Mechanisms to Enhance Resilience and Post-traumatic Growth in Residential Care: a Narrative Review

Residential care is primarily considered most appropriate for young people with complex needs, often due to multi-type traumas. Children in care are disproportionately disadvantaged, with children in residential care most vulnerable, which is why it is so important to understand mechanisms that support resilience and post-traumatic growth for this group of young people. This review aimed to advance our understanding of how interventions, reflections upon experience, and constructs of resilience can enhance developing resilience in children’s homes for young people in care. International quantitative and qualitative studies were sought to identify features and mechanisms of care that underpin developing resilience. Following a systematic search of six databases, 25 papers were selected for review, with a total sample of 3198 individuals up to the age of 30 years old who were either receiving residential care (N = 3037) or who were care leavers (N = 161). Themes from the quantitative studies and a narrative synthesis of qualitative studies were developed. Therapeutic mechanisms and processes to support the development of resilience included experiencing love and trust with staff in homes through therapeutic relationships, nurturing self-compassion, promoting self-value and self-belief, positive future thinking, problem-focused coping, school engagement, constructing a positive origin story, and positive visualisations of a stable future. Measures of resilience could more accurately reflect post-traumatic growth and potential for resilience development for this unique group of young people, which in turn could inform intervention design and evaluation. Measures appreciative of intrapersonal, relational, community and environmental factors could be particularly useful for intervention design.


Introduction
UNICEF state at least 2.7 million children are in residential care around the world, although recognise this is a conservative estimate. In the UK, children in care are young people who have been placed in the care of their local authority for over 24 h (NSPCC, 2021). In 2021, there were 80,080 children in care in England, most of whom (72%) were placed with a relative, friend or approved carer (National Statistics, 2020). Foster care has consistently been the placement of choice over the past decade, although out-of-home placement options, educational residencies and therapeutic communities are more suitable for some young people with specific needs. Residential care involves a non-family based, out-of-home care setting such as a secure unit, children's home, residential school or semi-independent living accommodation. The total number of children's homes in England increased by 7% to 2460 in March 2020 (Ofsted, 2020).

Key Points
• Children in care are disproportionately disadvantaged socially, educationally, in terms of their health and economically, with children in residential care most vulnerable. • This is the first mixed method review to explore resilience, resilience growth and post-traumatic growth for young people in residential care.
• Key mechanisms that support the development of resilience growth for these unique young people have been identified. • Resilience measures are potentially not suitably sensitive for care experienced youth. • Young mothers experience a double transition and often additional stigma, which leads them to be especially vulnerable socially, economically and in terms of their health. Residential care has been termed a 'last resort' (Holmes, Connolly, Mortimer, & Hevesi, 2018) and children's homes are primarily considered most appropriate for young people with complex needs, often due to multi-type traumas (Berridge, Biehal, & Henry, 2012).
Young people living in children's homes are considered to be some of the most vulnerable people in the care system and society as a whole (Parry, Williams, & Burbidge, 2021). Exposure to trauma is prevalent worldwide (Magrunder, McLaughlin, & Elmore Borbon, 2017), yet youth in this type of out-of-home placement have been found to display particularly high levels of trauma exposure (Brady & Caraway, 2002;Briggs, Greeson, Layne, Fairbank, Knoverek, & Pynoos, 2012), such as abuse and neglect, which are the main reasons children are taken into care (NSPCC, 2021). Trauma experiences in childhood have been linked to a range of neurobiological, psychosocial, and somatic consequences (Herzog & Schmahl, 2018). The impact of trauma is dependent on various factors, such as the type of trauma experienced, duration of the event, and protective factors present (Gahleitner, Frank, Gerlich, Hinterwallner, Schenider, & Radler, 2018) and young people who have experienced significant trauma often appear to demonstrate incredible resilience (Quisenberry & Foltz, 2013) and even post-traumatic growth (PTG; Milam, Ritt-Olson, & Unger, 2004). However, too often, children in care are expected to cope with too much too soon for their developmental stage and personal resources, which can reduce opportunities for resilience (Parry & Weatherhead, 2014;Stein, 2006). Children in care are disproportionately disadvantaged, with children in residential care most vulnerable (Department for Education, 2021), which is why it is so important to understand mechanisms that support resilience and PTG.
Resilience is a dynamic developmental process through which individual, contextual and environmental factors influence the ability of a person to cope and adapt (Pinheiro, Magalhães, & Baptista, 2021), whereas PTG is considered to be the ability to move beyond pre-trauma levels of functioning (Ogińska-Bulik, 2015), absorbing learning with enhanced agency. Although the two concepts are different, both are associated with recovery from traumatic events and are considered to be particularly relevant to highly traumatised youth in residential care (Lou, Taylor, & Di Folco, 2018;Masoom Ali et al., 2020). It is important to attend to resilience within a field dominated by trauma, adversity and hardship because "resilience information in addition to adverse experiences can increase the richness of studies measuring the impact of program interventions" (Leitch, 2017, p.5). Through exploring the role of resilience separately to trauma-responses and coping, features that enhance resilience and thriving can also be prioritised within interventions for children in care.
A relatively recent systematic review by Lou et al. (Lou et al., 2018) explored resilience for children in residential care. Promoting interpersonal relationships, developing a future focus, and enhancing motivation were particularly salient amongst the resilience factors identified. Further, the authors found that adolescents in residential care displayed greater vulnerability and developmental challenges compared to their peers. Importantly, higher resilience levels were found to be related to better developmental outcomes, highlighting resilience as an important mechanism for improving outcomes for children. There has been limited literature surrounding PTG in youth within residential care and no review exploring this area to date.
Building upon the work of Lou et al. (2018) and the recent review of Pinheiro et al. (2021), the current review specifically focused upon intervention studies published between 2017 and 2021 that explore resilience and PTG for young people (< 30 years old) with experience of residential care. Lou et al. and Pinheiro et al. recommend future research includes qualitative and mixed methods to elicit the views of care leavers (people with experience of residential care), to gain a better sense of the long-term influences and in-depth understanding of processes related to resilience-focused interventions. This review aims to update and synthesise our understanding of how the development of resilience and PTG can be supported through childhood, adolescence and emerging adulthood through identifying mechanisms of change. The review includes quantitative and qualitative intervention studies that either report on intervention outcomes or capture reflections on engagement with intervention practices, offering nuanced insights into mechanisms of change and therapeutic processes underpinning resilience and PTG. In summary, this review contributes a comprehensive and timely understanding surrounding how resilience and PTG can be enhanced for young people in children's homes to inform sector reform and improve outcomes for young people.

