say that, because in my experience … the person I had … was so key for me." (FG1) Practitioners also identified poor relationships as being a risk factor and typically focused on poor family relationships. This was recognised by young people too.
"I think a really key factor is, yeah, like family relationships, erm especially like parental relationships and whether or not you feel you can express and communicate yourself to your parents." (FG1) "It's those who don't have a family network where parents are really present in their life, erm, or able to put boundaries in place, or to really connect with (continued)  (85) However, despite this primary emphasis on the assessment of current need, practitioners did also recognise that they consider children's future as part of their work though often they described this as being an informal, rather than formal, assessment.
"Not really in a sort of formalised way  Similarly, practitioners thought the risk score might help them evidence their concerns and help them secure services for the child. "It might add weight certainly to, if it's a tool that you can get, um, sort of managers to agree or services agree could be part of a like recognised thing, then how do they, um, how do they ignore it? It's much harder isn't it? If you can then attach something to your email to say 'this is how this person scored, it's a horrendously high score for this poor outcome, and I'm suggesting I think this should happen and I need you to help me fund it', I think that would add a much greater weight to the argument." (P12) Practitioners felt that the risk calculator could help ensure they didn't overlook any victimised children who are at high risk. " … you may not have seen it because you didn't know that a certain answer, or you weren't fully aware that that thing could make them greater risk, or that particular behaviour or problem … we're not perfect, are we? So, we might not realise that that behaviour could contribute to poor mental health or whatever outcomes so that could be helpful from that point of view." (P12) Some practitioners also suggested that the risk calculator could be used to help identify service need and plan provision.
"Perhaps it would be beneficial for identifying service need and therefore possibly, kind of, creating that access to resources or more resources being created (continued) 1.3. Accuracy 9 (69) Young people and practitioners expressed concern about the ability of the risk calculator to correctly identify victimised children who are -and are not at risk.
"If certain lower socioeconomic groups have poor mental health outcomes and maybe, like, I'm from a poor background and I have a really "ok" mental health, but I'm going to be put as "at risk" because I'm at that low group … I think like a solution to that is … having more people put data into it, like a wide group of people." (FG1) "What is the accuracy of it? And are you potentially setting somebody up to suggest they're going to have poor outcomes when maybe they might not? And there's always … those that don't follow those patterns … how do you work out those anomalies?
… Are you labelling people, by doing that, into a bracket that they don't even necessarily belong in?" (P12) Practitioners noted that the accuracy of the risk calculator would be reliant on the input of accurate information which could be limited. "To me it's about the thoroughness of the ability of the people filling it in and the systems that they have, also that work across boroughs, across health districts, across different professionals. You know, systems are only as perfect as the human systems behind them." (P6) Young people and practitioners were dubious about the comprehensiveness of the risk calculator. "It doesn't account for everything. So, like, if a child has very niche things, like they're disabled, if they are from a different ethnic group, it doesn't account for all of that." (FG1) "If we don't consider these children as individuals then we may miss something … I'm not a psychologist or a statistician, I just find it hard to imagine that we could reduce young people and their outcomes to a certain number of categories of questions and how we would determine those." (P4).
Some young people and practitioners perceived human behaviour and mental health to be more complex and less predictable than physical health. They felt that this limited comparisons between this risk calculator and those used in medical practice. "I think, we have to be careful to equate it fully with physical health in the sense that it's an individual, it's not, like, a medical problem just to be dealt with, like, from that perspective. I think there's a lot more to a human being and they need to look at them as a whole human being." (FG1) "People and mental health and all of the factors that impact on that are so much more complex than sort of a physiological assessment and therefore it could be that certain factors aren't representative of what's actually happening or it might be that actually somebody is high risk of developing a mental health difficulty but if they don't tick those boxes then it might not show up as high risk." (P8)    (92) Participants reflected on whether practitioners should involve children in completing the risk calculator and whether they should share the resulting risk score with them or simply use it to inform their own work with that individual. There was a tension between wanting to protect the child from any potential negative impact of the risk calculator, their desire to work openly and honestly with them, and the child's right to access information about themselves.
"I know obviously it is not very ethical, but like, is it helpful to say to a child, like, you're at risk of developing mental health issues, when actually they haven't even raised that they're having any problems. So, maybe [it is] just something to have in mind or considered." (FG1) "If you could kind of do it, and just do it as a professional network … to try and contain the impact of the results on the young person, because if it's, I don't know, but then, because if it came out as 'you're five percent likely', that might be nice for them. I guess I'm wondering about not sharing [the risk score information], um and kind of implications of that, whether or not that's, you know, possible." (P3).
Practitioners and young people suggested using language and visual images to communicate risk to children and their families.
"Instead of giving a number to the child … use language instead, to go like, 'you're a bit at risk', or 'severely at risk', and that sort of reduces the impact." (FG1) "Sometimes when presenting percentiles for people I find using more of a colour grid and almost translating the bell curve into, we've used like a (continued) Practitioners and young people identified the need for professionals to be very sensitive in their communication with children about their risk. I think an issue such as [self-fulfilling prophecy] is more to do with the interaction between the practitioner and the patient, rather than the tool itself … and more to do with their training and interpretation of it." (FG1) " … how do they explain that to me? How do I go away from that feeling that I want to change something as opposed to it terrifying me and then feeling I'm never going back?" (P6) Practitioners believed the risk calculator should be used alongside -not in place of -professional judgement and referred to examples of existing tools that explicitly promoted this.
"I think just it's always good to have professional judgement running alongside these things and to not say 'actually this takes over professional judgement, so we don't need it anymore.' So, I think just for practitioners to be aware that those two things come hand-in-hand and one doesn't exclude the other." (P8) "I'm thinking about other tools, so the CAADA DASH risk for domestic abuse … I think you've got to score fourteen on that to get a referral to … MARAC [multi-agency risk assessment conference] … Um, but also, they say if you don't score fourteen but you have a space for professional judgement then you can still get [referred]." (P9) 1.5. Response 1.5.1. Professionals

