A retrospective study examining the adverse effect of childhood abuse among adult psychiatric service users in Britain

: The long-term effects of childhood abuse have been well studied though the effects of abuse of speciﬁc types in adult psychiatric service users are less known. This study aimed to assess the association between childhood abuse and the development of harmful social and behavioural outcomes among adult psychiatric service users. Adult psychiatric service users were accessed from secondary mental health services in South London. A retrospective analysis was conducted of a randomly selected sample of 342 mental health records. Chi-square tests and logistic regression models were used to examine associations between childhood abuse – sexual abuse, physical abuse and psychological/emotional abuse – and health and behavioural outcomes. This study identiﬁed that 109 (31.8%) psychiatric service users had some history of childhood abuse and more often presented with high-risk or severe behaviours. Sexual abuse in childhood was linked with social isolation related to loss of friends (odds ratio (OR) = 2.68, P < 0.01), risky behaviours such as binge drinking (OR = 2.15, P < 0.05) and self-harming (OR = 2.86, P < 0.01), while childhood physical abuse was associated with drug abuse in adulthood (OR = 1.88, P < 0.05). Revictimization (adult domestic violence) also impacted on service users’ quality of life in terms of loss of housing (OR = 2.21, P < 0.05) and loss of friends/family contact (OR = 2.73, P < 0.01). These ﬁndings suggest childhood abuse may play an important role in shaping risk and vulnerability for mental health problems across a lifespan. In acute mental health services, the incorporation of a trauma-informed nursing care model is necessary to generate a shift in culture in the delivery of care.


INTRODUCTION
Childhood abuse describes deliberate acts or threats of harm committed against a child, which can be physical, sexual or emotional. It can have injurious effects throughout the life span on the child's physical, developmental and psychosocial well-being (Giardino, Lyn & Giardino 2019), and ultimately diminish children's potential to thrive as adolescents and later as adults (Anda et al. 2006). Although child abuse is a fundamental risk factor for poor adult health and is associated with a substantial social and health burden, in the UK, the government begun to embrace evidence that child abuse occurred on continuums of severity and chronicity (Woodman et al. 2012) only when the Munro Review of Child Protection (2011) was published. Since then, a number of policies were put forward to safeguard children (e.g. Children and Families Act 2014, Working Together to Safeguard Children 2018, Keeping Children Safe in Education 2019). Nevertheless, less has been done in terms of cross-cutting policies to prevent and/or reduce the cumulative effect of early childhood experiences in (young) adults with childhood abuse experiences which have been linked to risks such as substance and alcohol abuse, deprivation, poor educational attainment and psychiatric problems as the extant literature illustrates.

