What do young people who self-harm find helpful? A comparative study of young people with and without experience of being looked after in care.

Background Self-harm amongst young people is an increasing problem, with looked-after young people at higher risk. Despite this, little research exists on what young people who self-harm find helpful. Young people who self-harm have mixed views about CAMHS. Differences in the pattern of access and preferences for support between looked-after and non-looked-after young people should be reflected in service availability and commissioning.

• Little is known about what young people who self-harm find helpful, particularly for lookedafter young people. • CAMHS was named amongst the most helpful and the least helpful services by both lookedafter and non-looked-after young people. • Social services and A&E were frequently cited amongst the least helpful sources of support.
• Young people report finding informal support helpful, including family and friends. • There appears to be a need for explorations of the reasons behind the apparent negative perception of statutory services in young people, and what is required to shift that.
Self-harm is an act of self-injury or self-poisoning with or without suicidal intent (Hawton et al. 2007). The age distribution of those who self-harm is markedly skewed towards adolescence (Hawton et al. 2007) with 7-14% of adolescents having self-harmed at least once by the age of 15-16 (Hawton et al. 2002). Rates of self-harm are increasing amongst young people , presenting a growing problem.
In 2018, 75420 young people were registered as living in the care of the state (known as lookedafter) in England (Department for Education 2019). Looked-after young people have an increased risk of a number of mental health difficulties including self-harm (Stanley, Riordan & Alaszewski 2005, Evans et al, 2017, with 45% meeting criteria for at least one mental health disorder (Meltzer et al. 2003) compared with 14.6% in disadvantaged private households and 8.5% in other private households (Ford, Vostanis, Meltzer & Goodman 2007). Proposed causes for this include that by the time the child has entered the care system they have experienced high levels of psychosocial adversity, which are often compounded by stressors and vulnerabilities of being in care (Ford et al. 2007) In the UK, Child and Adolescent Mental Health Services (CAMHS) work alongside other agencies such as social care and education to form a multi-agency network (Hay et al. 2015). Not all young people who self-harm are seen by CAMHS, and despite in some cases the involvement of multiple agencies, care for young people who self-harm is often disjointed and sub-optimal. Taylor et al's. (2009), systematic review of the international literature on attitudes towards clinical services for individuals who self-harm, showed that many participants' perceptions of their management were negative. For young people who self-harm there is a lack of research into sources of support which are not professional services.
Self-harm is the strongest known risk factor for eventual suicide (Carrol, Metcalfe & Gunnell 2014). The suicide prevention strategy in England (Department of Health 2012) notes a need to "tailor approaches to improve mental health in specific groups" (p.21) including looked-after young people. Despite these aspirations, not enough research has been done to explore what looked-after young people find helpful and unhelpful and what support they would like which they do not currently have access to.
We used an Audio Computer Assisted Self-Interview (ACASI), this has been shown to (i) enhance disclosure on sensitive topics (Lessler et al., 2000) (ii) avoid problems with literacy because participants hear the questions read aloud over headphones and give the appropriate response with a mouse click. It also has excellent criterion validity (Paschall et al. 2001). We have previously used ACASI with young offenders who self-harm (Knowles et al. 2011) and young people who self-harm referred to CAMHS (Glazebrook et al. 2015).
An advisory group of young people who had personal experience of self-harm, a number of whom also had experience of being looked-after, worked with us to shape the research and prioritise the research questions. They played an important role in ensuring that our study materials were suitable for young people and worked with us throughout the project to refine research questions and dissemination plans.
The key questions this study aims to answer are: 1) What services do young people who self-harm find most helpful? 2) What services do young people who self-harm find least helpful? 3) Are there differences between the views of looked-after young people and those who have never been in care with regard to what they find supportive to promote recovery and reduce distress? 4) What support would young people like which they are not currently receiving?