Eligibility Criteria
Eligibility was assessed in two stages: stage 1 involved title and abstract screening of the papers identified in the initial scoping or database searches, whereas stage 2 involved fulltext screening of the papers that were identified as being potentially relevant in stage 1. Informed by Lou and et al., (2018 p.84), the inclusion criteria for articles were to: focus on a population aged 30 years old or younger who had experience living in a residential care setting (residential care or treatment) either at the time of the study or prior to it. Residential care settings include any non-family-based 1 3 residential childcare setting where staff do not appear to live in; papers including a qualitative, quantitative or mixed methods empirical design, methods section, and results section; authors investigate resilience or post-traumatic growth by either measuring it directly (i.e. via a resilience or posttraumatic growth scale) or by defining it then measuring it indirectly (e.g. via qualitative techniques, scales measuring positive characteristic(s) representing resilience); published between 2017 and 2021; written in the English Language and full-text available; published and unpublished papers including dissertations. The exclusion criteria were as follows: focus on a population with experience in a residential care setting that is only for individuals with moderate or severe learning difficulties; authors define or measure resilience or post-traumatic growth with only negative outcomes such as recidivism or trauma symptoms; numerous populations investigated but do not sufficiently distinguish between the residential care setting population and other populations within the results section. The current review included papers focusing on a population up to the age of 30 years rather than 18 years to capture as many papers involving care leavers as possible to include their valuable reflections on their experiences as adolescents and emerging adults leaving care. Lou et al. conducted searches up to November 2017; therefore, the inclusion criteria in this review required papers to be published between 2017 and 2021 to ensure that only the most recent papers in this area were included. The criteria relating to conceptualisation and measurement (inclusion criteria C and exclusion criteria B) were devised to ensure it related to PTG rather than solely resilience, as this review also aimed to explore PTG for people with experience of residential childcare. Consequently, the current review builds upon and expands the original review of Lou et al. SC and SP reviewed the final set of papers for review, conducted independent scoping searches to check for updates, and concluded the searches in July 2021.

Search Strategy
The review of Lou et al. (2018) informed the initial search strategy and process of selection for data extraction. The review was prospectively registered on PROSPERO. Two sets of terms were used to search six databases between February and July 2021: Medline, ASSIA (Applied Social Sciences Index and Abstracts), PsycINFO, YourJournals@ OVID, Scopus, and OpenGrey. The first four databases were used in Lou et al. (2018) and two additional databases were also searched (Scopus, OpenGrey) to help ensure all relevant papers were identified. The first set of terms were as follows: (residential OR accommodated) AND (resilience OR protective) AND (child* OR teen* OR youth OR young) and the second set of terms were: (residential OR accommodated) AND "post traumatic growth" AND (child* OR teen* OR youth OR young). All of the searches were limited to papers written in the English Language and papers published between 2017 and 2021.

Data Items
In total, 25 papers met the criteria and were subsequently included for review. Supplementary File One displays the data extracted from the final papers (see also Fig. 1).

Characteristics
Fourteen studies explored resilience using a quantitative design and the remaining ten studies investigated resilience using a qualitative design. Nine of the quantitative studies and all of the qualitative studies appear to be cross-sectional; however, only five papers explicitly state this in their paper (Chulakarn & Chaimongkol, 2021;Greenbaum & Javdani, 2017;Isakov & Hrncic, 2021;Segura, Pereda, Guilera, & Hamby, 2017;Suárez-Soto, Pereda, & Guilera, 2019). Four of the remaining studies used a pre-post design to investigate resilience and each measured resilience before and after a different type of intervention/programme. In pre-post design studies, resilience was measured before and after a "Girls Group" psychoeducational programme (Berry, Tully, & Egan, 2017), a group-based mental health intervention (Garoff, Kangaslampi, & Peltonen, 2019), a "Strong Teens" programme (Marvin, Caldarella, Young, & Young, 2017) and spiritual intervention programmes (Pandya, 2018). Two studies examined resilience using a longitudinal approach, one of the studies looked at resilience at the time of disengagement from care and 1 year following this (Van Breda & Dickens, 2017) whereas the other study used a pre-post design to explore resilience before and 1 year after spiritual intervention programmes (Pandya, 2018). One study adopted a quantitative case study design that explored resilience in one individual (Berry et al., 2017). However, all of the other resilience studies involved more than one individual in their sample. Like most of the resilience studies, the study investigating solely PTG (Masoom Ali et al., 2020) used quantitative methods and appears to be cross-sectional.
Fifteen studies were carried out in developed countries (USA, UK, Canada, Finland, Israel, Spain, South Africa), six studies were conducted in developing countries (Thailand, Ghana, India, Indonesia, Nigeria) and two studies took place in countries undergoing economic transition (Serbia, Slovakia; Central Intelligence Agency, n.d.). One study was 1 3 carried out in numerous countries with various different economies (Pandya, 2018). The PTG study (Masoom Ali et al., 2020) was carried out in Pakistan, which is classified as a developing country (Central Intelligence Agency, n.d.).