BACKGROUND
The downstream effects of childhood abuse or living in an unsafe household have been described in several studies exploring the long-term effects of childhood adversity (Felitti et al. 1998;Gilbert et al. 2009;Hughes et al. 2017;Lippard & Nemeroff 2020). Since Felitti et al.'s (1998) study on Adverse Childhood Experiences (ACEs), research has confirmed the detrimental impacts of toxic stress on health Brown et al. 2009;Kelly-Irving et al. 2013).
Research findings indicate a strong and cumulative association between exposure to adversity during childhood and the adoption of health-harming behaviours and poor mental health across the life course (Bellis et al. 2015;Bellis et al. 2014). Studies demonstrate that the more childhood adversities someone experiences the more detrimental the effect on (a 'graded dose-response') numerous health, social and behavioural problems throughout their lifespan (Felitti et al. 1998). Experiencing more than four childhood adversities is considered to be the tipping point for prediction of future health problems (Mitchell et al. 2015) with many problems related to childhood being comorbid (Cougle et al. 2010;Koball et al. 2019).
Moreover, comparative studies have examined abused and non-abused adults, demonstrating that abused children experience multiple types of victimization (Radford et al. 2013), which is a substantial risk factor for adult revictimization in violence-exposed children (Thoresen et al. 2015;Trickett, Noll & Putnam 2011). Mediation analysis has highlighted how childhood abuse mediates disease risk and course, showing it increases risk for developing psychiatric disorders (e.g. mood and anxiety disorders, post-traumatic stress disorder (PTSD), antisocial and borderline personality disorders, and alcohol/substance use disorders) (Alisic et al. 2014;Carliner et al. 2016;Lippard, & Nemeroff 2020;McLaughlin et al. 2012). Psychiatric disorders are interrelated with revictimization and are a risk factor for revictimization through symptoms of post-traumatic stress disorder (PTSD) (Ullman, Najdowski & Filipas 2009;Widom et al. 2008). Experiences of child abuse are associated with childhood risk of pscyhiatric disorders, contributing to enduring vulnerability to psychopathology that persevere into adulthood (Green et al. 2010).
Longitudinal studies have generally confirmed and extended previous research findings on the impact of child abuse on adult psychiatric outcomes. In an early prospective longitudinal study in the United States, Widom, DuMont and Czaja (2007) reported that children who were physically abused or experienced multiple types of abuse were at increased risk of lifetime major depressive disorder in early adulthood. Further to these results, Fergusson et al.'s (2013) 30-year longitudinal study examining the consequences of child sexual abuse for multiple outcomes in a birth cohort highlighted that greater exposure to child sexual abuse was associated with growing risks of adverse outcomes from age 18 to 30. Results showed that after covariate analysis (e.g. socio-demographic background, family functioning, and child factors such as being female, low IQ aged 8-9 and regular/more severe physical punishment < 16 years), child sexual abuse was linked to adverse adult developmental outcomes including mental problems (e.g. suicide attempts/ideation, alcohol/drug dependence), psychological well-being (e.g. number of PTSD symptoms score and self-esteem scores) and socioeconomic well-being. Although in this study the individual effect sizes for child sexual abuse ranged from small to moderate, it underscored that cumulative adverse effects on adult developmental outcomes were considerable. Disparities in the long-term effects of different types of abuse were found in Kisely et al.'s (2020) study, a prospectively collected recordlinkage analysis examining whether child abuse (notified and/or substantiated) was associated with adverse psychological outcomes when participants were 30. After adjusting for possible confounders, disparities in the long-term effects of different types of abuse were noted, with physical abuse having a stronger association with depression, and sexual abuse with PTSD. Substantiated abuse or multiple forms of abuse had generally the worst outcomes.
