Health and wellbeing of survivors of sexual violence and abuse attending sexual assault referral centres in England: the MESARCH mixed-methods evaluation

Aims

• To evaluate the evidence for psychosocial interventions for survivors of sexual violence and abuse.

• To examine SARC models of service delivery by exploring work practices, workforce and technology factors and the integration of SARCs in the broader context of a health and community response to sexual violence and abuse.

• To undertake a 1-year follow-up study in a cohort of survivors of sexual violence and abuse to explore the effect of different SARC models of service delivery, access to health care and other SAAS on PTSD, depressive symptoms, quality of life (QoL), substance misuse, violence re-exposure, sexual health and costs.

• To identify the effect of delivering post-crisis trauma-focused counselling interventions in the voluntary sector compared with the NHS on health and other outcomes.

• To engage young SARC service users (age range: 13–17 years) and explore impacts of exposure to sexual violence and abuse on their lives and quality of care and support from SARCs.

• To draw on the cohort sample using maximum variation sampling to ensure a broad range of subgroups represented, supplemented by a community sample and to conduct a qualitative investigation of the experiences at SARCs and outcomes of SARCs and barriers and facilitators to access.

Research questions

1. For individuals who have experienced sexual violence and abuse, do psychosocial interventions reduce PTSD and other poor health outcomes? What are providers’ experiences of delivering such psychosocial interventions? What are the experiences of survivors and supporters in accessing such psychosocial interventions?

2. What are the implications of four inter-related aspects of SARCs – the everyday work they do, the workforce, the technology and the organisation – for the delivery of SARC services? What is the work of SARCs including the types of interventions delivered? Who is the SARC workforce? What are the technologies that enable SARCs to get work done? What is the organisational context of SARCs and to what extent are SARCs embedded within the overall response by statutory and voluntary sector organisations to the needs of survivors of sexual violence and abuse?

3. What are the health and cost trajectories of those who attend SARCs? How can these be compared for different SARC models of service delivery and access to health and SAAS?

4. What is the effect of receiving post-crisis counselling in the voluntary sector when compared with that provided through the NHS?

5. What are the experiences of children and young people (CYP) of receiving care and support from SARCs?

6. What are the experiences of access to SARCs by survivors from marginalised or minority populations?

Oversight and approvals

Approvals were granted by research ethics committees and the Health Research Authority before any data collection. Approval dates and reference numbers are provided in the Ethics Statement at the end of the report. In addition, five NHS SARC sites and one onward referral NHS agency confirmed site capacity and capability. Amendments to the study protocol and documentation were discussed with the SSC, and approval was sought and obtained from the relevant ethics committees. These amendments are outlined in V3.3 of the protocol. 22 The DMEC met on three occasions over the project course with the purpose of monitoring data collection and analyses across all studies, risks to the project and making recommendations on the ethical considerations where appropriate.

Ethical considerations

We now outline key ethical considerations that were made across the studies that involved primary data collection (SARC process evaluation, cohort study, embedded qualitative study and the CYP’s study). In the presentation of findings and quotes, we have ensured that participants are not identifiable. Where we had permission to use direct quotes from participants, we have used pseudonyms (chosen by participants or the research team) or quotes have been attributed to an unidentifiable label. In addition to the project team receiving regular training for conducting interviews with survivors of sexual violence and abuse, we developed a safety protocol drawing on safety and distress protocols from our previous published work, and based on the expertise of the MESARCH team, members of our Lived Experiences Group (LEG) and SSC, staff at Coventry Rape and Sexual Abuse Centre, and informed by the Charter of Survivors Voices. 23 The safety protocol outlined different procedures the project team would follow in relation to key risks. The procedures are as follows:

Data management

A comprehensive data management plan was developed for our studies involving data collection. We regularly monitored our processes regarding where and for how long personal data and anonymised data sets were stored, as well as who had access to these data. To manage the personal data and safety notes of participants in our 1-year cohort study, a secure tracking system was developed for use by staff members who were interviewing participants, enabling accurate recording and tracking of participants’ personal data (such as their contact details), when they were due follow-up interviews, and information about safety concerns (this included the RAG system) and any actions implemented.

Background

Patient and public involvement refers to involving patients, members of the population directly or indirectly affected by the target problem and/or members of the public in the research process. In MESARCH, we worked with survivors of sexual violence and abuse. While a research context could be empowering and beneficial to survivors in their recovery from trauma,24,25 negative engagement could lead to re-traumatisation. 23 The use of PPI to shape how research with survivors is carried out can lead to high-quality practice and positive experiences for participants. However, previous research suggests that there is considerable scope for improving the quality of PPI in sexual violence and abuse research. 23,26,27 Our research aimed to advance practice in this area.

Methodology

Drawing on principles of co-production, we worked in partnership with those with lived experience of sexual violence and abuse, referred heretofore as ‘the MESARCH LEG’. MESARCH reflects the shift to research being carried out collaboratively with members of the public who share decision-making rather than models of lay consultation and tokenistic involvement. 28,29 Several theoretical positions underpinned our approach. First, we value a multiplicity of knowledge and have afforded great weight to experiential knowledge within the research process thus creating opportunities for epistemic justice, a concept coined by philosopher and feminist Miranda Fricker30 where voices that are often silenced or marginalised are amplified. Second, building and maintaining strong relationships with the LEG was central to our approach to lived experience involvement, premised on theories of relational engagement, dialogical ethics and an ethics of care, and reflecting the significance of the human and socially interactive element of engagement. 31–33 Lastly, drawing on the concepts of citizenship and democracy, the active and meaningful involvement of those with lived experience of sexual violence and abuse in our research promotes the empowerment of survivors. 34 This is particularly relevant in relation to involving survivors of sexual violence and abuse who will have experienced disempowerment through abuse, communities and institutions. To ensure good practice in involving survivors of abuse and trauma in our research, we worked closely with Survivors’ Voices (a survivor-led organisation that harnesses the expertise of people affected by abuse), drawing on their Charter, which details principles and good practice for good survivor engagement in research. 23 We were also guided by the Rape Crisis National Service Standards, adapting a set of quality service standards to meet the needs of survivors of sexual violence and abuse. 35 These principles provide a framework against which we have monitored the extent to which our research is trauma-informed and achieves authentic survivor focus.

Methods

Our LEG was supported by a dedicated public engagement officer to enable a co-design and co-production approach across the project. LEG members were recruited through survivor support groups, delegates at conferences, participants on related research studies and recommendations from other researchers. Over the lifecycle of the project, the group included 11 members participating at different times and brought 794 hours of expertise into the project. Members of the LEG were diverse across sex, age, ethnicity, sexuality, education, employment and life opportunities. They brought a wide range of professional and lived expertise as well as various engagement with the CJS, NHS, therapies and third-sector services. We consistently worked to expand diversity as members moved on and new members joined, recognising the association between diversity and gains being achieved across the research.

The involvement of the LEG in the project was guided by a terms of reference document that was co-produced and regularly reviewed. Involvement was enhanced through formal meetings (8) administered through the lead university, participation of two LEG members in SSC meetings (10) to provide a lived experience perspective and contribute to the oversight of the project, and ongoing communication through other means as and when it was required. The work of the LEG was fully resourced by the NIHR covering people’s time, travel, childcare expenses and conference attendance. We also engaged with survivor support groups, organisations led by survivors, specialist sexual violence and abuse organisations and key stakeholders to further embed PPI across the project.

Patient and public involvement activity across the project lifecycle

Patient and public involvement was embedded across all aspects of the project, from its design and delivery to the interpretation and sharing of the findings (see Figure 1).

FIGURE 1.

Embedding lived experience and ‘experts-by-experience’ across the MESARCH project. CYP, children and young people; ISVA, independent sexual violence adviser; LEG, lived experiences group; PPI, patient and public involvement.

Knowledge sharing

The LEG has been heavily involved in sharing knowledge and insights gained from the research, for example, shaped plain language statements across outputs such as our Cochrane Reviews (see Chapter 3). Other examples include a fully-supported LEG-led production in the form of ‘photovoice’ where members presented photos to convey recovery journeys (https://vimeo.com/547572348/2709483bd8). The end of our project knowledge-sharing event (7 December 2022; see Figure 2) was co-designed and co-delivered by the LEG and researchers (http://mesarch.coventry.ac.uk/whats-our-project-creating/). The LEG also shaped our ‘#whatareyoudoing’ campaign highlighting the problems of sexual and domestic abuse in young people’s relationships39 partly funded by NIHR as part of MESARCH, and raised this work as important at the NHSE survivor forum.

FIGURE 2.

Visual minutes from the MESARCH knowledge-sharing event.

Impact and evaluation of PPI

The overall impact of the LEG/PPI work summarised in Figure 1 on the research has been tremendous and is understood to explain the low incidence of adverse events and known harms arising from the research. This expertise was also part of operational success. There were also direct and unanticipated positive benefits for the LEG members themselves as captured in the testimony of one member (see Appendix 1, Box 43). LEG members also told us that being part of the group has contributed to recovery by enabling safe exposure to triggers. The conversations and interactions (online, by email, in person and in-print) that were part of the everyday work of the project normalised talking about sexual violence and abuse. Members have also commented on how being involved in MESARCH and being heard by MESARCH amplifies individual voices and provides a platform for creating change and being part of influencing better responses to others affected by abuse. PPI also enhanced the research team’s confidence to undertake this work, deepening our understanding in relation to trauma and abuse, and our LEG members’ commitment to the project was hugely motivating for overcoming the challenges encountered during the project. Our work with the PPI group has also inspired future research around understanding long-term physical health outcomes for survivors of sexual violence and abuse, and influencing the implementation of trauma-informed practice across the NHS (considering clinical services beyond mental health); building a toolkit of methods for doing better research in this field; and growing the evidence base on novel interventions for survivors that reflect survivors’ individual needs and preferences.