Method
The "Listen-up" study Data are taken from the "Listen-up" study, a mixed methods investigation which aimed to improve knowledge and understanding of the experience of self-harm in young people. Two studies were conducted as part of this work. The first involved qualitative interviews with young people (e.g. Wadman et al. 2018) and developed and tested the Card Sort Task for Self-Harm (CaTS) (e.g. Townsend et al, 2016). The second study used the ACASI, the data presented below are the first to be published from this. Recruitment 126 Participants aged 11-21 years were recruited for this study. Initial recruitment was through CAMHS (both inpatient and outpatient services) and Children's Services (social care) in two East Midlands counties of the UK. Potential participants were identified and initially approached by their usual care team. If the young person gave permission, their details were passed to the research team. Further participants who had experienced self-harm were also identified in the community through self-harm organisations, youth clubs, secondary schools, leaflets, social media and a project website.
All participants aged 16 and over provided written informed consent. For those under 16, assent was sought from the young person and consent was gained from their parent/legal guardian.
Ethical approval was given by the Social Care Research Ethics Committee (NHS Health Research Authority), REC number 14/IEC08/0015, and the departmental ethics committee. A referral pathway to clinical support was available if any participant became distressed during the research.

Interviews
Participants completed an ACASI, the questions for which were discussed with the study's Young Person's advisory group, to ensure their acceptability and suitability to elicit information. Interviews were conducted in a private space of the participant's choosing (home, volunteer centre, private room at school). The participant also chose whether to complete it online on their own or with a researcher present. The ACASI included questions about the participant's demographics, positive and negative factors which lead to and influence recovery from self-harm and services/therapies/supports which may help to promote recovery from self-harm. Participants were given a series of Likert scales asking them to indicate the extent to which they agreed/disagreed with the statement "I have found XXX helpful in promoting recovery from self-harm and reducing distress". The potential sources of support included services such as counselling, professionals such as psychiatrist, individuals such as girlfriend/boyfriend and activities such as exercise. For each question, the participant could answer whether they: strongly disagree, disagree, neither disagree nor agree, agree, strongly agree, never received this treatment or would prefer not to say.
Participants were then asked to select from the supports they had received which they had found the most and least helpful. They were able to respond using free text.
Finally, participants were asked what services, therapies and supports they would like but are currently not receiving and were able to use free text.

Analysis
The analysis looked at the data as a whole then compared responses between the two study groups, looked-after and non-looked-after young people.
For each question, frequency tables were generated to assess how many young people chose each response on the Likert scale.
The percentage of those who indicated that they had used/received each service/therapy/support was calculated. Those who agreed and strongly agreed were combined to calculate a percentage who found the support helpful, likewise those who disagreed and strongly disagreed were also combined to form one category.
For the free text responses, answers were grouped where it was clear the same response was meant but different spellings were used, for example "CAMS" and "CAMHS", and frequencies of responses calculated. All groupings were checked and agreed by two authors (JH and KS) to ensure reliability.
Chi squared analyses were used to compare looked-after and non-looked-after young people with regard to the services/therapies/supports they had experienced and what they found helpful. Bonferroni corrections were applied to all calculations due to multiple comparisons being made.
Results 126 young people aged 11-21 participated in this study, of these, 53 had experience of being in care and 73 did not. The mean age was 16.49 years (standard deviation (s.d.) 1.98), (16.00 (s.d. 1.32) for looked-after young people and 16.85 (s.d. 2.29) for non-looked-after young people). 84.1% of the sample identified as female (81.1% of the looked-after young people and 86.3% of the non-lookedafter young people). 1) What services do young people who self-harm find most helpful? Across all participants, the ten sources of support which had been experienced as helpful by the highest percentage of young people are shown in Table 1; the top three were friends, distraction techniques and pets.
2) What services do young people who self-harm find least helpful? Across all participants, the ten sources of support which had been experienced as unhelpful by the highest percentage of young people are shown in Table 2; the top four were psychiatrists, harm minimisation techniques, medication and A&E.
3) Are there differences between the views of looked-after young people and those who have never been in care with regard to what they find supportive to promote recovery and reduce distress? When asked what the most helpful service/therapy/support for self-harm was, there was a wide range of responses, as reflected in the low frequencies presented. The three most frequent responses for looked-after young people were CAMHS, friends and pets, and CAMHS, Counselling and Harmless for non-looked-after young people (Table 3).
Since they were able to type their responses with free text, some gave a reason to explain their answer.
"Having someone who fully understands your life, my counsellor and mental health support worker have been most helpful." (Non-looked-after young person)