Measurement
Various methods and instruments were used to measure resilience. Most of the qualitative studies used semi-structured interviews to explore resilience (Frensch, Ashbourne, MacLeod, Bartlett, & Preyde, 2020;Frimpong-Manso, 2018;Lukšík, 2018;Schofield, Larsson, & Ward, 2017;Sekibo, 2020;Sulimani-Aidan, 2018). Other approaches used focus groups (Bermea et al., 2019;Mishra & Sondhi, 2019), unstructured interviews (Van Breda & Hlungwani, 2019) and participant drawings (Mishra & Sondhi, 2019) were used as well. One qualitative study used a life-story interview (Refaeli, 2017) and this interview appears to be semi-structured, like the majority of interviews used in qualitative studies within this review. A range of instruments were used to measure resilience in the quantitative studies (see Table 1 for a complete overview). All of the instruments used appear to be self-report aside from the Social Emotional Assets and Resilience Scale (SEARS; Marvin et al., 2017), which is a cross-informant measure involving both a self-report form and a teacher/adult report form. Although no instrument was consistently used across all of the quantitative studies to measure resilience, some instruments were used in more than one study. For example, the Adolescent Resilience Questionnaire (ARQ; 24) was used in two studies (Segura et al., 2017;Suárez-Soto et al., 2019) and versions of the Resilience Scale (Isakov & Hrncic, 2021) were used in two studies as well (Isakov & Hrncic, 2021;Pandya, 2018). The majority of studies used established instruments created by other authors; however, some authors did use an instrument that they had helped to develop (Van Breda, 2017;Van Breda & Dickens, 2017) or an instrument that was still undergoing evaluation (Garoff et al., 2019) to examine resilience. The instrument used by Masoom et al. was the Posttraumatic Growth Inventory for Children-Revised (PTGI-C-R; Masoom Ali et al., 2020), which was also self-report.

Analytic Approach to Qualitative Synthesis
A narrative approach was adopted to synthesise findings from the qualitative papers (Marriott et al., 2014) as it is widely accepted that qualitative studies offer a depth of personal insight that other methods do not (Bogar & Hulse-Killacky, 2006). Therefore, a retelling of original narratives provides an additional degree of nuance and perspective to the understanding of resilience for this under-researched group. The narrative synthesis aimed to identify and summarise prior research in relation to the role of resilience in residential care across nine articles. A narrative synthesis provided an opportunity to explore similarities, differences, patterns and relationships within the combined texts to inform the development of nuanced and novel insights (Lisy & Porritt, 2016). With the objective of interpreting and  Semi-structured interviews • Young people described hobbies (e.g. singing, karate, being with animals) as helpful for emotional regulation and focusing on positive thoughts Relationships with family and friends sought out, mothers most common source of support, positive nature of grandparent relationships and learning skills, 'process perspective' upon familial relationships • Emotional resilience supported aspirational educational and employment goals, which were nurtured through 'positive and close relationships with immediate and extended family, instrumental, and emotional support from parents, engaging personal hobbies, and time with friends' • Overall, a sense of accomplishment supported positive sense of self, which acted as a resiliency growth factor for relationships, aspirations and endeavours, which nurtured success Semi-structured interviews • Resilience facilitators/factors: social support, self-reliance, informal social networks, borrowing from fellow care leavers, sharing accommodation with peer care leavers and intimate partners (mostly young women), one's ability, to "encourage yourself" and "be determined", problem-solving skills Garoff  The Brief Resilience Scale (BRS as cited in Greenbaum & Javdani, 2017) • In comparison to CSG participants, WRITE ON participants showed a significant increase in positive mental health attributes • Increase found to be driven by a significant increase in resilience rather than other variables (self-esteem, positive affect) • Increase remained present even when the number of treatment minutes (treatment dosage) was controlled for Isakov and Hrncic (2021), Serbia • Youth in foster care displayed significantly higher resilience scores than youth in residential care, but no significant differences found between youth in residential care and youth in kinship care Lukšík (2018) • At week 6 of the intervention, no significant differences found between resilience in the treatment group and the wait-list group based on self-report and supervisor ratings • Based on the group therapist ratings, the treatment group displayed significantly higher resilience compared to the wait-list group at week 6 of the intervention • After the full 12-week intervention, resilience significantly increased in both groups based on the group therapist and supervisor ratings, but only significantly increased in the wait-list group based on the self-report ratings • Pre-programme (phase 1): range = 8-13 years; male = 58% (980) • Post-programme (phase 2): range = 9-14 years The 25-item Resilience Scale (Wagnild & Young, 1993 as cited in Pandya, 2018) • Average resilience scores at phase 2 (postprogramme) were significantly higher than those at phase 1 (pre-programme) • At both phases, boys displayed significantly higher resilience scores compared to girls • Younger children aged 8-10 years displayed significantly higher resilience scores at phase 1 compared to older children aged 11-13 years • A significant interaction effect was also found between sex and age at phase 1 indicating that younger boys displayed higher resilience scores • At phase 2, children who said they had learnt about positive thinking-futuristic attitude had significantly higher resilience scores than those who said they had learnt about unconditional love and forgivenessletting go • Children who self-practised had significantly higher average resilience scores than those who did not at phase 2 Refaeli (2017) synthesising the findings of the selected studies, the articles were reviewed narratively to 'tell a story' informed by the stories of the original participants, interpreted and reported by the authors of the included studies. Narratives were explicitly mentioned in six of the nine articles reviewed, providing strong epistemological narratives upon which to construct an emancipatory narrative of resilience and PTG for the review to inform interpretations and the discussion (Frimpong-Manso, 2018;Lukšík, 2018;Refaeli, 2017;Schofield et al., 2017;Sulimani-Aidan, 2018;Van Breda & Hlungwani, 2019).