In an effort to identify any gaps in our PPI provision in MESARCH, we commissioned an independent, survivor-led organisation to evaluate it. 40 A summary of the conclusion reached by Survivors’ Voices is provided in Box 1.

Study design overview

The main component of this study was a process evaluation of SARCs. A process evaluation is broadly recognised as a method that seeks to understand the work of an intervention, how that intervention is delivered and its effect, alongside contextual factors that may affect outcomes. 53 In completing this process evaluation, we undertook qualitative investigation of professional and survivor voices regarding SARC service delivery and integration of SARCs within the SAAS pathway. From the mapping of all SARC sites in England, we identified several diverse sites to undertake the qualitative enquiry. Data used to map SARCs were collected between March and July 2019. These data were supplemented with national SARC indicators of performance (SARCIPs) data supplied by NHSE54 and gathered from SARCs between April 2018 and April 2019. Data were collected from professionals and survivors participating in the qualitative study between July 2019 and July 2021. We also drew upon qualitative data gathered with survivors as part of the linked cohort study (see Chapter 5) between September 2019 and May 2022.

Changes to protocol

We planned to use Normalisation Process Theory, an effective approach when exploring a new intervention or service. 55 However, when it came to applying Normalisation Process Theory to our data, the services, staff and integration of SARCs were not newly implemented. Therefore, we sought an alternative theoretical approach to capture the importance of effective integration between different organisations and professionals, and drew upon theory related to inter-professional collaboration.

Theoretical approach

Bronstein’s theory of interdisciplinary collaboration48 has been applied extensively across other fields and in two studies exploring integration within Dutch and United States (US) models of SARC provision. 49,52 Coordination and collaboration have been hypothesised to be linked with the effectiveness of Sexual Assault Responses Teams in the USA. 50 These studies point to the theory’s relevance to capturing interdisciplinary collaboration and integration for SARCs in England. Bronstein defines interdisciplinary collaboration as a process that facilitates the achievement of goals that cannot be achieved by individual professionals. 48 Bronstein’s model comprises five components: interdependence, newly created professional activities, flexibility, collective ownership of goals and reflection on process. Interdependence refers to the dependence between and reliance upon different professionals in working together to achieve their own goals and requirements. Newly created professional activities are collaborative acts, programs or structures that can help achieve more than that if undertaken independently by professionals. Flexibility captures deliberate acts of role blurring in which professionals working together can achieve greater efficiency and effective working. Collective ownership of goals relates to a shared responsibility among professionals to achieve goals with a commitment to client-centred care. Finally, reflection on process relates to professionals acknowledging their collaborative roles and evaluating how to effectively achieve and strengthen these relationships. Bronstein argues that inter-professional collaboration can be affected by several characteristics including professional role (e.g. clear and well-defined roles), structural characteristics (e.g. manageable workloads), personal characteristics (e.g. how collaborators perceive each other) and history of collaboration (e.g. prior experiences of collaboration). Previous research has highlighted challenges to inter-professional collaboration, for example, perceived power imbalances and different professional focus/orientations between staff members;49 these challenges will be explored in the context of this research.

Sampling, recruitment and participants

We mapped SARCs across England using data collected from SARC managers through an online survey (36/48 sites returned data) and SARCIPs data. We produced a data set, which allowed sampling of sites for the qualitative process evaluation and the cohort study (see Chapter 5).

Using this data set, we recruited a stratified random sample of seven SARCs from all 48 sites, with strata defined according to service delivery model (police-led, NHS-led, charity-led or private sector); size (small, medium and large) and level of integration of services (on-site, embedded ISVA service or not).

At each site, we invited a range of professionals based at the SARC to participate in an audio-recorded interview. We also approached professionals external to the SARCs whose organisations were part of inward and onward referral pathways. We recruited directly through the UK Association of Forensic Nurses to ensure representation of Forensic Nurse Examiners (FNEs) within the sample. We also invited survivors who had attended these SARC sites through the staff. We undertook 72 interviews with professionals (see Appendix 2, Table 23). Five interviews were conducted with survivors initially. We subsequently recognised the wealth of qualitative data being gathered as part of our cohort study in reference to accessing care at SARCs (see Chapter 5). The qualitative data from our cohort study provided data from a further 293 survivors who accessed 21 SARCs including the eight SARCs where we undertook the in-depth evaluation of SARC services.

Analysis

Thematic analysis (TA) was the chosen process for analysing the two sources of data: in-depth interviews with survivors and professionals (n = 77) and data arising from asking open-ended questions about care during the baseline cohort study interviews (n = 293). Specifically, Braun and Clarke’s Reflexive TA approach was applied. 56–58 This method adopts a flexible and organic approach to coding and subsequent theme development, and it is considered theoretically independent. Primarily, inductive coding was undertaken, being led by the content of the data, with semantic coding used (the data are considered to represent the explicit meaning of the data). The standard TA method was followed including familiarisation and initial code development, clustering of codes to initial preliminary themes, theme development and subsequent finalising of the thematic themes. 56 Coding was undertaken by four members of the team, and during the initial coding phase, regular meetings were undertaken to inform, develop and refine the coding book. Once all the initial coding was completed, one team member checked all codes for quality purposes and to ensure that the data fitted to the code label. This process also refined codes into preliminary themes. The analysis then focused on key themes that explained the research questions in ensuring that these themes were distinct and closely aligned to the data.

What is the work of SARCs including the types of interventions delivered?

Sexual assault referral centres provide a vital early or the first point of care for survivors and providing a frontline response to survivors constitutes the majority of their work. There is a standard care offering at SARCs based on Service Specification 30. 19 In the research, survivors focused on the elements of core services that were salient for them. Crisis support was strongly valued by participants – the emotional support, information and options presented, and the referrals made. ‘[SARC staff] tried their hardest not to cause me any more distress, of course there was distress and it was uncomfortable, but not their fault’ – the FME is an extremely demanding experience, and for one participant, akin to ‘having a post-mortem while you were still alive’. ISVA support, which may be provided directly through SARCs, was also highly valued, particularly in relation to practical support through investigative and criminal justice proceedings and the techniques to support recovery. Counselling, increasingly offered through SARCs, was considered a largely positive experience, although wait-lists to access counselling, and the restrictions on the number of sessions meant that some survivors ‘felt I was getting somewhere and then it has to stop, I didn’t think it was good enough that it stopped’.

Another core function of SARCs is access to other services; SARCs opened a range of referrals for clients. However, drawbacks were ‘I was getting bombarded with a lot of people calling me … I had five or six agencies calling me the day after, my phone didn’t stop ringing’ and the consequence of this as ‘I had to keep going over and over my story’. It was highlighted how survivors find it difficult to navigate the services and suggests the value of SARC and ISVAs working closely to assist this understanding: ‘I think victims and survivors find it really difficult to understand the landscape of what’s going on out there in terms of the case, who’s providing what’. (ISVA service delivery manager)

While service specifications set out the core areas of delivery for SARCs,19 variation arose with different leadership models, commissioning arrangements, geography and the needs of the local community. Participants commonly identified typical journeys through SARC services. They also noted divergence from primary routes. Access needs and how people engaged with the services varied with individuals’ circumstances (e.g. specific or complex needs, location); the time since the trauma occurred (e.g. non-recent abuse or CSA); and external factors such as being on wait-lists for therapeutic care and the status/progression of criminal justice proceedings. The research also documented the effects of the COVID-19 pandemic, which saw a step-change in access and service provision and witnessed clear evidence of collaboration within and between the services as commented on by this FNE: ‘you can’t deliver a service like this without being a close-knit team’.

While standard care requirements are set out for SARCs by NHSE, there is much scope for additional work to enhance provision and access. The research documented examples of SARC teams working creatively to ensure that their services could accommodate the range of needs of clients from different communities/locally including those whose access to services had been delayed, non-linear or complicated by different factors. One police officer demonstrated reflection on process in how they worked with their local SARC: ‘I attend the SARC a lot because if there’s been an operational issue, even though it may be north of the county, I will get invited, and then [SARC manager] and I and others will sit down and go, “ok, how can we deal with [this], what can we do going forward to make sure that this doesn’t happen again?”’ Teams developed new initiatives such as in-house emotional support or counselling; acting as a holding contact point while clients awaited access to other SAAS and other forms of care (e.g. community mental health, longer-term counselling); and enabling video interviewing/live link technologies. SARC staff expressed a commitment to clients, ‘we’re here, checking in, or there’s always someone here twenty-four seven … people do call us and say, “I’m suicidal” or ‘I don’t know where to turn to, I haven’t heard from this person”’. They wanted to ensure that their services could support people at any point after sexual violence and abuse, and they saw a role for being a point of contact for survivors of CSA and informing survivors about options for reporting and referrals to ISVA and third-sector support. SARCs expressed that efforts should be made to address public perception that SARC support means police involvement. They demonstrated how self-referral to services was fostered and also noted the ongoing difficulties for self-referring survivors in terms of physically reaching their services in some areas of the country (by contrast, those who report to the police are usually transported to SARCs by police officers). SARC staff operated in line with trauma-informed care; professionals were keen to see that choice was exercised in key areas such as reporting mechanisms as noted by this crisis worker: ‘I think it’s really important to allow people that opportunity to validate what’s happened to them … to try and re-balance, regain some power in a situation where they’ve felt powerless and to have that choice’. Staff highlighted the longer-term storage of forensic data and anonymous submissions to police intelligence as an underutilised service.