TABLE 4 HERE
When asked about their least helpful service/therapy/support the most frequent responses were CAMHS, A&E and Social services for looked-after young people and CAMHS, GP and CBT for nonlooked-after young people (Table 4).
Again free text responses sometimes gave additional information.
"Inpatient treatment-being unable to self-harm in the units made it worse when I came out." (Nonlooked-after young person) "Social services tend to make me feel worse but I think this is only because of my past experiences being involved with them." (Looked-after young person) Post-hoc analyses compared participants who had placed CAMHS as the most and least helpful source of support but found no difference in the group age, sex distribution or whether they had been looked-after.  Table 5, the most accessed supports for non-looked-after young people were harm minimisation, distraction techniques, counselling, friends and pets. For looked-after young people these were: foster carers, harm minimisation, counselling, distraction techniques and friends.
When comparing looked-after with non-looked-after young people, significantly more non-looked after young people had experience using distraction techniques, meditation, GP, websites and psychological therapies for support.
More looked-after young people had experience of accessing A&E for support with self-harm and using CAMHS.
In terms of what young people did and did not find helpful, a higher proportion of looked-after young people found support groups helpful. A higher proportion of non-looked-after young people found distraction, medication and their siblings helpful. 4) What services would young people like which they are not currently receiving?
When asked about supports they would like, which they were not currently receiving, participants gave a wide range of answers. For the non-looked-after young people the most frequent responses were n/a, "none" and support from those with similar issues. See Appendix A for a frequency table of responses from each group. Some described particular characteristics of this support.
"Support groups for young people across the country where all that is said is confidential… Getting advice from those with similar experiences." (Non-looked-after young person) "At the moment I could do with someone to chat to who's been in a similar position to me -dealing with a parent with terminal cancer." (Non-looked-after young person) A number of different psychological therapies were suggested including: counselling, CBT, dialectical behavioural therapy, art therapy and drama therapy. Others also requested a better experience when using an existing service.
"In an ideal world a really safe place you could go if you felt like hurting yourself/after you'd hurt yourself, where there were people trained in self-harm who understood." (Non-looked-after young person) "Nice A&E experience (when I need it)" (Non-looked-after young person) The most frequently requested supports from the looked-after young people were n/a, "none" and "don't know". Three asked for CAMHS support. Fewer responses were offered with very few suggestions of psychological therapies. Some used this as a place to note times they had felt unsupported.
"More help from the personality disorder team who have let me down as well as many others" (Looked-after young person).
"Made me feel even lower in myself" (Looked-after young person).

Discussion
This study provides unique insight into the experiences and views of a difficult to reach and vulnerable patient group, young people who self-harm, and in particular looked-after young people who self-harm.
1) What services do young people who self-harm find most helpful?
When considering the sources of support reported as helpful by the largest percentage of young people, most were individuals around the young person such as friends, partner, pets, parents and teachers or relatively easily accessible interventions including distraction techniques, exercise and harm minimisation techniques. The only professionally-delivered service in the top 10 was counselling, which is also a more accessible and non-specific intervention.
These findings show that people around the young person in their daily lives are important sources of support. This highlights that to optimise support and its reach for young people, approaches should come at a number of levels, including helping families and caregivers ( including foster carers and residential workers) understand self-harm, and training and resources for those working in education.
In 2017, the UK government published its Green Paper for transforming children and young people's mental health services (Department of Health and Social Care & Department for Education 2017). Both this, and the NHS long term plan (The NHS Long Term Plan 2019) have put in place plans to provide additional mental health support through schools and colleges in the form of developing Mental Health Support Teams. A recent survey shows that schools would welcome further support and staff training in supporting those with self-harm (Evans et al. 2019). The findings from this study, that teachers ranked amongst the most helpful sources of support, endorses the importance of schools as a place where young people seek out help with self-harm and the need for these provisions.
Where professional agencies are involved, part of their role should include mobilising informal sources of support. This may present more of a challenge for looked-after children, who struggle to access consistent, supportive adults in their day-to-day lives. Peer and adult mentors could thus have an additional role to play. For those who require CAMHS input, these findings support consideration of family interventions such as family therapy and parent/carer focused interventions such as Theraplay but further research is needed to measure the clinical effectiveness and acceptability of these.
Further research is needed to better understand which websites are found helpful by these young people, since previous research has shown that websites can also be unhelpful in self-harm in young people (Mitchell et al. 2014).
2) What services do young people who self-harm find least helpful?
The low percentages in Table 2 show that there is variation amongst young people in what they find unhelpful, supporting findings that young people self-harm for various reasons and have differing needs (Cottrell et al. 2018). Sources of support most likely to be reported as unhelpful included services such as: psychiatrists, medication, A&E, psychological therapies, CAMHS, GPs, social workers and counselling. Parents also fell into this list, as well as some techniques that others reported as helpful: harm minimisation and distraction techniques.