Study Selection
All but one of the studies involving care leavers (Van Breda & Dickens, 2017) and eight of the studies involving individuals who were receiving residential care at the time of the study had a sample size of less than 50 within their final analysis. One study included 30 social workers rather than individuals with residential care experience in their sample (Sulimani-Aidan, 2018), although was included because of their in-depth reflections upon directly caring for young people in residential care and because the voices of frontline workers are also scarce within this field of research.
Most studies included individuals of various different ages in their sample, with the majority including either children and adolescents (18 years or younger) or adolescents and young adults (19-30 years) rather than just one age group (e.g. adolescents, young adults). The three studies that included only young adults in their sample involved care leavers rather than individuals who were currently receiving residential care (

Study Characteristics
Twenty-four studies in this review explored resilience and one study explicitly explored PTG (Masoom Ali et al., 2020), although there were features of PTG within many of the studies reviewed, as discussed. A comprehensive overview is available in Table 1.

Sample Characteristics
This review identified a total sample of n = 3198 individuals up to the age of 30 years old who were either receiving residential care at the time of the study (n = 3037 across seventeen studies) or had previous experience of receiving residential care (care leavers; n = 161 across seven studies). Sample sizes ranged from n = 1 to n = 1689. The review also includes reflections on resilience from 30 social workers based on their experiences with children in care (age range 8-17 years old; Sulimani-Aidan, 2018).

Victimisation, Resilience and PTG
Young people with residential care experience were found to be particularly vulnerable to victimisation. This finding bears out across the literature, with children in residential care settings recognised as at increased risk for childhood sexual abuse (Roache & McSherry, 2021).  et al., 2000) has also been employed to examine total emotional and behavioural difficulties, Masoom Ali et al. (2020) reported that 43.94% of their sample scored within the high/ very high range and Garoff et al. (Garoff et al., 2019) found that over 30% of their sample scored within the 'abnormal' range before and after a mental health intervention. Further evidence of vulnerability to mental health difficulties was provided by Marvin et al. (Marvin et al., 2017) who noted how all of the youth at their study site had a clinically diagnosed psychiatric disorder. Despite this evident vulnerability, the young people with residential care experience were also found to display resilience and PTG on scales measuring the two concepts. Emergent resilience was found for certain groups of youth in residential care (boys, younger children; and there was evidence of youth in residential care displaying moderate (Suárez-Soto et al., 2019) and high levels of resilience (Chulakarn & Chaimongkol, 2021), along with high rates of PTG (Masoom Ali et al., 2020). Alongside this, the resilience of youth in residential care was found to increase post-intervention in all but one of the studies employing a pre-post design (Garoff et al., 2019). Two papers compared the resilience of youth in residential care to the resilience of youth in various other settings (Isakov & Hrncic, 2021;Van Breda & Dickens, 2017). Isakov and Hrncic examined the resilience of youth in alternative care settings and found that youth in residential care displayed significantly lower levels of resilience in comparison to youth in foster care, but no significant differences in resilience were found when compared to youth in kinship care. Van Breda (Van Breda, 2017) explored the resilience of youth in two residential care settings and five schools. These authors found that youth in one residential setting (for young people in multiple provinces) displayed the highest level of resilience overall, but youth in the other residential setting (for refugee and unaccompanied children) displayed the third lowest level of resilience.

Conceptualisation
All the included quantitative papers used measures of resilience. Mazur (2018) measured three variables: self-compassion, compassion towards others, motivation to change. Most authors also defined resilience directly in their paper using an array of linguistic terms, cumulatively describing resilience as an active response to a challenging experience that required flexibility and action, with the terms 'adapt/adaptation' and 'adversity' being mentioned most frequently. Further, synonyms such as 'adjusting well', 'managing', and 'cope positively' were used to define resilience to 'difficult experiences', 'trauma', and 'stress'. Most authors did not explicitly state that they adopted an ecological conceptualisation of resilience; however, several papers used measures that examined resilience in internal and external domains, thus implying that this conceptualisation had been adopted (Chulakarn & Chaimongkol, 2021;Segura et al., 2017;Suárez-Soto et al., 2019;Van Breda & Dickens, 2017). Masoom Ali et al. (2020) used the Posttraumatic Growth Inventory for Children-Revised (PTGI-C-R; (49)) instrument to examine PTG directly through five different subscales: new possibilities, relating to others, personal strength, appreciation of life, and spiritual change. These authors defined PTG directly as a '…potentially positive psychological outcome that transcends pre-trauma functioning ' (2020, p.64). In terms of interpretation of experience, it is the 'positive psychological outcome' within PTG, rather than the ability to manage and cope, that appears to most starkly differentiate between resilience within the papers and PTG.

Quantitative Thematic Summary
Numerous variables were mentioned within the quantitative findings regarding resilience and PTG in relation to 23 domains, which have been synthesised according to four themes. The four themes reflect the groupings of content according to overarching attributes.