In terms of people’s access, there was widespread endorsement of anonymous locations and for sites where survivors can provide alternative reasons for being there (e.g. hospital site). The staff members were aware that co-location with police forces was challenging for survivors. SARC buildings ranged from purpose-built to adapted, and survivor-centred design features were strongly supported. One such example was entering and exiting the space through alternative doors to signal the beginning and end of the SARC journey (at services where clients generally do not return for other care and support). Staff reported a range of issues with SARC estates from lack of physical and confidential spaces for client support and administrative work and shared entrance/common corridors. Forensic rooms were perceived as overly clinical and suggestions were made by professional and survivor participants to address this through features such as wall/ceiling decorations. The staff members reported efforts to ensure that the SARC was welcoming CYP, that furniture was comfortable and that coverings provided to survivors were appropriate (e.g. participants cited instances of gowns being transparent). SARC teams expressed concerns about reaching clients in remote areas. The shift to greater use of digital services and remote care in response to the COVID-19 pandemic had the effect of mitigating some of the geographical barriers affecting survivors’ access to SARCs and ISVA care. Many survivors also identified a preference for remote contact with providers, feeling ‘more relaxed’ and that ‘the distance of a telephone call is better for me’ when talking about difficult topics.

Who is the SARC workforce?

Across service delivery models, staff roles generally seen at SARCs were SARC manager/deputy and an administrative member/team, crisis workers and FMEs/FNEs. Flexibility in roles at SARCs was common ‘my role covers a lot of areas’, for example, managers and administrators trained as crisis workers. Depending on the degree of integration at the site, ISVAs enacted their roles in flexible ways, as core staff at a SARC; employed by the voluntary sector but co-located at a SARC; with largely visiting/remote role; or based physically at a voluntary sector organisation. Thus, many models could be drawn upon to suit the landscape, for example, some sexual health services have ISVAs. NHS SARCs tended to employ clinical staff directly, whereas private sector SARCs made greater use of staff banks. With SARCs increasingly offering short-term psychosocial interventions, counsellors were also integrated into the staff of SARCs, frequently on a part-time/co-located basis. Finally, several roles were noted to lend themselves to specialisation, with some sites training crisis workers and ISVAs to specialise in the care of particular groups, for example, CYP, individuals with learning disabilities. Within the scope of the current research, we focus on two prominent workforce issues: (1) the nature of the crisis worker’s role and related training and (2) challenges to effective interdisciplinary SARC teams.

The crisis worker’s main purpose is to support clients of the service. However, the role involves a diverse range of task and balancing the sensitivities of cases and needs and expectations of vulnerable clients with the constraints of what can be provided:

One morning I might be helping with a forensic medical, and in the afternoon we might be doing call-ups for email enquiries, and reaching out to people that have accessed that way. I think it’s an incredibly difficult role to learn, if I’m honest, because it’s so varied and wide. There are some repetitions within it that are very quick and easy [to learn] … but there’s other stuff that makes it much trickier, like around geographical location. So understanding what services are purchased in relation to paediatric FMEs in a geographical location, which means that if somebody is ringing from [name of city] and they [want to] come in, do we see them or not? Or there’s understanding and learning what the service offers outside of the area as well as in the area.

Crisis worker

Without prior experience of the region, crisis workers’ practice might suffer. Training varied from across sites with most involving formal learning components, for eample, safeguarding, and a period of shadowing: ‘my first client interaction that I shadowed, it was kind of a steep learning curve because it’s not like anything else’ (crisis worker).

Shadowing, and training more broadly, was noted to be variable and rely on standards set by providers. The crisis worker’s role has not attracted much attention within the research literature beyond the toll it places on individuals,45 an issue also recognised in our research. While the role was consistently valued by service users for whom the crisis worker is the first SARC professional they encounter, ‘I found it helpful that I received emotional support from [the crisis worker] when I got there because it was scary and going with the police’ (survivor), professional participants viewed the role as lacking incentivisation (e.g. opportunities for progression).

‘When a client or when a victim is at the SARC and you’ve got [private company who run forensic services], and you’ve got the police, and you’ve got a crisis worker, surely all three agencies should be working together because we’re supporting that one person’. (SARC manager) – meeting the client’s needs was a shared goal for staff at SARCs and there was frequent reference to ‘we are quite a tight knit group’ (ISVA). However, in terms of the challenges to effective interdisciplinary SARC teamsthe nature of staff contracts was a common workforce concern (subcontracting of SARC roles, on-call/zero hours contracts, staff employed by different providers). A crisis worker participant contrasted the scenario of where FNEs are employed directly by SARC and the experience of working with FNEs/FMEs contracted on an on-call basis: ‘the other doctors, you only see them every now and then, so you don’t really build up the same rapport with them, and obviously they all work slightly different … try and suggest the way that you would do it normally, just to try and make it easier and quicker, because obviously we want to put the victims first, so we want to make it easier for them … I do think that it’s so much better when you work together’. Implicit here is that placing survivors ‘first’ requires effective teams and staff who are familiar with each other and the setting. While all participants agreed that consistency in staffing was important for team-working and reducing isolation, flexible contracting was desirable for some professionals balancing other livelihoods and commitments. SARC managers did have particular concerns about crisis workers in this regard: ‘[crisis workers] do an amazing job but I think they do just feel a little bit disconnected from the service’.

Despite the workforce concerns that were voiced across the staff members and SARCs included in the study, we found very high levels of intra-SARC interdependence, particularly among crisis workers and FNEs. There was strong recognition of the importance of clear communication and inter-professional respect, supporting each other to reduce the impact of their work, and a commitment to providing an integrated service to the client. Simple examples such as gathering information from their clients together and sharing information between professionals as appropriate benefited clients in not having to repeat information or re-tell about their experiences and evidenced newly created professional activities. Collective ownership of goals in pursuit of client-centred care represents a key aspect of effective interdisciplinary collaboration. 48 Reflecting the impact of this care, one survivor commented: ‘I felt safe with the doctor, I felt safe with [crisis worker name] who was in the room. I didn’t feel that anything bad was gonna happen’.

What are the technologies that enable SARCs to get work done?

This section focuses on two key issues: (1) use of remote/digital services in providing care and (2) challenges emerging in information sharing across the services. The advent of the COVID-19 pandemic meant that SARCs used digital technologies to facilitate service delivery, including triaging of clients before they attended the SARC. Staff reported that they carefully determined the appropriateness of remote care for different clients. Some benefits to triaging were realised in that clients spent less time at the SARC once they arrived. Survivors reported both positive and negative responses to the increased use of digital technologies, highlighting the need for developing remote services as part of a suite of options and tools. The second key issue related to information sharing. Different IT systems for different organisations meant that SARC staff members spent considerable time making manual referrals through online systems or secure email systems, which added considerably to workload:

I referred a lady to adult social services because I felt her needs would be met with [social care provider]. But I couldn’t go direct with [social care provider]; I had to go through adult social services, and they have to make the referral. And I was told on that phone call that “somebody would be contacting you.” And today, I’ve not had a contact, and it’s been about two weeks, so that’s something that I am going to have to chase up. So that is quite time consuming.

ISVA

This work, described by one SARC staff member as their ‘nagging work’, was frequently highlighted with regard to sharing information with partner organisations. Survivors experienced interruptions or delays in access to onward care:

I haven’t received any care, I got told I’d have sexual health screening, that hasn’t been sorted or anything, I had to go to the doctor’s about getting an examination done, they were supposed to help sort that out, but they didn’t.

Survivor

To what extent are SARCs embedded within the overall response to the needs of survivors of sexual violence and abuse?

The most important external partnerships for SARCs included the police, sexual health, ISVAs, third sector and social care. For example, our data suggested well-established and collaborative relationships between SARC staff and police officers. Single point-of-contact officers in police forces worked directly with SARC managers to develop pathways for survivors and training by SARCs to ensure police officers understood local provision and access. Collaboration with third-sector organisations was more challenged: ‘with any multi-agency working, we all seem to work parallel with each other quite often and not come together at the top to deliver’ (crisis worker). There was a strong sense that collaboration was driven by individual senior members of staff at the SARC and/or the partner organisation and involved local implementation. Positive relationships at senior levels trickled down and fostered positive cross-working, seen to ultimately benefit survivors. Examples of cross-working included co-delivery of training and sharing of resources and space. Professionals expressed a wish for change. For some, implementing a more consistent SARC model was the answer, with more services under ‘one-roof’ such as integrated ISVA and sexual health services. Others recognised existing high-quality provision in these areas within their communities and that pathways and relationships should be strengthened. Another example was Achieving Best Evidence interviews, with some believing these should be conducted consistently at SARCs while others saw SARCs as a place solely for acute crisis care and considered that survivors may not wish to return to these locations. Participants pointed out that SARCs have all evolved in different ways in the communities they serve and, in expressing how the sector needs to develop in the future, emphasised the need for simplification and better approaches to supporting clients with complex needs.