To explain some of the adverse outcomes associated with chronic and multi-type abuse, complex trauma is often used. It embodies the combination of simultaneously occurring and interacting symptoms, disorders and multiple adverse experiences, and the broad range of cognitive, affective and behavioural outcomes associated with prolonged trauma in early in life and involving an interpersonal abuse (Price-Robertson et al. 2013;p.5). Individuals with complex trauma may display an inability to relate to others, often affecting their capacity to form healthy relationships. Other features include inability to regulate emotions, changes to the sense of self, changes in sense of well-being, potentially leading to despair and hopelessness as well as elevated risks to personal safety and somatic symptoms (Price-Robertson et al. 2013). It is suggested that these manifestations function as adaptations to the experience of abuse which then become problematic for the individual. For example, behaviours such as aggression, selfinjury or substance abuse may be an attempt to manage distress or to self-soothe (Pearlman & Courtois 2005).
In England and Wales, there have been inconsistent measures of abuse experienced in childhood and a lack of complete statistics. The Crime Survey for England and Wales (CSEW) first-ever compiled data in 2019 from different sources of data thereby enabling a better understanding of the breadth and set of conditions related to child abuse including data on different types of abuse (sexual, physical, emotional and neglect) (ONS 2020a). The estimates. related to adults aged 18 to 74 years highlight that one in five adults experienced at least one form of child abuse whether emotional, physical, sexual, or witnessing familial violence or abuse, before the age of 16 years (8.5 million people), and that 3.1 million of these adults were survivors of sexual abuse before the age of 16 years (ONS 2020a). It is shown that many cases of child abuse stay hidden and do not come into contact with the criminal justice system (ONS 2020b).
Research in primary care settings adopting clinical audit of 11 practices, reported that the rate of children with at least one maltreatment-related code was 8.4/ 1000 child years (Woodman et al. 2012). Studies on childhood abuse in adult psychiatric population are scarce although childhood abuse is more prevalent among individuals with psychiatric disorders than healthy individuals (Bonoldi et al. 2013;Fisher et al. 2010). For instance, outpatients with a clinical diagnosis of mood, schizophrenia, psychotic, adjustment and anxiety disorders have higher prevalence rate compared to community samples (Devi et al. 2019). Nevertheless, some studies suggest they have higher global symptom severity, are more likely to self-harm and kill themselves (Felitti et al. 1998;Hepworth & McGowan 2013), to be prescribed antipsychotics and antidepressants (Anda et al. 2007) and have increased problems associated with the inpatient care pathway (Hepworth & McGowan 2013). Studies have shown the relationship between childhood adversities and increased health care utilization among adults (Chartier, Walker & Naimark 2010;Koball et al. 2019) reporting lengthier duration of hospitalization and younger age at first admission (Koball et al. 2019). As a result of high health service utilization, costs to public health systems (Koball et al. 2019;Saied-Tessier 2014) and productivity losses in adulthood (Thielen et al. 2016) are considerable. Despite the relevance of enquiring about childhood abuse in acute mental health settings, mental health professionals do not routinely enquire about childhood abuse among service users (Hepworth & McGowan 2013;Read et al. 2018a). In a study, nurses expressed their concerns about increasing the service users' distress (Gallop et al. 1995) and a reluctance to talk about the subject of child abuse (Gallop et al. 1998;Lab et al. 2000).
To better respond to the mental health needs associated with child abuse among individuals in psychiatric settings, 'agency surveys' (Jud, Fegert & Finkelhor 2016) like the present study are necessary. As psychiatric populations with experience of childhood abuse are under-researched, this study aimed to investigate the association between childhood abuse and the development of harmful social and behavioural outcomes such as binge drinking, (other) substance abuse, selfharm and experiencing loss of friends and family in this population, accounting for experiences of adult victimization (e.g. domestic abuse). Uniquely, this study highlighted the co-occurrence of childhood abuse sub-types in a sample of adult psychiatric service users, underscoring the interconnectedness of child abuse experiences with an array of multiple and interacting symptoms such as binge drinking and (other) substance abuse. These externalizing behaviours were high risks for revictimization (e.g. domestic violence), with associated risks to both loss of housing and loss of friends or family contact. The study illustrated the high utilization of care in mental health services among adults with experience of childhood abuse.