Demographic Theme
The Posttraumatic Growth Inventory assesses positive outcomes reported by people who have experienced traumatic events. The 21-item scale includes factors of New Possibilities, Relating to Others, Personal Strength, Spiritual Change, and Appreciation of Life (see Tedeschi & Calhoun, 1996). Overall, girls showed significantly higher total growth and 'New Possibilities' PTGI subscale scores compared to boys (Masoom Ali et al., 2020), although boys displayed significantly higher resilience scores compared to girls in Pandya (2018), although the mean resilience score for girls was significantly higher than that of boys in Sobana (2018), which raises questions as to gender sensitivity in terms of language and item structure across measures. It was also found that girls demonstrated significant inverse associations with self, family, and community resilience domains (Segura et al., 2017) and girls displayed higher scores in various resilience characteristics (communication, empathy, help seeking, goals for future and aspirations) but boys displayed higher self-esteem compared to girls, indicating the importance of meta-and systemic influences for girls.
There were further mixed results in terms of the measurement of resilience and PTG in terms of age, as younger children (aged 8-10 years) displayed significantly higher resilience scores at phase 1 compared to older children aged 11-13 years (Pandya, 2018), although age negatively correlated with 'Personal Strength' PTGI subscale scores (Masoom Ali et al., 2020). Older children displayed higher scores in various resilience characteristics (communication, self-esteem, empathy, goals for future and aspirations) but younger children displayed higher help seeking. The mean resilience score of older children was significantly higher than that of younger children (Sobana, 2018).
With regard to 'Conduct Problems' as measured by the SDQ, subscale scores were significantly negatively correlated with 'Relating to Others' and PTGI subscale scores (Masoom Ali et al., 2020). The 'Relating to Others' PTGI subscale scores significantly negatively predicted by 'Conduct Problems' from SDQ subscale scores (Masoom Ali et al., 2020), although it is worthy of note that the SDQ items 1 3 and calculations that lead to the sub-scales have not been tailored for children in care who have experienced multitype traumas.
There was a negative association found between a history of sexual abuse and self-compassion, which could have important implications for psychological support. Interestingly, a positive association was found between a history of sexual abuse and compassion towards others (Mazur, 2018).

Intrapersonal Theme
A range of intrapersonal factors were positively associated with resilience, including self-concept (Chulakarn & Chaimongkol, 2021), self-esteem, positive affect, shame, guilt and negative affect (Greenbaum & Javdani, 2017). Further, extraversion explained an 8.9% variance in resilience and was found to have a significant positive effect on resilience (Nurani et al., 2018). Additionally, children who said they had learnt about a positive thinking-futuristic attitude had significantly higher resilience scores than those who said they had learnt about unconditional love and forgiveness (Pandya, 2018), which again appears to be an important mechanism for nurturing resilience. All of the resilience domains (self, family, peers, school, and community) except residential care support showed a significant inverse association with either internalising symptoms, externalising symptoms, or both (Segura et al., 2017). Importantly, lifetime victimisation was found to have a significant inverse association with self, family, and community resilience domains (Segura et al., 2017) and young people reporting a presence of suicidal phenomena within the last 6 months displayed lower scores in all resilience factors (Suárez-Soto et al., 2019).

Relational Theme
Overall, there was an indirect effect on resilience via school engagement and self-concept (Chulakarn & Chaimongkol, 2021). Relationships were also shown to be a protective factor as the 'Relating to Others' item on the PTGI subscale scores significantly negatively predicted by 'Intrusion' (sum of items 1 + 4 + 8 + 9 on the CRIES-13 subscales relating to intrusive thoughts; Masoom Ali et al., 2020). Similarly, 'arousal' significantly positively predicted 'Relating to Others' PTGI subscale scores and significantly negatively predicted 'Personal Growth' PTGI subscale scores (Masoom Ali et al., 2020), highlighting the need for interventions to attend to settling arousal through mechanisms such as grounding techniques.

Strategies Theme
Overall, both problem-focused coping (Chulakarn & Chaimongkol, 2021) and extraversion*problem-focused coping variable were found to have a significant positive effect on resilience (Nurani et al., 2018). School engagement was also found to have a significant direct positive effect on resilience (Chulakarn & Chaimongkol, 2021). In terms of independent living outcomes, higher resilience levels at the time of disengagement from care were found to be positively associated with higher levels of independent living outcomes 1 year after leaving care (Van Breda & Dickens, 2017).

Narrative Synthesis
The findings and results section of each article was read for relevant data, which was then coded for concepts, themes or narrative pathways. These collections of data were then reinterpreted and synthesised to form a reconstructed narrative of resilience for young people with experience of residential care across two narrative layers, which is outlined in Fig. 2. Tabulated summaries can be helpful in the analysis and reporting of a narrative summary to organise data and add transparency to the process (Baños et al., 2017), so a table of concepts, thematic interpretations and synthesis interpretations was developed and discussed within the team.