CAMHS was ranked amongst both the top three most and least helpful sources of support for both groups. CAMHS is a broad term that includes both community and inpatient services and is provided by different organisations including health and local authorities. Each young person's interaction
with CAMHS is likely to differ in terms of: waiting times, practitioners, treatments offered and follow up. Clearly for some, CAMHS is providing a helpful service, however, for others, they do not feel supported by this specialist service. A qualitative study exploring the experiences of these young people would give further insight into the factors that under-pin these split opinions.
Harm minimisation and distraction techniques also received split responses. Harm minimisation techniques focus on ways to minimise the physical consequences of self-harm. Distraction techniques involve learning about the patterns of emotions surrounding the self-harm and diverting attention to other activities until the urge has passed. For harm minimisation, around a third of participants found it helpful, however, almost the same percentage found it unhelpful. For distraction techniques almost half of the young people found it helpful, however, around 20% indicated it was not helpful. As shown in Table 5, these are some of the most highly accessed sources of support. The differences in young people's opinions may be due to differences in techniques used, level of practitioner training or differences in the young person themselves. It is important to note that recent mixed-methods research indicates that young people view harm minimisation to be unhelpful and it may even end up being used as self-harm (e.g. using very thick rubber bands to break the skin) (Wadman et al. 2019).
3) Are there differences between the views of looked-after young people and those who have never been in care with regard to what they find supportive to promote recovery and reduce distress?
Since looked-after young people have allocated social workers, it is not surprising that they feature more highly in their responses; these findings suggest that they are often not found to be supportive by the young people themselves in relation to their self-harm. This may be due to the practical role of a social worker, perhaps implementing plans following self-harm e.g. change of placement which a young person may perceive as punishments.
There are a number of different factors which influence whether young people seek help from professionals, and if so which professions (Michelmore & Hindley, 2012). More non-looked-after young people reported experience of seeking help from their GP for self-harm, with over two thirds of them having used this service; in contrast a significantly higher proportion of looked-after young people had attended A&E with their self-harm, with over 70% reporting this compared with half of the non-looked-after young people. Each looked-after young person receives a medical review with a paediatrician once a year, these appointments, as well as placement moves, may mean lookedafter young people are less likely to develop a relationship with a GP and use them for support. However, this may also reflect the severity and timing of the self-harm or the preferences of those caring for these young people.
Both looked-after and non-looked-after young people listed A&E among the least helpful services. As shown by Owens and colleagues (2016) those who self-harm try to avoid attending A&E; when they attend they often find the experience negative. A&E attendances are often triggered by others rather than the young person themselves, this may mean that during attendances young people may not feel that is the time or place to receive support . GPs featured amongst the top 10 most helpful for non-looked-after young people but also appeared in the least helpful list, this may be because some GPs do not feel able to offer specialised support without the necessary infrastructure or training (Michail & Tate, 2016). It is also possible that this is a reflection of the time pressure on GPs to keep appointments to 10 minutes. Despite this, other UK research has found that most adolescents who self-harm report having seen their GP in the previous 6 months (Sayal et al, 2014).
A higher proportion of looked-after young people had experience of CAMHS. As looked-after children have higher rates of psychiatric illness than non-looked-after young people (Meltzer et al. 2003), this is reflected in commissioning of specific looked-after CAMHS teams. However, this finding may also be due to looked-after young people presenting to A&E rather than GP and thus triggering more quickly a referral to CAMHS, or it may be due to referrals by paediatricians or social workers.
Non-looked-after young people had more experience of using distraction techniques, meditation, self-help books and websites. This may be due to better access to books, computers and the internet or may be due to whom they have sought help from e.g. GPs may direct patients towards self-help.
A higher proportion of non-looked-after young people found distraction techniques helpful for their self-harm. This is likely to vary on an individual basis, but one reason for this may be how comfortable a young person feels in their living environment, which may be unstable and distressing for a looked-after young person.
More non-looked-after young people also found medication helpful. Again this will vary between individuals but may reflect that on average looked-after young people may be less likely to be prescribed medication since they will have experienced greater levels of complex trauma, insecure attachment and impairment secondary to being in care (Stanley, Riordan & Alaszewski 2005), the effects of which may not be easily helped with medication and require attachment-focused therapeutic approaches and multiagency bio-psycho-social interventions.
A higher proportion of non-looked-after young people found their sibling helpful with support for their self-harm; there is a higher likelihood that the non-looked-after young person live with their sibling but also the nature of sibling relationships may be more complex when young people have adverse childhood experiences which have brought them into care.
Although only a fifth of the looked-after young people had experienced support groups, all those who had experienced them found them helpful, this could represent an underused source of support.
4) What services would young people like which they are not currently receiving?
When asked to state what supports they would like, participants provided a wide range of responses. When young people responded with 'n/a' or 'none' it is not clear if this was because they felt adequately supported or they did not feel anything would help. Responses were mainly services and therapies, including improved existing services such as a more confidential support group, a more helpful A&E or more safe and accessible inpatient care. This may indicate a need for better implementation and adequate resourcing of existing services rather than service transformations. The non-looked-after children provided a larger variety of psychological therapy suggestions, this may be due to a higher percentage of this group having accessed psychological therapies.