The role of SARCs is vital, but there was real concern about the invisibility of SARC services. Survivors relied on the police and others to learn about SARCs, ‘I wouldn’t have known about them if it hadn’t been for the detective’ (survivor). In response to the lack of public awareness about SARCs, there were several examples of teams initiating outreach services in their communities, often targeting groups seen to experience even greater barriers to receiving help (e.g. men, minority communities). Gaps in knowledge about service provision by SARC extended to important professional groups: ‘there are so many more people that need to know about the SARC’ (crisis worker). GPs were identified as a professional group lacking awareness of SARC services, with referrals from GPs being infrequent. This SARC manager talked about the difficulties of engaging with GPs:

Probably in defence of GPs, they have just got so much demands on their time and they don’t have free time, so we’ve put on training events especially for GPs. We do the open days, we try and target GPs … some of them are just not aware of what we’re doing, they’re really out of date, the information that they are giving clients.

From the perspective of a survivor: ‘GPs have got such little understanding …. They don’t take it in. And then they’re like, we don’t know who you can speak to … they just don’t know what the right support is’.

Summary of main findings

This work highlights the excellence in care and support being provided within SARCs, as such, the value of SARCs in the provision of care and support to survivors is supported. The contribution of SARC provision to the care pathway, extends beyond provision of forensic services, importantly fulfilling an early role in coordinating care for survivors. However, this research also documents the challenges that SARCs, and the SAAS sector as a whole, experience in addressing survivors’ needs. Drawing on 77 interviews (72 professionals, 5 survivors) and 293 survivor responses, the novelty of this qualitative enquiry is evident in the breadth of participants spanning both inward and onward referral routes, as well as SARC staff themselves, along with the voices of survivors in sharing about the care and support they received.

In addressing the first aim of this work, findings have highlighted key issues and concerns associated with the work of SARCs, their workforce and the use of technologies. In doing so, this research documents the high level of dedication of SARC staff and captures their commitment to client-centred care and an ethos consistent with trauma-competent practice. 60 This commitment of professionals at SARC was further evidenced in their response to the COVID-19 pandemic, where they effectively utilised technology to ensure that care could be continued while enabling greater flexibility and choice for survivors. Appreciating several distinct roles are involved in producing the range of care and options for survivors, our findings emphasised crisis workers as vital to a survivor’s care journey at SARC. Attention to the crisis worker role is fundamental in further developing the workforce of SARCs.

We found evidence of effective interdisciplinary collaboration spanning across Bronstein’s five components of interdependence, newly created professional activities, flexibility, collective ownership of goals and reflection on process both within SARCs and across the sector. Despite being hampered by the complexity of how different services are contracted, SARC staff worked hard to achieve cohesive teams and ensure that an excellent service is provided for their clients. We saw many examples of work that reflected good quality inter-professional collaboration, specifically inter dependence and collective ownership of goals in delivering client-centred care. Matthew and Hulton called for research to identify whether there is a link between effective SARC collaborative practice and survivors’ experiences of SARC services. 51 This research answers that call with our findings showing high levels of effective inter professional collaboration found between SARC staff (specifically between crisis workers and FMEs/FNEs), and subsequent positive endorsement of SARCs by survivors.

However, there were factors that can hamper inter-professional working within SARC settings. The ways in which professional roles at SARC are commissioned can lead to short-term teams and unstable working hours of staff, whereby collective ownership of goals in pursuit of client-centred care is difficult to achieve. In comparison, consistent teams enabled strong inter-professional working and evidenced best practice. This generated cultures of transparency and respect, more compassionate support and offered a high-quality experience of care for survivors.

Across the sector, and moving beyond the SARC setting, the second aim of the research was to explore the integration of SARCs in the broader context of a health and community response to sexual violence and abuse. In doing so, our research gave a positive picture of interdisciplinary collaboration across the wider sector across criminal justice, health and third-sector organisations. Professionals across these sectors clearly engaged in collective ownership of goals48 in recognising the importance of effective, and cohesive care and support for survivors. This runs counter to US research where research noted cross-system coordination being non-existent or lacking consistency, with subsequent challenges to the effectiveness of SARC provision and integration. 61 However, barriers remain, primarily in the form of inefficient communication systems for sharing information. Effective information sharing systems were highlighted as being critical to enabling strong inter-professional working and importantly contributed to efficient working practices, reducing the workload burden of SARC staff.

However, this study has demonstrated challenges to optimising the access to and fully realising the benefits of SARC services, specifically relating to the invisibility of SARCs. SARCs need to be more apparent as a care option for those in crisis, particularly if the person experienced rape, sexual assault or abuse recently. This includes wider community awareness (e.g. families, workplaces, friends of survivors) given patterns of disclosure in personal/social networks. 62 It is paramount that public awareness of SARCs is prioritised, for example, through investment in public awareness campaigns. This is particularly important for groups of survivors who are underrepresented at SARCs and experience barriers to accessing them. NHSE made efforts to address this problem with a campaign in 2022 targeting specific groups including black and Asian women, men, and survivors from lesbian, gay, bisexual, transgender and/or queer (LGBTQ+) communities with a focus on youth across these groups. A broader campaign was also run in South-East England targeting all ages.

Strengths and limitations

The study’s originality resides in the breadth of the experiences brought to bear on questions about the role and effectiveness of SARC services, with input from 35 SARC professionals and 37 stakeholders outside of SARCs as well as the voices of 293 survivors who have accessed 21 SARCs in England. While prior work49–52 has explored aspects of the work captured, this study benefited from the large-scale nature of data collection which allowed triangulation of experiences of multiple informants. There were important gaps in representing those inside SARCs (medical examiners). The study also lacked GPs and social workers, two professional groups that are integral to early intervention in sexual violence and abuse.

Implications for health care

Recommendations for research

Given the gaps in including informants from social work and general practice, future research could examine the specific challenges these professional groups perceive in participating in a community response to sexual violence and abuse. Furthermore, while this research has identified excellent practice in relation to collaboration, continued research attention is needed to evaluate these practices to ensure that this captures the complexity of how SARC services are delivered both nationally and internationally. Overall, we hope to have highlighted the dedication and creativity of staff delivering client-centred care and support in the SAAS sector, which was reflected in survivors’ experiences of high standards of care across the sector.

Study design overview

The study design and methods have been published in a protocol study. 63 This was a cohort study of mental, physical, sexual health and cost outcomes over 1 year in adult survivors of rape, sexual assault and abuse who had received care through SARCs in England. Quantitative measures were implemented at baseline (≥ 4 weeks after index SARC access) and at 6 and 12 months post baseline. The study also gathered qualitative responses at each time point related to quality of care. Recruitment commenced in September 2019 and the last 12-month interview was conducted in December 2022.

Changes to protocol

The main deviation from the protocol (V3.3)22 is that loss-to-follow-up was even greater than anticipated. We interviewed only 223 people at 12 months instead of the projected 270 people. We did, however, reach the target sample size at baseline. Our cohort underwent several changes before V3.3 to cope with challenges to recruitment and these are tabulated in the study protocol. Briefly, the most significant amendments to our original protocol were reduction in the target sample size from 750 to 338 and reduction in the follow-up of 2 years to 12 months (each due to the difficulties reaching a trauma population). The COVID-19 pandemic triggered a change in inclusion criteria: eligibility no longer required service users to present in person at SARC services, and we also began to admit self-referrals into the research (as opposed to requiring referral by a professional). Early consultations with our SSC and LEG (see Chapter 2) after the study launch led to additions/changes in measures; for example, we added assessments of eating problems, suicidality, self-harm, drug use and capability. We also changed the minimum age for entry into the cohort study from 16 to 18 years because care pathways are different for those under 18 years, and it was seen as a better approach to focus on young people separately (see Chapter 7).

Sampling, recruitment and procedures

Using data supplied by NHSE and based on our process evaluation study (see Chapter 4), we divided all SARCs in England into strata based on service delivery model, size and level of integration of services. Next, we randomly selected sites from strata and approached them to participate. In line with our planned site recruitment, 15 SARC sites joined the research. We also recruited 22 of their onward referral partners, mainly ISVA services, because the primary recruitment model relied on referral of SARC service users to research by ISVAs. The requirement to recruit this number of SARC referral partners was unanticipated; we intended to have more of the major SARC sites nationally where ISVAs are employed by SARCs and deliver care on-site and therefore make referral to the research more seamless. However, several of these sites were already participating in another NIHR-funded study. 66

Recruitment of individuals was undertaken in three phases; the first involved SARC staff screening service users for eligibility. People were considered eligible if they were service users at SARC, aged ≥ 18 years and had presented in person at any of the 15 participating sites. After 23 March 2020, the research welcomed survivors who accessed SARC services remotely as well as self-referrals, these being participants who had not necessarily accessed one of the originally recruited 15 SARC sites. The main implication of this was that we could not collect service use data about participants from those sites; in most cases, we asked the participants for the information, where appropriate and safe to do so (e.g. What date did you access the SARC? Who was the perpetrator of the offence?).