(31)
Young people expressed concern that practitioners may have an overly medicalised response to children's risk score.
"Maybe some people just need like support … to talk to someone that they trust … maybe they don't need anti-depressants, or, erm, like psychological intensive therapy … maybe, they literally just need someone to just say 'I feel really rubbish at the moment' or 'I'm going through something really difficult' … I think, we have to be careful to equate it fully with physical health." (FG1) Practitioners expressed concerns about how the risk score could be used by some practitioners as well as the potential for organisations and private businesses to (mis) use the risk score information.
"I can imagine some very crude care planning happening on the back of a score." (P2)

"[if] the young person is perhaps high risk
… what the impact could be for them in terms of employment and insurance and, I don't know, all sorts of things, whether there's the potential for that to be misused or to experience discrimination as a result?" (P4) 1.5.2. Family 12 (92) Young people and practitioners identified the potential for parents to have a negative response to a child's risk score. Young people felt that parents may respond in ways that are unhelpful whereas practitioners also empathised that parents may feel blame and guilt.
"If the GP says to a child's parent 'your child's really at risk for mental health problems' then that child … might be treated differently by their family, and then that could actually go on to actually cause a lot of issues for the child." (FG1) "I think it could be quite disheartening for parents who are doing a really good job but maybe struggling for reasons out of their control … I'd be (continued)

Explanation
Illustrative quote(s) a very cautious to have another thing that's in these families' lives that's going to be a knock, like another potential blaming tool for them." (P9) Young people and practitioners identified that receiving a high risk score might evoke negative emotions such as feeling scared or hopeless and may fuel a self-fulfilling prophecy in terms of developing poor mental health and functioning.
"You're just going to think to yourself that I'm always going to be at risk, there's no point of me trying to better myself." (FG1) "It has to come with a package [of support] because otherwise that child is then left on their own and it's almost like 'well it's inevitable. I'm going to have these poor [outcomes].'" (P9) Despite this, young people and practitioners also recognised that families may have a positive response to a high risk score, suggesting that it had the potential to educate, empower and motivate them. "It gives young people kind of a greater opportunity to become self-aware with their experiences and what they've gone through … I think it is good for the young person to come to terms with what they've been through, erm, but actually feel in control of that and like they can change." (FG1) "It could be that drive that families need to see." (P9) 2. Protective factors and prevention 2.1. Individual 9 (69) Young people and practitioners identified that having a focus was an important protective factor for victimised children.
"Having a structured routine, you get up at this time, you go to school and work at this time, and that sort of helps you out because you have to get out of bed for work, it gives you a very subtle incentive." (FG2) " … something to focus on, to channel their energy." (P5) Young people highlighted the protective role of children having future aspirations. "I didn't want to go through teen pregnancy. I didn't want to go through really awful menial jobs. I wanted to have something better, like, a better life." (FG2) Young people and practitioners considered whether intelligence was protective. Whilst some believed it to be, others questioned whether this was necessarily the case.
"Academic intelligence is probably the best opportunity that somebody could have to get out of a situation that's really poor." (FG2) "Higher level of intelligence  "Education about relationships and sort of healthy social relationships in school." (FG2) "I think it becomes part of how the system at school, you know how you have like physical checks in school and stuff like that, I think it's important to get young people to understand that mental health should be seen in the same way as physical and to have those check-ups on a regular basis. Um, I think that would also help raise awareness around mental health in general." (P7) In particular, the potential for Personal Social and Health Education (PSHE) lessons to do this was highlighted by young people.  though that any work that was done with a child wouldn't necessarily be massively obvious to stand out, label them to peers or anything like that so if anything was done it would be done as part of their school day or to help them integrate as much as possible or be done out of school altogether at some sort of appointment or meeting that they would go to." (P12) Some practitioners felt that stigmatisation was more of a risk in families of particular cultural or socioeconomic backgrounds "So, I don't necessarily think stigma would be a problem although there are some who I think, definitely some communities as well it's still kind of looked down on." (P11) Despite this potential risk, the apparent acceptability of preventive intervention among children was noted. (continued)