Narrative Layer 1: Experiencing the Development, Maintenance and Growth of Resilience Through Relationships
A strong narrative layer within the analysis highlighted the importance of relationships to nurture resilience and to help it grow. Learning love and trust through experiencing positive care with staff in homes also seemed essential to nurture hopefulness, a recognised characteristic of resilience (Frensch et al., 2020;Schofield et al., 2017): "What makes me resilient is that I know there is hope for the future" (Sekibo, 2020). Some authors used metaphor to describe "a wall of bricks that they (young people) build out of all the meaningful adult figures they met", which could lead to the development of "strong interpersonal relationships" through friendship, care or mentorship that fostered young people's belief in themselves and others (Sulimani-Aidan, 2018).
This positive sense of self and others appeared pivotal in terms of how able young people felt to ask for help when they needed it and aspire to fulfilling future goals (Frensch et al., 2020). Sulimani-Aidan (2018) illustrated a bi-directional process through which a young person asking for help could enhance a therapeutic relationship because the therapist then felt "more connected and empathic" or perhaps needed and invited to connect, which meant the therapist was more likely to nurture a sense of connectedness and 1 3 resilience for the young person. Further, young people who were able to form social networks with friends, colleagues and former carers from the residential homes found innovative ways to support one another to nurture their resilience and reduce adversities, for example, sharing homes to reduce costs (Van Breda & Hlungwani, 2019), returning to the care facility for short respite when needed (Mishra & Sondhi, 2019), or reconnecting with extended family to pursue future goals (Frensch et al., 2020).
Relationships with peers were also important to foster a sense of self-belief and resilience. For example, when young people witnessed 'similar others' in residential care achieving and succeeding, there was a sense of "I felt if they can, so can I" (Mishra & Sondhi, 2019). Young people's narratives across the papers also emphasised the importance of socialising with peers who were not care experienced. However, their ability to integrate with peers in the community was constrained by environmental, socioeconomic and interpersonal barriers, a lack of acceptance from people and lack of support from systems. Within these hostile community environments, marriage, faith groups or churches offered a solution to financial hardship, danger and loneliness (Sekibo, 2020). Membership to a supportive system by way of a marriage or church appeared to offer similar benefits in terms of friendship and a sense of belonging: "is like marriage. It's a fresh start. They (church members) don't care about where I come from. I feel I have a family" (Frimpong-Manso, 2018). Having a connection to another person or community offered some protection against the many risks posed by systems that did not suitably integrate to form a safety net. Developing a safety net through relationships meant ongoing stressors could be reduced, which eased the transition from care and provided more opportunities to continue building a sense of resilience, rather than continually struggling with overwhelming adversities. However, although the young people recognised the need for the informal support relationships could offer, some also struggled to form meaningful supportive equal relationships with non-care experienced peers. There were also stories of not knowing what to look for or not recognising support when offered, which meant opportunities could be missed (Refaeli, 2017).
Overall, resilience largely centred around personal agency: agency to help oneself and agency to seek help when needed. To be a 'survivor' in their stories, young people needed the ability to connect with others through relationships and to embrace personal agency. There were ongoing risk factors to resilience for many of the young people, especially the threat of loss of employment and relationships, which seemed to increase a sense of helplessness and victimhood as resources for agency depleted. However, for young people who had positive stable relationships to fall back on, and those young people who talked about residential Health Housing Employment RelaƟonships Importance of strategic relaƟonships, knowing what help to look for, recognising opportuniƟes for help and support, developing collaboraƟons (e.g. to borrow from one another, living together to share costs and enhance safety), to witness similar others succeed. 'Survivors' -ability to rely on personal agency and connect with others.
Marriage and connecƟon to a church -offering friendship, belonging, a fresh start, family and could reduce financial hardships, loneliness and danger PosiƟve narraƟves of abiliƟes, strengths and achievements to develop a dominant narraƟve of resilience -romanƟc self-narraƟve as a hero, survivor, heroes in their lives -reduces impact of negaƟve experiences Care leavers as young mothers had double transiƟon and could face double sƟgma Experiences of rejecƟon can help painful memories resurface The absence of care and support can also be experienced as further rejecƟon People not workers -staff as friends/family Safe people who want to help Offer help when needed ProacƟvely offer opportuniƟes (e.g. housing, employment) DeterminaƟon, self-reliance, ability to ask for help when needed, self-aware, higher selfesteem, believe they are worthy, believe they can cope and have helpful coping strategies ResidenƟal homes that provide training and preparaƟon as well as somewhere to live Opportunity to keep in touch for care leavers and return for short period when need a rest Job prospects improve when have longer to transiƟon with support Experience loving relaƟonships and grow within them, enhancing self-esteem, sense of worth and resilience Resilience Fig. 2 Narrative summary visual representation carers as 'people' not 'workers', and their residences as 'at home' with 'family' or 'friends', they could find resources to rely on when their personal resilience was under attack. Importantly, when carers demonstrated their commitment to 'not giving up', modelling resilience, this helped fortify the young people to maintain agency and therefore their own resilience (Schofield et al., 2017). One group who appeared particularly vulnerable were care leavers who were also young mothers. Their double transition out of care and into a caring role of their own highlighted how many barriers care leavers may have to contend with. As they emerged from one care system to becoming carers with little support from other systems, they often experienced hostile social and financial environments with many barriers and few facilitators. More than any other group, young mothers described the extent of the stigma they experienced, whereby adults around them would actively prevent their own children from socialising with them, leaving the young mothers and their babies with depleted social support networks and few relationships to lean on or feel safe within (Bermea et al., 2019).
Finally, one means of talking about resilience was that of becoming a 'survivor', which could come about through relationships, through enhanced self-reliance, or both; "Becoming a survivor for these young people was about discovering their own value through caring relationships" (Schofield et al., 2017). The environment and culture of the residential setting also influenced emerging resilience and coping strategies, with multiple strategies often needed to cope with multi-layered adversities. Some adversities could be resolved with help from others, but some young people could struggle to first identify what help was needed and to know where to look. There was also the risk of potential exploitation if the wrong person was trusted, which added high stakes to help seeking (Lukšík, 2018). Young people also had to find survival strategies in relation to their birth families, which offered roots and risks. An example of this was to find cognitive ways to align with birth the family, such as through pursuing similar employment or occupation pathway, rather than becoming personally or emotionally integrated into the birth family after leaving care (Refaeli, 2017).
In summary, whilst relationships could offer protection, support and a safe space to heal and grow, there were also risks, which some young people struggled to navigate without the support of a trustworthy other. Positive supportive relationships were necessary for resilience to develop and flourish, although there could be many systemic barriers to developing positive relationships outside of care. Falling between the gaps left behind by support systems that did not appropriately integrate to offer a safety net for care leavers was a common theme. A process of a downwards spiral was described, whereby a lack of employment meant people were forced to live in cheaper accommodation, often in dangerous areas and without basic amenities (Frimpong-Manso, 2018;Sekibo, 2020;Van Breda & Hlungwani, 2019). This spiral would negatively impact upon their wellbeing and deplete their fledgling resilience. This cycle was further impacted through a reduction in social support as their location could act as a barrier for connection, for example, "who would visit me here?" (Frimpong-Manso, 2018).