Returning to the main model of recruitment, people were excluded by services if, in exercising their own judgement, the responsible member of SARC staff conducting screening anticipated a service user may be too unsafe to participate or may encounter difficulties in providing informed consent due to severe mental health, cognitive or learning problems. We also excluded those who were incarcerated, as conducting follow-up assessments in prisons was outside the scope of the project.

Sexual assault referral centre staff completed a screen form for all those aged ≥ 18 years presenting to the service, recording basic demographic data, and offence and referral characteristics and indicating whether the person was eligible or not, with any reasons for exclusion indicated. These screen forms were assigned a study identification number, containing no identifying information, and passed onto the project team to provide an indication of the characteristics of the pool of service users pre-enrolment. SARCs routinely refer on to other agencies, and nearly all survivors are offered access to ISVAs; thus, our approach was to train ISVAs (working inside SARCs and in the voluntary sector) to invite their clients referred from SARCs to consider joining the research. We co-produced recruitment materials with our LEG (e.g. video),67 which the ISVAs shared with their clients. ISVAs gained consent to pass on the person’s name and phone number and preferences for safe/convenient contact to the research team. Study identification numbers were passed on not only from the SARC to the research team but also from the SARC to ISVA services. Thus, when a person consented to be in the research, we were able to link the person to the service data we held. This minimised the number of questions (including sensitive questions) that we needed to ask. A trained project team member established contact within 1 week of receiving the consent from the ISVA or SARC or in a small number of instances, other recruited sites/settings such as sexual health, and contact was made in line with service users’ preferences. Once contact was made by the research team, we followed our recruitment and safety protocols, explaining the study purpose and gaining full consent (see Chapter 2). When consent to participate was established, baseline data were collected and the person was enrolled in the cohort study. Follow-up was undertaken according to participants’ preferences 6 and 12 months after the baseline interview.

Sample size

The primary outcome of interest for evaluating service user journeys after accessing a SARC was PTSD, measured using the Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5). 68 Our interpretation of models based on the PCL-5 focused on what constitutes a minimal clinically important difference (MCID) for PTSD. Stefanovics et al. concluded that differences of approximately 0.5 of a standard deviation (SD) could be deemed clinically important both cross-sectionally and for within-person change. 69 As this study involved neither randomisation nor placebo control, the analyses focused on changes in PTSD symptoms between baseline and follow-up. We considered the recommendations of Stefanovics et al. in terms of quantifying the change-score group-difference we would expect that separates an effective treatment from an ineffective one. Based on guidelines,68 we anticipated this MCID of 0.5 SD to approximate to a 10-point difference on the PCL-5 change-score; however, Stefanovics et al. propose that an SD-based approach is more robust, and we also recognised that the magnitude of change on the raw-scale metric would be influenced by the length of time between baseline and follow-up. To allow for four levels in a predictor variable (e.g. four types of SARC service provider); up to 10 covariates; two time points; an expected small-to-moderate Cohen’s f, which corresponds to the MCID for our primary outcome variable; and an α error probability of 0.05, 270 participants were required at study completion to achieve a power (1−β error probability) of 0.90. Based on previous studies,64,65 our pilot research and early patterns of recruitment and retention, we estimated attrition at 20% by the 12-month follow-up, requiring a baseline target of 338 individuals. Participants who did not complete a 6-month follow-up were still eligible to complete 12-month follow-up measures.

Data collection and outcomes

The form used by SARC staff to screen service users for the research provided initial data on gender, age, ethnicity, offence characteristics including nature of the relationship to the perpetrator and time since the offence occurred; referral into SARC and indicators of vulnerability at the point of service access (e.g. history of mental health problems, disability, substance use). At baseline, we collected additional detailed sociodemographic information. Baseline data were gathered using computer-assisted telephone interviews by staff trained in survivor-centred research practice (see Chapter 2). We enquired about adverse childhood experiences (ACEs) using the World Health Organization’s ACE International Questionnaire, which measures exposure to 13 adverse experiences under the age of 18 years. 70 We screened for lifetime intimate partner violence as well as in the year up to baseline using the Afraid/Controlled/Threatened/Slapped or physically hurt Screening Measure tool. 71 All other measures applied at baseline focused on standardised timeframes and were used to assess change in survivors’ health and well-being journeys over the year period (see Table 2). Sexual health will be reported elsewhere. Our approach to measuring costs is outlined in the economic evaluation (see Chapter 6). After enquiring about service use at each timepoint, we asked participants to rate the extent to which four different sectors/services (SARC; voluntary sector; NHS; policing and justice) had been beneficial, neutral or harmful to their health and well-being using a scale that extended from minus 100 (maximum harm) to plus 100 (maximum benefit).

Qualitative methods

In addition to the quantitative measures and scales indicated above, interviews also included open-ended questions about participants’ experiences of care and support in the context of four different sectors. This produced an extensive longitudinal qualitative data set to complement the quantitative data. One of the study aims was to explore how survivors experienced mental health care in the NHS compared to the voluntary sector. As there was so much overlap and movement of participants between voluntary and statutory mental health services, it was not possible to examine this aim quantitatively. Thus, we applied Braun and Clarke’s Reflexive TA approach56–58 (see Chapter 4) to examine our longitudinal qualitative data set.

Statistical methods

Characteristics of the setting and flow of participants through the study

Overall, 337 people were recruited into the cohort study. It was later determined that two of these individuals, referred into the research by ISVAs and participating up to the 12-month interview, had not actually accessed a SARC; their data were excluded from analyses. Figure 3 shows the flowchart forparticipants in the different stages of the study commencing with screening for eligibility at SARCs.

FIGURE 3.

Flow of participants through different stages of the MESARCH cohort study.

The sample of 335 people had accessed care at 21 different SARCs featuring all four of the models of service delivery under investigation: 67 participants (42%) had attended or otherwise accessed care from a SARC led by private sector companies (10 sites); 115 (34%) accessed care at NHS-led SARCs (five sites); 139 participants (20%) had accessed a police-led SARC (four sites); and 14 participants (4%) had accessed charity-led SARCs (two sites). Six of these sites (affecting experiences of 141 service users) had integrated ISVA services, whereas accessing any of the other 15 SARCs (affecting 194 service users) would have required service users to utilise ISVA care in settings external to the SARC (usually charity sector, but we did have some ISVA services located in sexual health).

There was a high variability in the cluster sizes (i.e. the number of participants linked to each of the 21 sites) ranging from 1 to 104, with an average of 16. This variability was due to varying levels of commitment to the research by SARCs but mainly, by the ISVAs on whom we relied for referrals. Occasionally, a participant self-referred to the research (n = 8), and for one of these individuals, the SARC they attended had not been formally recruited into the project, thus adding a site. Another situation arose when an ISVA referred a person in the usual manner, but we determined that the person had actually accessed a SARC in a different geographic location (i.e. location of the offence/reporting). Five individuals joined the study in this way, generating five additional SARC sites. Thus we had 21 regions/sites altogether, six of which were generated through service users joining rather than the team recruiting the SARC and its referral partners first. The most effective model from a research recruitment perspective was a committed ISVA team co-located at the SARC. The perception in the charity sector that the research was only concerned with SARC practice and related outcomes may have been a barrier to engaging with the project. All agencies were remunerated for recruitment (on a per-client-referred basis).

There was a high degree of variability across the sample in the period between a person’s access to a SARC and participation into the study. For 8% of the sample, a year or more had lapsed since they attended SARC. Given how difficult it is to involve people from a trauma population in research, especially observation research, we had not applied an upper limit on time since SARC access. The duration range was 3 days (for a person who had accessed SARC 3 days earlier for a video-recorded interview) to 5 years (for someone who self-referred to the research). The median duration was 104 days, which was more in line with what we anticipated at the outset, considering most people not being approached at the SARC, waiting for them to be referred to an ISVA service, and for the ISVA to introduce the research at a time they felt was appropriate for their client. Then, it took some time for the research team to undertake the baseline interview.

The first baseline interview was conducted on 18 September 2019, and the final person was recruited on 3 November 2021; the final 12-month follow-up was conducted on 21 December 2022. Baseline interviews took, on average, 1 hour 21 minutes (range: 30 minutes to 3 hours 10 minutes), and most of them were conducted as computer-assisted telephone interviews. The average duration of 6- and 12-month interviews was 67 (range: 20–135) and 64 (range: 15–120) minutes, respectively.

At 6 months, we reached 224 participants, and at 1 year, we reached 222 participants. Thus, overall, the study lost a third of the sample at each time point. However, some individuals not interviewed at 6 months did participate again at 12 months. While we had high levels of loss-to-follow-up at both follow-up time points, intentional withdrawal from the study was low at 9% (see Figure 2 for explanations on why people exited the study). We recorded one adverse event (see Appendix 1, Table 22).