METHOD Design
We conducted a retrospective analysis of 342 electronic care records of psychiatric service users in secondary mental health services in South London. We included service users of secondary mental health services aged between 18 and 64, and present in the inpatient hospital during survey periods (1 April 2014-31 March 2015; including closed cases open at any point in the period). We excluded from the analysis psychiatric service users from the following categories: those with mild-to-moderate mental health needs (as rated by Mental Health Clustering Tool (MHCT)); Child and Adolescent Mental Health Services only (CAMHS); Liaison only; Assessment only (i.e. not yet taken on as clients); Section 136 of Mental Health Act only (police powers to take/keep someone to a place of safety); and Duplicates (i.e. count in just last episode of care).
Operationally, the types of adversity were defined by using clinicians and psychiatric service users' identifications. When surveying the records, descriptors such as 'sexually abused as a child' were considered sufficient to indicate abuse having occurred and coded as such. If information was vague or allusive of adversity, records were then independently examined by the two researchers (initials to be inserted). Only those cases rated as '95% or more probable' to have taken place by both researchers were included in analyses.

Data collection and analysis
Information Services generated data for this study. This service manages local databases storing details from hospital patient administration systems. The total number of uniquely registered users of secondary mental health services in South London, who fitted the aforementioned criteria were 3045. From these cases, a random sample of 342 was selected, which was generated through the NEWID (Transact-SQL) technique (Microsoft 2008). The sample size was calculated taking into account population size, confidence levels, accuracy and expected incidence.
Extracted case notes were examined by researcher [initials to be inserted] who took an average of 120 min per record. For this study, a standardized data collection sheet was created to collect clinical and demographic information, which included noted incidents of (childhood) abuse (<16 years old) anywhere in the file, such as sexual abuse, physical abuse and emotional abuse. Additionally, to control for any potential effects of adult victimization on health and behavioural outcomes, data for adult (>16) domestic abuse (indicating intimate partner violence) and adult community violence (violence perpetrated by stranger or acquittance) was collected. Outcome measures such as heavy/binge drinking, other substance misuse, self-harm, loss of housing/ having to move, loss of employment or education opportunity, and loss of friends or family contact were collected from clinical risk assessment record section in RiO electronic care record system. These are clinically operationalized and each outcome has a tab where clinicians routinely collect this data. In their data collection sheet, and for each outcome, the researchers noted with a 'yes' or 'no' to indicate that each tab contained descriptions of these adverse outcomes. Before embarking on data collection, in order to ensure a high degree of care in identifying the appropriate data, the two researchers (senior researcher . . ., and post-graduate psychologist . . .) were trained by Trust staff. The standardized training record sets RiO electronic care record system was adopted for this purpose.
Descriptive data are presented in the form of frequencies and percentages. Child abuse relationships with adult victimization and health, behavioural and social care outcomes were examined using chi-square tests with odds ratios (ORs) and 95% confidence intervals (CIs). To analyse the association of different types of child abuse controlling for adult victimization and socio-demographic factors, multivariate logistic regression models were constructed for each outcome measure. Multivariate tests of association are presented as beta and Wald chi-square values, (adjusted) ORs with CIs and Nagelkerke Pseudo-R2 as a measure of goodness of fit for the models. The level of significance in analyses was set at P < 0.05. All statistical analyses were completed with SPSS, Release 25.0 (SPSS, IBM).

Ethical approval and research governance
Ethical approval to conduct the study was granted in March 2015 by the National Research Ethics Service (NRES) Committee London (reference number:15/LO/ 0340). Local approval was obtained from the mental health trust to ensure the study met their standards for research governance.

Socio-demographic profile of sample
The socio-demographic characteristics of the 342 users of secondary mental health services selected for analyses are shown in Table 1. Just over half were female, with a wide age range. Almost two-thirds were white with the remaining sample evenly split between African descended, Asian and mixed/other ethnic classifications. Child abuse was identified in 109 (31.8%) cases. There were trends for different rates of child abuse experience according to service users' gender and ethnicity, reflecting (marginally) elevated proportions in females and in those individuals identifying as white. Almost half of psychiatric service users with experience of childhood maltreatment experienced further victimization (Table 1); rates of domestic and community violence victimization were significantly higher for service users with experience of childhood maltreatment (OR = 2.75, CI = 1.71,2.44 and OR = 2.57 CI = 1.46,4.54, respectively). Women were overwhelmingly more likely than men to have had experience of domestic violence (44.6% versus 19.2%; v 2 = 25.29, P < 0.001, OR = 3.39, CI = 2.08-5.52) but prevalence of community victimization was comparable between women and men (20.0% versus 15.0%, v 2 = 1.50, P = 0.222).

Child abuse frequencies
Physical abuse was the most frequently identified child abuse type from case notes (61 or 17.8%), followed by sexual abuse (56 or 16.4%) and then emotional/psychological abuse (43 or 12.6%). There was a notable overlap, with almost 40% (43/109 or 39.4%) of individuals with any identified child abuse experiencing more than one type. Not all case notes specified the age at which the abuse began; for those cases (N = 41) where age of abuse onset was noted, this ranged from 2 to 15 years of age.