Strengths and limitations
This study has a number of strengths. Firstly, it gained responses from a large number of young people with a history of self-harm, particularly looked-after young people who are a hard to reach group for research. The use of a computer interview allowed young people to express their views in a confidential and non-confrontational way. The use of free text answers allowed the young people to answer however they wished without being influenced by what was available.
Since young people were also able to answer when they had never experienced a service/therapy/support this helped to give an indication towards which services are being used. Although this does not provide us with information on the referral pathways or reasons for these differences, it does provide a starting point to try and understand this.
This study also explored how helpful young people found sources of support. This gives a more detailed picture of the support around a young person; often clinicians focus on the services a young person has been referred to without thinking of wider resources a young person may have, such as their family and friends, current coping strategies and information gathering techniques.
One limitation of this study was that the computer interviews were quite long and included a number of questions in a very similar format. This may have caused young people to feel less engaged with the interview or to skip questions. To try and keep the interviews from being too long some sources of support were grouped e.g. different therapies or CAMHS interventions, this limits our understanding of what people are referring to in their answer.
The use of Likert scales also caused a loss of richness of data, it does not tell us why the young person found a service helpful or unhelpful, only whether they did. A qualitative study, looking in greater detail at why young people answered in the way they did would give further insight into what is and is not supportive, and why; this would be particularly helpful to understand issues raised by this study such as polarised views regarding CAMHS and why some young people did not find GPs or social workers supportive.

Conclusion
This study has provided important insight into the views of a large sample of young people who selfharm. Focusing solely on professional services for the management of self-harm ignores what the young people find most supportive which includes the individuals around them. Specialist services were reported to be supportive by some but not others. Universal statutory services such as social workers and A&E were often found to be unhelpful. There are differences in access to certain sources of support between looked-after and non-looked-after young people and in what they find helpful. Qualitative work within this area is urgently needed to try and better understand the factors that have contributed to these perceptions and attitudes and how to improve support for these young people.
Tables Table 1 The ten sources of support which had the highest percentage of young people who found them helpful. Parents 15.1 10 GP 13.5 Table 3 The top 10 most frequently cited most helpful sources of support by looked-after and nonlooked-after young people.   Table 5 The percentage of looked-after and non-looked after young people who had experienced each service and the Bonferroni corrected p values for Chi-squared comparisons between groups.

Ranking Support/Therapy/Service
Where there was a significant difference between groups the bold value represents the group with the higher percentage who had experienced/found this helpful. **since all those who used apps found them helpful, it was not possible to perform a chi squared calculation.