Characteristics of the sample

Baseline characteristics of the total group of participants who accessed SARCs are presented in Table 1. At the time of the baseline interview, participants were aged 18–75 years; 50% of the sample was aged 18–29 years, with one in five participants aged ≥ 45 years. Women constituted the majority of the sample. Overall, 28 men were included in the study; no participants were transgenders, and five people were non-binary or gender-fluid. Overall, 15% of the sample identified as having an ethnic or cultural background other than ‘white’ and included people of dual heritage. Ninety-three per cent of the sample was British, and 95% were residents of the UK. A third had a religious affiliation, that is, Christianity (76), Islam (8) and other religions or beliefs (24). The sample was diverse on indicators of deprivation or opportunity: just under half of people were working full- or part-time and around a half had no financial difficulties. Five per cent of people were in supported accommodation or living in an unstable housing situation. Forty-five per cent of participants (152) had children aged under 18 years.

Poor mental health was highly visible in the sample. ‘Long-term mental health problems’ (see Table 1) reflects the proportions of participants identified by SARC or ISVA services as having a pre-existing mental health problem or condition at the time they accessed SARC. If this was missing at referral, we obtained it in the interviews based on medication and psychiatric histories. The proportions for ever having attempted suicide or self-harmed were similarly high at 53% and 61%, respectively. Eight per cent of the sample had attempted suicide in the month before the interview, and 16% self-harmed in that period. It is noteworthy that nearly 70% of those with longstanding illness or disability that limited activity also had long-term mental health problems.

In terms of offence characteristics, 13% of participants had reached the SARC for help after exposure to sexual abuse during childhood; a quarter of people wanted to seek help for sexual violence perpetrated by partners or ex-partners; and the majority (60%) of people in the study sought help after assault by strangers or acquaintances. Only under 5% had been victimised as adults by family members other than partners or by persons in positions or trust or authority. While most people reached SARCs through referrals by the police, over a fifth of participants had contacted SARCs themselves. The rest of the referrals were by the charity/voluntary sector (6%) and a small number from general practice, sexual health, mental health, education and social services.

The burden of trauma across the life course is summarised in Table 1. The mean number of ACEs was 6 (range: 0–13) across the 334 participants who answered the questions, and nearly four-fifths indicated exposure to four or more adverse experiences. As indicated above, more than 13% of people sought help in relation to CSA; however, our research found that more than half of the sample had exposure to contact sexual abuse in childhood. Exposure to domestic abuse in the household during childhood was another commonly endorsed ACE, affecting half the sample. Nearly 80% told us they experienced domestic abuse at some point in adulthood including a third in the year up to baseline. We examined whether individuals retained (n = 222) and lost (n = 113) at the 12-month follow-up differed on any of the baseline characteristics. A greater proportion of those lost to follow-up were out of work than those who were retained. Table 2 presents the mean levels on the various assessments of health and well-being in the sample at each time point. For our primary outcome measure, that is, the PCL-5, 71% of participants had scores above the cut-off of 33 for ‘probable PTSD’ at baseline. This rate persisted at 6 months (69%) before reducing to 55% at 12 months. This reflected a mean difference (reduction) of 9 points on the PCL-5 [95% confidence interval (CI) 6.6 to 11, Cohen’s d = 0.53] equating to a MCID change based on the guidance used for the study. Although there were trends towards better health across the measures, the most visible change was on PTSD. Modelling of changes on other outcomes will be reported elsewhere.

Direct assessments by survivors about quality of care

Survivors’ assessments of harm and benefit of different sectors for meeting the needs of survivors

The research identified levels of harm and benefit derived across sectors in relation to meeting health and social care needs after sexual violence and abuse. Overall, there was no difference in participants’ evaluations of benefit and harm between the SARCs and voluntary sector (see Table 3).

TABLE 3 - Harm and benefit of care and support across sectors

A&E, accident and emergency; CPS, Crown Prosecution Service; IAPT, Improving Access to Psychological Therapy.

With the exception of SARC, which was most appropriately rated by participants at baseline, sector scores represent participants’ responses in the 6-month interview. In cases where 6-month data were absent, these scores were supplemented by 12-month and baseline data.

	Types of services accessed/counted	Number of respondentsa	Minimum	Maximum	Median	Mean	SD
Sexual assault referral centres	Immediate remote and in person crisis care, forensic services, sexual health care, emergency contraception and pregnancy testing, referrals to (on-site or off site) ISVAs and other services, some aftercare and follow-up	292	−100	100	90	75.6	33.6
Policing and justice	Any contact about the case with police officers, CPS, HM Courts and Tribunals Service	270	−100	100	50	29.6	68.3
Voluntary sector services	ISVA care provided through the charities, brief and long-term counselling services, support groups, domestic violence services, drug and alcohol, helplines people may have accessed over the course of the study period	264	−100	100	90	72.0	40.9
NHS care	GP and other primary care, IAPT, sexual health, crisis and community mental health, A&E, in-patient care, psychiatry	257	−100	100	75	51.9	56.1

The figures below provide further insight into the distribution of participants’ scores. We have presented median scores given the non-normal distribution of people’s responses. The median score is visible in the figures by the black horizontal line, indicating that 50% of the ratings lie above the line and 50% below it. In Figure 4, half the participants rated SARC care above 90 on our scale, with only four people of 292 rating SARCs as having harmful effects. Similarly, participants rated the voluntary sector positively, with 50% of people having a score of  > 90, and 11 people identified harm (see Figure 5). The ratings by survivors for the responses by the NHS and policing were extremely polarised. For NHS care, half of the scores were rated above 75, which indicates many people gained a high degree of benefit; however, 15% of observations were experienced as ‘harmful’ to survivors (see Figure 6). With regard to police and justice, 50% of responses were rated above 50, suggesting many people found benefit through their experience of interacting with the police (see Figure 7). However, a quarter of survivors rated the police and justice response as harmful.

FIGURE 4.

Evaluating sexual assault referral centres care in meeting needs of survivors.

FIGURE 5.

Evaluating voluntary sector care in meeting the needs of survivors.

FIGURE 6.

Evaluating NHS care in meeting needs of survivors.

FIGURE 7.

Evaluating policing and justice care in meeting the needs of survivors.

Main model

Unconditional mixed-effects models

Figure 8 shows all available data on PCL-5, with each line representing the observed data from one participant. While there is a sense of a downward trend in symptoms across the follow-up period, there is substantial heterogeneity in both scores at baseline and the nature of this trend in terms of its magnitude and sign. Furthermore, by plotting follow-up data in terms of months-since-baseline rather than by wave, there was variation in the timing of subsequent waves, particularly for the final wave.

FIGURE 8.

Plot of all the repeated-measures data for the PCL-5.

Table 6 shows fixed- and random-effect estimates obtained from a trio of random intercept/slope linear mixed-effects models.

TABLE 6 - Parameter estimates from unconditional mixed-effects models for repeated data on PCL-5

ICC, intraclass correlation coefficient; N, number of participants in model; Obs, number of observations in model; σ², residual (level 1) variance; τ₀₀, intercept variance (spread at baseline); τ₁₁, slope variance; ρ₀₁, correlation between the intercept and the slope.

Predictors	PCL-5 (complete sample)			PCL-5 (2 + Measures)			PCL-5 (maximal sample)
	Estimate	95% CI	p-value	Estimate	95% CI	p-value	Estimate	95% CI	p-value
(Intercept)	44.9	42.5 to 47.3	< 0.001	44.9	42.9 to 46.9	< 0.001	44.2	42.4 to 46.0	< 0.001
Time year	−8.5	−10.6 to −6.4	< 0.001	−8.1	−10.0 to −6.2	< 0.001	−7.8	−9.6 to −5.9	< 0.001
Random effects
σ2	89.55			90.61			90.68
τ00	196.75 ID			189.71 ID			189.32 ID
τ11	73.83 ID.time_yr			79.02 ID.time_yr			78.37 ID.time_yr
ρ01	−0.31 ID			−0.29 ID			−0.29 ID
ICC	0.69			0.68			0.68
N	182 ID			258 ID			333 ID
Obs	546			698			773

When we focused initially on the complete-case sample of 182 participants, we observed an estimated mean symptom score of 44.9 points at baseline and an average improvement (i.e. reduction) of 8.5 points across the 1-year follow-up.

The baseline random effect has a variance of 196.8 (τ₀₀) corresponding to an SD of 14.0 indicating that we would anticipate 95% of the participants to have symptom scores between 16.8 (i.e. 44.9 − 1.96 × 14.0) and 73.0 (i.e. 44.9 + 1.96 × 14.0) at the first wave. The variance in slope is also substantial (τ₁₁ = 73.8) corresponding to an SD of 8.6 indicating that we would anticipate 95% of the participants to have an improvement in the range of 25.3 to −8.3 across the year of follow-up. The other role of the random-effect SDs here is that they link back to the original power calculation and the quantification of the MCID. The MCID for symptoms scores at baseline is at 14.0 × 0.5 = 7 points on the PCL-5, and the MCID for improvement across the year is approximately 4.25. Finally, it is worth commenting on the correlation between the two random effects (ρ₀₁ = −0.31), which is modest in size, and negative. Thus, there is a tendency for the variance in score to decrease over time, or in other words, greater improvements are observed among participants who entered the study with a greater burden of symptoms.

Changes are subtle if we move to the right in the table and introduce the participants who provided an incomplete set of PCL-5 measurements. We can observe the average yearly improvement reducing slightly to 7.8 points, which is to be expected given the previous observation that participants providing one measurement had slightly lower scores at baseline, and the observation above that there is a negative correlation between baseline scores and the rate of improvement.