Health, behavioural and social outcomes
Binge drinking, (other) substance abuse and self-harm were highly prevalent in the sample of secondary mental health service users, with rates of 45.3% (155), 42.7% (146) and 45.0% (154), respectively. Adverse social outcomes were also common; loss of housing was identified in 112 (32.7%) service users, loss of employment or education in 187 (54.7%) and loss of friends or family contact in 143 (41.8%).
Univariate associations showed that the prevalence of binge drinking, other substance abuse and self-harm were significantly elevated (approximately 60% across variables) in those service users who had experienced childhood abuse (Table 2), with the strongest relationships for childhood physical abuse (ORs range 1.81-2.79) and sexual abuse (ORs range 2.01-2.81). Univariate associations with social outcomes were more specific; loss of housing was significantly more frequent in those with childhood physical abuse only while loss of friends or family contact was (in separate analyses) strongly related to childhood physical abuse, sexual abuse and emotional/psychological abuse (ORs range 2.13-3.31).
Tables 3 and 4 illustrate multivariate logistic regression models for outcomes. Adult domestic violence, community violence and demographic factors (gender, age and ethnicity) were always included in models given their relevance to social, behavioural and health status, although to ensure model parsimony (and variable levels were sufficiently large), age was collapsed into three levels (18-35, 36-55, 56+ years) and ethnicity into BAME (Black, Asian and minority ethnic) and whiteunivariate analyses revealed there were no

THE ADVERSE EFFECT OF CHILDHOOD ABUSE IN ADULT PSYCHIATRIC SERVICE USERS
significant differences between collapsed age categories or among collapsed ethnic classifications on any outcome measure or prevalence of child abuse sub-types with the sole exception of ethnicity and self-harm (African descended = 17 or 41.5%; Asian = 7 or 20.6%; mixed/other = 18 or 54.5%; v 2 = 8.31, P = 0.016). Models included only those childhood abuse variables significantly associated with outcomes in univariate analyses (in the case of loss of employment/education this was none).
Results indicated that childhood sexual abuse was significantly associated with binge drinking, and selfharm in secondary mental health service users (Table 3), with increased odds (relative to those service users who did not experience sexual abuse as a child) of 2.15 and 2.86, respectively, while childhood physical abuse was significantly related to (other) substance abuse (OR = 1.88). Notably, being a victim of adult violence (domestic or non-domestic) was not significantly related to any health and behavioural indicator, although the associated risk of binge drinking was approximately halved in female service users and service users identifying as BAME (or, in other words, doubled in male service users and service users of white background).
A different pattern of associated factors emerged in multivariate models concerning social outcomes (Table 4). Domestic violence was significantly related to both loss of housing and loss of friends or family contact, more than doubling the associated risk of each. The odds of loss of housing and loss of employment/ education were more than halved in female service users (or, in other words, more than doubled in male service users). Notably, childhood sexual abuse remained significantly related to loss of friends or family contact in service users; the associated risk was more than doubled in those service users who had experienced child sexual abuse.