Explanation
Illustrative quote(s) a " … mental health seems so much more talked about and accepted in schools and among young people than it was before." (P8) Practitioners' highlighted the importance of ensuring preventive interventions are appropriate for the child in terms of content, timing and support available to them in order to minimise unintended negative consequences. " … you don't carry on with an intervention because it's for you, you're negotiating it and you're also managing the risk of 'ok this means you don't sit your exams but this will be for a short period of your life and then you will able to access education and it will help you' versus 'this isn't the right time for you.'" (P6) Practitioners also identified the withdrawal of support at the end of a preventive intervention as carrying risks for the child and family.
"I guess the first thing that came to my mind was the support withdrawing and how do families manage the withdrawal of support, there's potentially got them to a place where they're feeling confident and obviously it would be, well I don't know because if it's funding, you may get your six months and then regardless of where you are and regardless of if you're ready or not that support may have to be withdrawn and you could be stepped down but you know transitions tend to be very difficult for chaotic families." (P3) Practitioners noted the potential for an intervention to impact negatively on an individual's willingness to seek help in the future. The source of a quote is denoted by the focus group number (FG#) or practitioner number (P#). appropriate for a particular child they wish to screen whilst upholding core ethical values of honesty, openness and protecting from harm. Finally, our findings suggest that successful implementation of the risk calculator will require training for the practitioners who will use it. We found that people were very knowledgeable about the wide range of factors that could buffer or accentuate the risk of poor outcomes following exposure to victimisation (themes 1.1. "risk factors", 2.1. "individual" and 2.2. "environmental"). However, awareness of this complexity and of individual differences in response to victimisation gave rise to scepticism about whether a statistical tool could accurately predict such outcomes (theme 1.3. "accuracy"). Increasing practitioners' understanding of the risk calculator as a means of individualrather than averageprediction and evidencing its accuracy will therefore be critical in promoting their acceptance of the tool. This is consistent with the What Works for Children's Social Care recommendation for the use of prediction models and their limitations to be included in social worker training (Leslie et al., 2020). Additionally, practitioners will need to be trained in how to interpret the risk scores. In particular, further work is needed to determine a clinically appropriate cut-point to guide practitioners regarding what they should interpret as a "high" risk score that requires action (subtheme 1.4.4. "interpretation").

Limitations
We acknowledge some limitations of our study. First, our sample of young people was relatively small. Second, the self-selecting and predominantly female sample may limit the generalisability of findings as those who volunteered may have a greater interest in research and a proclivity towards the use of a screening tool. However, the views of our young people and practitioners were not unequivocally positive with regards to the acceptability and implementation of a risk calculator. Moreover, men are indeed underrepresented in health and social care roles in the UK (Les, 2017;Skills for Care, 2017) such that our practitioner sample reflects the female-dominated gender composition of these professions. Despite attempts to achieve a balanced gender composition of young people, the low number of males means that this sample is not representative of children who experience victimisation (Radford et al., 2013). Third, the study did not include parents/carers or teachers who also have a key role in supporting children and young people who have experienced victimisation. Ascertaining the views of these stakeholders is also important, particularly given our finding that schools and/or multi-agency meetings may be a useful setting for the risk calculator to be used. Finally, this qualitative study was undertaken while the risk calculator is still in the early stages of development and validation. Therefore, we were unable to demonstrate to participants exactly what the risk calculator would look like or how well it performs in external samples. Nonetheless, finding out what young people and practitioners' feel are the main considerations is critical to inform the further development of the risk calculator and future implementation efforts.

Conclusion
Notwithstanding these limitations, the findings of the current study provide useful insights into the acceptability and feasibility of implementing a risk calculator to identify victimised children who are most at risk for developing psychopathology and poor functioning. Overall, young people and practitioners recognised the potential for an accurate screening tool like this to enhance health and social care practice but also highlighted key considerations and challenges related to its successful implementation. This will be useful to inform future work in this area.