Multivariable models

In the multivariable models that follow, we consider each factor of interest and estimate its association with the baseline and change in PCL-5 (i.e. intercept and slope random effects in multilevel-model terminology). This is achieved by introducing the risk factor as both a main effect and an interaction with time. To aid interpretation, we present parameter estimates in terms of two models – a ‘baseline model’ and a ‘slope model’; however, all estimates are obtained from a single linear mixed-effects model with the relevant interaction terms.

For all models presented, a series of covariates has been included; however, their own parameter estimates are not shown. The primary goal of model adjustment here is to address potential bias problems due to variation in the timing of response and loss-to-follow-up. The specific adjustments made for each risk factor are listed in the table footnote. Finally, results are shown for the sample of participants providing at least one measure of PCL-5 across the three waves. The analysis sample for each risk factor varies slightly because of a minor amount of missing data, and this is also summarised in the footnote.

Summary of main findings

The cohort study fulfils two major goals of the MESARCH project. Firstly, it addresses questions about the effectiveness of SARCs and the onward pathway of care, adding longitudinal findings to several existing mixed-methods in this field. 21,47,83,84 Second, it enables analysis of health and well-being journeys for survivors of sexual violence and abuse after contact with services. We are not aware of any other UK-based study that has followed-up a sample of survivors of rape, sexual assault and sexual abuse. There are some longitudinal data available in relation to young survivors in the UK, with smaller samples and shorter timescales,85 and a small number of international studies with adults,86–89 with the South Africa longitudinal study conducted by Abrahams et al. being the closest in terms of design. 12,64,65

Participants in our cohort were SARC service users aged ≥ 18 years. They were largely similar to SARC service users nationally,90 although this shifted somewhat in the wake of the COVID-19 pandemic where people accessed SARCs remotely, the time to care profile shifted and likely interacted with how the COVID-19 pandemic led to reductions in some forms of sexual violence and abuse (stranger and acquaintance assaults), increases in other areas (partner-perpetrated and online forms) and a pattern of non-recent survivors, such as survivors of CSA, coming forward for help. 91 Our research captured this varied picture and recruited people approximately 100 days after they had accessed a SARC. Cohort participants were predominantly women (90%) and belonged to non-minoritised ethnicities (85%), and their age ranged from 18 to 75 years.

Survivors reported a high burden of lifetime trauma. Although suffering from ACEs is not deterministic, having four or more ACEs is related to an increased likelihood of experiencing chronic health conditions such as cancer, cardiovascular disease, type 2 diabetes, respiratory diseases and poor mental health. 92 Notably in our study, three-quarters of participants had four or more ACEs, contrasting sharply with rates of approximately 10–15% reported in population-based studies. 92 Nearly half of the survivors had pre-existing mental health problems at the point of accessing SARC, more than half had lifetime experience of attempted suicide, 56% had been prescribed psychotropic medications. Although our sample had a good spread on socio demographic factors, 17% faced major daily financial problems. For offence-related factors, there was weak evidence that a delay of more than 10 days to SARC was associated with more PTSD symptoms, and a similar level of evidence in support of an association for perpetrator type where it appears that when the perpetrator was the survivor’s partner/family the baseline score was greater. The profile of perpetrators differed from ONS data93 on experiences of sexual violence and abuse nationally with a greater proportion of abuse among SARC service user by strangers and acquaintances over partners. ONS data suggest that nearly half of the sexual violence and abuse cases is committed by partners. This signals an access issue for survivors of domestic-perpetrated sexual abuse.

Sexual assault referral centres and third-sector agencies achieved consistently high standards of care for survivors across sub-studies, and harmful encounters were at a minimum. This was in stark contrast to the NHS and police, where the frequency of harmful experiences affected 15% and 25% of people who rated these sectors, respectively, and our qualitative studies have generated many insights into how to improve care in these settings.

Post-traumatic stress disorder proportions and means at entry to our study were consistent with data cited at baseline in international intervention studies with survivors of sexual violence and abuse. 36 Seventy per cent of participants had ‘probable PTSD’ and achieved important reductions at follow-up (9 points on the PCL-5 equivalent to > 0.5SD, which was our threshold for judging a minimal clinically meaningful difference). 69 We found an important triad of risk between ACEs, long-term mental health problems and socio-economic deprivation, and these were the main predictors of PTSD at baseline.

A third of the original cohort was lost-to-follow-up at 1 year, consistent with treatment studies in the field. 71,86 While those retained and lost differed on degree of socio-economic deprivation, people with more trauma symptoms at baseline were not more likely to leave the study. We consider this to be an important observation and reflect the impact of our trauma-informed research practice (see Chapter 2). The service use is described more fully in our economic evaluation (see Chapter 6), but there was strong evidence of a positive association between frequency of service use and symptom level at baseline. This was also consistent with the ACEs-related research on costs. 92

We observed a trend towards improvements on health measures for the sample over the year of the study. While examining in-depth aspects of health in our sample beyond PTSD is outside the scope of this research, we present the QoL data as an example of what is emerging from our work. When using the final timepoint for comparison, the MESARCH cohort sample was found to have poor QoL across three of the four domains on the WHOQoL-BREF relative to the general population. 94 In their study assessing the QoL in people with a variety of illnesses, Skevington and McCrate94 reported the following scores for the WHOQoL-BREF for the control group (a sample of more than 1300 participants who were considered ‘well’): Physical 76.49 (vs. 58 in MESARCH sample); Psychological 67.82 (vs. 49); Social 70.52 (vs. 64); Environment 68.20 (vs. 67). It is noteworthy that there was no difference on environment-related QoL in our sample that connects with important themes of access to health care and safe environments, which we highlight about in Chapter 8.

This study did not detect differences in outcomes on PTSD at 1 year by SARC service model accessed or on accessing a SARC-provided ISVA. This was an important finding and core part of our investigation. Although there was a trend towards benefit from charity-led SARCs over those led by the police, NHS and private sector organisations, we had too few service users accessing charity-led SARCs to have confidence in this finding. Without a control group, we cannot demonstrate causality; however, it is promising to observe improvements among those that have accessed the SARC care pathway. The main predictor of PTSD at 12 months was the baseline score. With 55% of people still exhibiting PTSD symptoms over a year later that would interfere with their daily functioning, there is still a gap in survivors accessing the right care for their recovery. This is crucial to address to avoid future costs of violence, especially given what our Cochrane Reviews have demonstrated in terms of the efficacy of interventions. There is also wide evidence from our qualitative work of a range of novel and alternative treatments and therapies seen as acceptable and beneficial to survivors that may offer cost-effective ways of promoting good health and recovery for survivors. A high ACE profile doubled the odds of meeting the PTSD diagnostic threshold at baseline, this reduced to 1.5 at 6 months and was absent at 1 year. This is a promising sign that the effects of childhood trauma may be alleviated through appropriate care and support.

Strengths, limitations and implications of this work for health care and research are addressed in detail in Chapter 9.

Resource use and costs

For the initial analysis, a UK NHS/Personal Social Services perspective was adopted when conducting the economic analysis; accordingly, only direct costs to the health service, social care and SARCs were deemed to be relevant. This was followed by a wider analysis, focusing on public sector resource use as a whole. Resource use data were collected prospectively to estimate the costs associated with different models of SARC service delivery. This included services attended after initial attendance at a SARC, to capture the whole service pathway experienced by survivors.

The resource use monitored included (1) the cost of service use within SARCs (e.g. consultations, treatment); (2) NHS and social care resource use after initial attendance at SARCs (e.g. GP visits, sexual health visits, contacts with social workers). This was distinct from the resource use within the NHS SARCs; (3) third-sector resource use after initial attendance at SARCs (e.g. counselling, information and advice); and (4) wider public sector resource use, for example, in relation to housing and other social welfare systems.

Relevant unit costs for health resource use and treatments were identified from established national sources103,104 (see Report Supplementary Material 2). The total cost associated with each service pathway, at different follow-up points, was estimated using a micro-costing method where unit costs were assigned to the number of utilised resources. 105 As healthcare costs are substantially skewed, 95% CIs around mean differences were estimated using the bootstrap method. 103 All costs are expressed in UK currency (£) for 2021/2. As necessary, costs were inflated using the Hospital and Community Health Services Pay and Prices Index. 103 Discounting of costs and outcomes was not required as the study was concerned with a 12-month period. As no information was publicly available on the costs associated with services in terms of different types of SARCs (police, NHS, charity and private-led services), we assumed that there was no difference in the costs associated with resource use such as consultations across the various SARC types (NHS-led, private etc.). There is currently considerable debate about the cost burden on different sectors, but no definitive data are available on actual cost differences. However, all costs were varied extensively in the sensitivity analyses.