DISCUSSION
This retrospective analysis reports the findings of a randomly selected sample of 342 electronic care records of service users in secondary mental health services in South London, which examined the long-term impact of childhood abuse among adult psychiatric population. This study identifieswhere it has been notedthat many individuals who are using mental health care services have some history of childhood abuse and present with more risk or more severe behaviours. The rate of childhood abusethe total number of incidents describing different types of victimization in psychiatric service users' fileswas almost 32%, with almost 40% of service users with any identified child abuse experiencing more than one type. The electronic case notes examined for this study did not have a standard question (or validated tool) to capture information about childhood abuse, hence, practitioners did not routinely ask about this (Mantovani & Allen 2017). It is likely, therefore, that incidents of child abuse sub-type reported were under-recorded. Studies adopting different methodologies indicate higher proportion of child abuse in psychiatric populations (Duhig et al. 2015;NCK 2015). Underreporting and under-disclosing of violence and abuse within psychiatric settings (Hepworth & McGowan 2013;Read et al. 2018a) may indicate that clinical staff do not ask about child abuse (Read et al. 2018b). Underreporting may reflect a combination of lack of recognition, under-recording and an emphasis on medical and diagnostic treatment which Data regarding ethnicity were missing for 31 service users; hence the model includes 311 participants only; reference categories were 36-54 years for age, male for gender, and white for ethnicity; for each model, summary chi-square were derived from omnibus test of model coefficients; significant factors/model effects are highlighted in bold; for each model, multicollinearity was within acceptable limits (none of the independent variables had a standard error > 0.418) and there was no serious influence of outliers (maximum Cook's distance was 0.022). SE = standard error; OR = odd ratios, CI = confidence intervals; *P < 0.05, **P < 0.01, and ***P < 0.001. may prevent exploration of individual, social and environmental factors impacting on the illness (Trevillion et al. 2014). Not asking about childhood experiences of abuse may affect potential pathways to return to good mental health, and a greater contact with and lengthy stay in psychiatric hospitals (Hepworth & McGowan 2013). A systematic review of psychiatric nursing care for adult survivors of child abuse found that although nurses and nurse managers attach a child sexual abuse assessment to the psychiatric admission procedure, there is a discrepancy between the importance attributed to the assessment of a patient's abuse history and the actual actions taken to obtain information about that history (van der Zalm et al. 2014).
Childhood abuse was significantly associated with poor health, behavioural and social outcomes in adulthood for psychiatric service users. After controlling for socio-demographic variables and presence of (current) domestic violence, the odds of perpetrating binge drinking, carrying out self-harm and experiencing loss of friends and family contacts more than doubled in those service users who have experienced childhood sexual abuse, while there was a slightly less than twofold (1.88) associated risk of substance abuse in those Data regarding ethnicity were missing for 31 service users; hence the model includes 311 participants only; reference categories were 36-54 years for age, male for gender and white for ethnicity; for each model summary chi-square were derived from omnibus test of model coefficients; significant factors/model effects are highlighted in bold; for each model, multicollinearity was within acceptable limits (none of the independent variables had a standard error > 0.399) and there was no serious influence of outliers (maximum Cook's distance was 0.024). SE = standard error; OR = odd ratios, CI = confidence intervals. *P < 0.05, **P < 0.01, and ***P < 0.001. with experienced childhood physical abuse. A metaanalysis of longitudinal studies assessing the relationship between a history of child abuse and the development of substance (mi)use in adult life, indicated that a history of physical and sexual abuse in childhood, increased risk of developing substance (mis)use by 74% and 73% respectively (Halpern et al. 2018). Substance abuse and binge drinking are relatively common among individuals who have been exposed to traumatic events, especially those who have experienced childhood adversity (Anda et al. 2006;Banducci et al. 2014) with high rates of self-harm among adult survivors child abuse (Bellis et al. ,2014(Bellis et al. , , 2015Bolen, Winter & Hodges 2012), behaviours that can be viewed as coping responses or survival strategies (Alexander 2008).
Relative to those service users who did not experience domestic violence in adulthood, the experience of victimization was independently associated with both loss of housing and the loss of friends and family, with odds increasing by more than twofold for each outcome. Research shows that the very actions survivors take to achieve safety may trigger a wide range of negative consequences (Thomas, Goodman & Putnins 2015); on leaving the abusive relationship, survivors may lose financial stability, housing, friends and other resources (Rizo et al. 2020). Australian research (Fuller 2016) provides insights into the many ways in which child abuse can affect survivors' lives and their families. It shows the frequent pattern of relocation across different countries that survivors and their mother experience as a result of feeling unsafe following the abuse, and the ensuing familial difficulties in putting down roots and attain some form of stability.