Narrative Layer 2: Conceptualising and Achieving Resilience
Across the papers reviewed, there were narrative layers referring to what facilitated resilience and what could act as barriers. A range of strategies were employed to foster resilience, such as narrating a positive origin story. These young people viewed themselves as survivors focussing only on successes, developing romantic saturated narratives of overcoming and resilience, which squeezed out space for negative events (e.g. being let down; Refaeli, 2017). Carers also described how important it was to help young people 'dig deep' to form a positive self-narrative, which could counterbalance low self-esteem and promote self-awareness of thoughts and feelings (Sulimani-Aidan, 2018).
Strategies and facilitators for resilience included the ability to form social networks that advanced accommodation and employment opportunities. A young person's judgement around situations and ability to interpret these situations to consider outcomes was an important cognitive process in the development and maintenance of resilience. Another key factor was the access to resources and ability to engage with those resources. Positive visualisations of a more stable future also nurtured hopefulness (Van Breda & Hlungwani, 2019).
Personal qualities related to self-esteem were also seen as essential for the development of resilience, such as selfvalue and self-belief. A 'resilient child' was said to have 'higher self-awareness', a 'sense of humour, creativity, emotional intelligence, and friendliness' (Sulimani-Aidan, 2018). Overall, there was little discussion of a young person's learnt skills, talents or ambition from carers, although personal abilities to "encourage yourself", "be determined" and develop problem-solving skills were mentioned (Frimpong-Manso, 2018).
Systemic thoughts around resiliency factors were discussed in terms of social support, informal social networks, 'borrowing' from fellow care leavers, sharing accommodation with peer care leavers and intimate partners (Frensch et al., 2020;mostly young women, Frimpong-Manso, 2018). The role of religion and cultural influences was also discussed, whereby figures who inspired young people within religions and global capitalist cultures provided very different stimuli but seemed to have a similar impact on 1 3 inspiration and aspiration (Mishra & Sondhi, 2019). However, the process of achieving strategies for resilience could be perturbed through socioeconomic hardships as young care leavers simply had fewer resources, which reduced their pathways and options. For example, some care leavers described the support systems in place as being overwhelming and bureaucratic. Rejections from support services could also reinforce feelings of victimisation and encourage painful memories of abandonment to resurface (Refaeli, 2017). An absence of immediate care could be experienced as rejection, which could mean that even when services were not actively working against them, they could be experienced as rejecting (Schofield et al., 2017).
Across the narratives of resilience were some clues that resilience could take different forms at different times. Some of these processes were clear, such as the aforementioned ability to know when to rely on oneself and when to seek support from trustworthy others. However, it was not always transparent within some narrative layers as to what was steadfast resilience and what was the acceptance of hardship, or whether these share a common quality of being able to persevere through immense hardships. There were also clues that some settings have greater potential to nurture future resilience than others. For example, schools and educational residential centres seemed to offer routes to resilience for the future, such as training for future employment and preparation for future transitions (Mishra & Sondhi, 2019), which could also enhance confidence for the future. There was also mention of the importance of continuity, whereby care leavers could return to a residential facility (Frimpong-Manso, 2018) or call previous carers when needed (Mishra & Sondhi, 2019), emphasising the need for safe-base support in the maintenance of resilience.