Study's limitations
This study shows that in individuals using acute mental health services, experience of childhood abuse was associated with an increased risk of and vulnerability to psychiatric problems across their lifespan, as witnessed by their poor outcomes. The cross-sectional design of this study precludes definitive statements about causality with respect to observed relationships. Specifically, it is difficult to establish whether the experience of childhood abuse was key in shaping vulnerability to (heightened) psychiatric problems or reflects an increased tendency of mental health service users with more severe behavioural and/or social problems to disclose child abuse in clinical assessments. Alternatively, other adverse familial factors not directly addressed in this study but often related to the experience of childhood abuse (e.g. parental mental illness and/or problematic alcohol or drug use) (Fergusson et al. 2013) may be important drivers of behavioural and social dysfunction in (adult) users of psychiatric services. Further, retrospective reporting of abuse in case notes may be subject to recall bias and/or lead to ambiguity in temporal sequencing of variables. It is possible, for example, that patterns of self-harm and substance abuse behaviours emerged prior to experienced childhood abuse in some cases. Finally, as noted above, some service users who experienced childhood abuse may have not been identified, although this would likely weaken observed relationships between childhood abuse and behavioural outcomes, which were robust for sexual (binge drinking, self-harm) and physical abuse (substance abuse) in any case.

CONCLUSION
Despite the aforementioned limitations, the findings reported here suggest the relevance of adopting routine enquiries about child abuse in acute mental health settings, accompanied by policies and guidelines to provide a supportive culture for clinicians in undertaking this challenging work. Screening tools should be used in relation to clear effective interventions behind which adequate resources are placed. As the findings suggest, there is a need for more efforts in early interventions to address childhood risk factors for psychiatric disorders and revictimization in abused children. Hence, cross-cutting policies are necessary to recognize and assist families and their children at risk of early adversity as a crucial first step (Scottish Government 2018). Not all people who experience abuse or traumatic childhoods become entrapped in a downward spiral and develop psychiatric disorders (Marriott, Hamilton-Giachritsis & Harrop 2014). Therefore, it is important to understand why some children do well despite early adversity and identify which factors can mediate the effects of adversity and trauma experienced in childhood. This could inform policy and practice to help more children to flourish and reduce revictimization and the development of psychiatric disorders. Further research is needed to examine the structural social context and socioeconomic disadvantage in which children are exposed to abuse and adversity. By shedding light to this context, the focus is placed upon conditions which may be adverse for child well-being, and how these conditions come about. There is evidence highlighting that poverty and deprivation underlie the exposure of certain populations to these adversities (Metzler et al. 2017).

RELEVANCE FOR NURSING PRACTICE
Early identification and treatment of child abuse are important to minimize and prevent the long-term consequences of abuse (Department for Education (DfE) 2018; Haynes et al. 2015); therefore, nursing staff are in a unique position to help to identify and support the person with the associated challenges of child abuse. In this study, the onset of child abuse, the frequency and severity of abuse were incompletely described in most files scrutinized, which supports the claim that these parameters are still missing in current routine clinical assessments (Posner et al. 2008), raising concerns around mandating enquiry about trauma not ensuring adequate clinical evaluation. Although clinical staff in our study routinely inquired about suicidality and substance abuse and other social outcomes, only in a handful of cases did they relate these behaviours to the history of childhood abuse. The adoption of routine enquiry for childhood abuse by psychiatric nurses would remove potential barriers to identification and disclosure of childhood abuse (Ford et al. 2019) while at the same time enabling nurses to make sense of service users' presenting symptoms, responding appropriately to disclosures of abuse and making appropriate referral (Larkin & Simpson-Adkins 2018). Improving clinical nurses' responses to childhood trauma-related disorders may reduce service users' distress and begin to address ingrained symptomatology and receive pertinent health care. Policies ought to help create a culture of nursing care responsive to and restorative/healing for trauma service users placing an emphasis on the user's perspective. Key aspects that may enable a shift in culture in delivering mental health care services are the adoption of trauma-informed care (Ashmore, Spangaro & Mcnamara 2015), the retraining of nurses and the incorporation of mandatory enquiry, which ultimately would support survivors to recover. In addition, not only (TIC) training increases psychiatric nurses' comfort, knowledge and screening confidence (Kalmakis et al. 2018), which are needed to support and maintain the well-being of mental health nurses in clinical practice (Kennedy et al 2020), but also heighten their empathy on understanding their patients' background (Kia-Keating et al. 2019).

FUNDING
The study was funded by Wandsworth clinical commissioning group (CCG) Ref. No. 14-15.