Discussion
This review aimed to advance understanding of factors that support resilience, resilience growth and post-traumatic growth for young people with experience of residential care. Overall, there were significant mental health needs across the sample, with PTSD (Masoom Ali et al., 2020, Garoff et al., 2019, depression (Pandya, 2018) and suicidality (Suárez-Soto et al., 2019) particularly evident. In terms of resilience development, younger children felt more able to seek help, although it was older children who demonstrated higher levels of measurable resilience (Chulakarn & Chaimongkol, 2021;Suárez-Soto et al., 2019) and PTG (Masoom Ali et al., 2020). Resilience as measured was found to increase post-intervention in all but one of the studies employing a pre-post design (Garoff et al., 2019). However, future research could include reflective discussion with young people following interventions to provide qualitative data as to how change occurred and what young people found to be most helpful. Specifically, further longitudinal research with care leavers to explore how resilience and PTG continue to develop or deplete over life course or in response to specific life challenges would be beneficial to inform early intervention opportunities.
Three core protective systems underpin resilience for children: individual capacities, attachment to a nurturing caregiver, and a protective community (Dermody et al., 2018). As core service systems, residential care settings are well placed to develop and pilot innovations in practice to nurture children and young people and provide traumainformed care within a protective community. Ecologically, the environment and culture of residential settings also influenced emerging resilience and coping strategies, with multiple strategies often needed to cope with multi-layered adversities.
Relationally, the experience of love and trust with staff in homes seemed essential to nurture hopefulness (Schofield et al., 2017), which can also underpin resilience through the development and witnessing of agency and pathways to achieve one's hopes (Snyder, 2002). Within the first narrative layer, there was an interesting description of a bi-directional process through which help seeking could enhance a therapeutic relationship when the practitioner felt connected through being needed (Sulimani-Aidan, 2018). However, relationships with adults could also pose risks though the threat of exploitation, the roots and risks associated with birth families, and high stakes in terms of decision making within relationships (e.g. marriage). Moreover, some of our findings (Mishra & Sondhi, 2019;Van Breda & Hlungwani, 2019) highlight the vitality of extra-institutional social networks and socialisation outside of formal 'cared-for' contexts. Existing research exploring the intersection between care-leaver experiences and their development of social capital through community activity (Martikke et al., 2019) provides one possible avenue for future research and intervention. Finally, relationships with peers were important to foster a sense of self-belief and resilience (Mishra & Sondhi, 2019), although there were many barriers to relationships with peers, especially for young mothers.
Peer relationships could help care leavers navigate the hostile community environments they could find themselves in, with peers offering ways of coping with environmental, socioeconomic and interpersonal barriers. However, one of the barriers to relational support and the continuing development of resilience was not knowing what to look for or not recognising support when offered (Refaeli, 2017). This implies relationship coaching could be beneficial so that young people can recognise relational opportunities and appreciate what to look for in supportive relationships. Based on these findings, three priorities for early intervention appear to be psychoeducation around 'safe' relationships and asking for help, experiencing love and trust in caregiver relationships, and preventative intervention to ensure young mothers in particular have good quality access to supportive social networks.
In terms of helpful therapeutic mechanisms to support the development of resilience, interventions to nurture self-compassion to address negative affect, and feelings of shame and guilt seem to hold relevance and promise. Promoting self-value and self-belief should also be prioritised within interventions as these factors were connected with fostering agency and resilience. Positive thinking-futuristic approaches, problem-focused coping, and school engagement were all found to be important factors for supporting the development of resilience. Within the second narrative layer, residential schools seemed to promote a positive future focus successfully and practically through training for future endeavours. Narrative approaches, such as constructing a positive origin story that focused on overcoming challenges also seemed to promote and maintain resilience, and positive visualisations of a more stable future (Van Breda & Hlungwani, 2019) nurtured hopefulness. In summary, although relatively little was explicitly discussed in relation to skills for resilience and learning, the therapeutic mechanisms to develop resilience were apparent across many of the papers reviewed.
The ability to 'borrow' resilience from others was cited as beneficial, as was feeling connected to another and therefore protected against hardship were also important in terms of maintaining resilience. However, an area in need of further conceptual research centres around the lack of clarity as to what can be conceptualised as resilience compared to the acceptance and endurance of hardship, and if there is an intersection between the two states. With further phenomenological clarity, it may be possible to develop tailored tools and measures to assess needs, to inform intervention design.
In summary, there were some inconsistencies and equivocality regarding the presence of resilience and positive psychological outcomes between age groups and genders. Future research should critically consider how sensitively resilience measures cater for younger children, girls and young women who may be facing multiple identity transitions simultaneously, cultural sensitivity and constructs of resilience, and those with significant mental health challenges so as to capture developing resilience and the prospect for resilience to grow over time. This developmental approach would reflect aspects of positive psychological outcomes and future-focused thinking associated with PTG, rather than solely the ability to manage and cope in the present, as found within some conceptualisations of resilience. The heterogeneity and geographical dispersion of this population group makes cross-case qualitative analysis and comparison somewhat challenging. Future qualitative research or evaluation would benefit from the refinement and development of systemised and theoretically informed approaches to qualitative data collection in this population, a prototype for which is described elsewhere (Webb et al., 2016). With this in mind, it may be helpful to carefully consider whether it is beneficial to separately measure resilience and PTG, or whether developing an approach to measuring resilience inclusive of personal and post-traumatic growth may be more effective and sensitive to this group of young people who have typically faced far more adversity earlier in their development than non-looked after peers. A developmental perspective upon resilience, appreciative of intra-, inter-personal and ecological factors, also aligns with recommendations for conceptualising resilience with young people (Masten & Barnes, 2018). Further, an approach that embraces temporal changes and intrapersonal factors may help differentiate between resilience inclusive of PTG and over-self-reliance, which appear to be conflated in some reporting.

Conclusion
Resilience as a concept within the field of residential care needs to account for the significant relational and systemic influences young people in care are exposed to, which can directly affect their resilience. Resilience measures should be tailored to reflect the needs of this unique group of young people, to recognise emerging resilience growth within a developmental and systemically influenced framework of resilience. Tentatively, we suggest that due to the ongoing adversities care leavers face, it may be helpful to take a developmental approach to measuring PTG within a developmental framework of resilience growth as sadly the trauma rarely ends when a young person enters care. This seems to be an area in need of urgent future research, which could build upon the work of qualitative reviews such as that undertaken by (Häggman-Laitila et al., 2018). Further, developing Masten and Barnes' (2018) resilience framework for intervention, it would be beneficial to construct a resilience framework for development and growth, sensitive to the needs and trajectories of care experienced young people to measure more accurately resilience growth, which in turn could inform intervention design and evaluation.
Funding Quality Protects Children.

Conflict of Interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, 1 3 adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.