Health outcomes

The primary clinical outcome for the MESARCH study was participant-reported PTSD symptoms, collected using the PCL-5 (see Chapter 5). Self-reported health-related QoL was assessed via the EuroQol-5 Dimensions, five-level version (EQ-5D-5L) questionnaire, completed by participants at baseline and follow-up points. The EQ-5D instrument is a preference-based measure consisting of five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression. 106 The EQ-5D instrument was selected as it is recommended by the National Institute of Health and Care Excellence (NICE) and has been used for patients experiencing PTSD and in other similar contexts. 107,108 Following NICE guidance utility values were calculated by mapping the 5L descriptive system data onto the 3L value set. 109 The mapping function developed by van Hout et al. 110 was used for the analysis, to allow consistency with NICE recommendations. The mapping function uses data obtained from a survey of the UK population to derive a utility-based value. Following the trapezium rule, the generated score was used to calculate quality-adjusted life-years (QALYs) gained at 12 months, to inform analysis of costs and utility. 111

The ICEpop CAPability measure for Adults (ICECAP-A) instrument was used to measure capability/ wellbeing. The instrument consists of the following five attributes: stability (an ability to feel settled and secure), attachment (an ability to have love, friendship and support), autonomy (an ability to be independent), achievement (an ability to achieve and progress in life) and enjoyment (an ability to experience enjoyment and pleasure). 112 This instrument was included, as there has been wide recognition of the limitations of EQ-5D and other health-related QoL measures in the context of mental health conditions. 113,114 The capability approach is concerned with the evaluation of a person’s ability to achieve ‘functionings’ in life that are important to them. Such ‘functionings’ can range from basic considerations, such as being nourished, to more complex aspects, such as self-respect. Overall, the capability approach attempts to capture a more comprehensive picture of individuals when evaluating programmes or services, rather than just limiting the space to health alone. 115

Changes to protocol

We had originally planned to undertake a cost-effectiveness analysis of different SARC models (based on organisational structure) in terms of the primary clinical outcome which was reduction in trauma symptoms (PTSD-based PCL-5). However, our analysis of the clinical outcome data demonstrated that significant clinical differences in terms of SARC organisational structures were not evident (see Chapter 5). It was therefore judged appropriate to conduct a full economic evaluation in the form of a cost–utility analysis only, focusing on cost per QALY gained. The data are initially presented as a descriptive analysis of costs and outcomes (in the form of an extended cost-consequence analysis), followed by a full economic evaluation in the form of a cost–utility analysis. This allows assessment of service use and costs for participants. It was originally planned to develop a decision-analytic model to evaluate the longer-term impacts of the different types of service delivery on costs and outcomes. However, this was not judged to be appropriate, based on the results of the cohort study, which did not indicate meaningful clinical differences based on SARC organisational structures.

Base case analysis

To compare the costs and benefits of different SARC service delivery models, a within-study analysis was undertaken. A descriptive analysis of the service use and costs associated with different SARC models was undertaken as the first step (a cost-consequence analysis). This involved analysing the service pathways followed by those accessing services. We assessed costs and outcomes in a disaggregated manner for SARC pathways to provide information on the costs and outcomes associated with each model. The main analysis was based on the reported health-related QoL, examining changes at 6 and 12 months. The secondary analysis examined changes in capability.

As shown in the previous chapter, analyses of clinical outcomes demonstrated that there were no significant differences in terms of the primary clinical outcome, which was a reduction in trauma symptoms (PTSD-based PCL-5) based on SARC type. Hence, a cost-effectiveness analysis based on the primary clinical outcome for the study was not deemed appropriate. A full economic evaluation in terms of a cost–utility analysis was performed to compare the cost per QALY gained at 12 months for different SARC types. Following an assessment of the level of missing data, a complete case analysis was undertaken for the cost–utility analysis, and only participants who returned the resource use and EQ-5D question elements at all time points were included in the analysis. It was not possible to employ multiple imputation methods because of the extent of missing data and a complete case analysis needed to be undertaken to explore changes in resource use and costs over time. 116 A similar approach as that outlined in Chapter 5 was adopted to analyse differences in characteristics between participants at baseline.

Sensitivity analyses

A series of one-way deterministic sensitivity analyses were carried out to assess the impact of the assumptions and estimated values included in the base case analysis. Uncertainties around costs and outcomes were analysed, and plausible ranges were specified using information from the study and from the literature. To account for uncertainty due to sampling, a probabilistic sensitivity analysis was undertaken involving a non-parametric bootstrapping approach, applied to derive paired estimates of mean differences in costs and utility values. The paired estimates were presented on scatterplots for the cost–utility analysis on the cost-effectiveness planes. STATA (StataCorp LP, College Station, TX, USA; version 17117) was used to perform the analyses.

Health outcomes

The analysis of the primary clinical outcome showed that there was no statistically significant difference based on SARC model type in terms of a clinically significant reduction in PTSD symptoms (see Chapter 5). Hence, this section focuses on describing the data on health-related QoL (HRQoL) collected during the study. The mean EQ-5D value at baseline was similar across all SARC types at baseline and improved over the 6- and 12-month periods, although remaining below population norms118,119 (see Table 12).

The analysis of mean QALY gain over the 12-month period was slightly higher for participants in police-led SARCs, but there was no statistically significant difference between the SARC models in terms of QALY gain (see Table 13). The results for the ICECAP-A instrument (will be reported elsewhere) showed that capability increased over the 12 months for the study participants.

Sexual assault referral centre and Independent Sexual Violence Advisor service use

As shown in Table 14, a wide range of services were accessed in all SARCs at baseline (in addition to FMEs, where appropriate). Data on SARC service use were available for 325 participants (97.0%). A similar level of service access was reported across SARCs. SARC service provision reduced over time, with very few participants reporting access to SARC services at 6 and 12 months. Table 15 shows the high level of access to ISVAs by participants in the study with 4056 contacts across the three time points, suggesting around five contacts per participant per timepoint.

NHS resource use and costs

Of the 335 participants interviewed at baseline, 225 reported accessing NHS care at their baseline interview (i.e. distinct from any health care accessed in their SARC visit). As shown in Table 16, the main form of care accessed was related to GP visits, followed by sexual health centre visits. A high proportion of the contacts related to mental health services and when the range of mental health services are combined, the rate of access was similar to primary care. Service use after the initial SARC visit was similar across the different models of SARCs. Overall, reported use of NHS services had decreased at 6 months. Once again, reported service usage mainly related to GP services, with continued access of mental health services. NHS service use further decreased by 12 months. The examination of costs associated with NHS service access across all time points showed that costs were similar across models (complete case analysis; see Table 17).

Third-sector and broader public sector costs

Reported resource use for other public sector services is detailed below. A wide range of services were accessed by participants at all time points, with reported levels of use decreasing over time. For policing and criminal justice services, contacts with police officers were the main contacts reported at all time points (see Table 18). For third-sector and other public sector resource use, see Report Supplementary Material 2. A broad range of services were accessed, with reported use decreasing over the time points.

Cost-utility analysis

The results of the cost–utility analysis are presented in Table 19. The police-led model resulted in slightly more QALYs but was slightly more costly than the NHS-led model. However, the differences in costs and outcomes were not statistically significant (95% level).

As a secondary analysis, broader societal costs were explored (see Report Supplementary Material 2). This assessed costs incurred by participants and their families. Impacts on participants and their families were substantial. For example, respondents reported important effects on their employment and productivity.

Sensitivity analysis

For details of the deterministic sensitivity analysis, see Report Supplementary Material 2. For the probabilistic sensitivity analysis, the results of 5000 bootstrap replications were plotted on a cost-effectiveness plane for the cost–utility analysis (see Figure 9 and Report Supplementary Material 2). This analysis suggested that there was uncertainty around the differences in costs and outcomes between the SARC models, as replicates were distributed almost equally across all quadrants (in terms of both effectiveness and cost). Accordingly, there is uncertainty around whether different SARC models are more or less costly than the alternative models and whether they are more or less effective.

FIGURE 9.

Cost-effectiveness plane (NHS vs. private-led SARCs).

Summary of main findings

The data collected in the study suggested that participants accessed a wide range of healthcare services and other public sector services in the period between attending SARCs and joining the study, but over time, service use decreased. The analyses undertaken suggested that the QALY gain and costs between SARC models were similar. The cost-effectiveness plane constructed to incorporate uncertainty around each point estimate in the results shows that there is considerable uncertainty around any estimated differences in costs and outcomes. This is similar to the findings presented in Chapter 5, which suggested that there was little evidence to suggest a relationship between PTSD as measured by the PCL-5 and the SARC type.

However, this study has only focused on those who have accessed SARCs. The lack of a comparator arm means that it is not possible to assess how resource use and QALY gains compare for survivors who did not access such services. Those who experience rape or sexual assault have a higher risk of developing PTSD and that PTSD is more likely to be of a longer duration. 120 Although economic evidence is limited, it is suggested that there are high levels of public sector service use for those who have experienced sexual violence121 and for those experiencing PTSD. 122 This study revealed that there were also substantial economic impacts on participants and their families (e.g. time off work). The results of this analysis suggested that access to services and costs were similar across SARC models. Participants reported improvements in HRQoL over the 12-month period under consideration, although this remained below population norms. 118,119 The absence of a comparator arm means that it is not possible to compare HRQoL for those accessing SARCs with those who did not access SARC services. Although the research base is limited, there is evidence to suggest that a high proportion of those who develop PTSD, are likely to face long-term impacts on HRQoL. 123

Strengths and limitations

This is the first analysis of health outcomes and resource use for those accessing SARCs in England. Detailed data on resource use and health outcomes were collected. The analysis also considered the broader costs for participants and their families associated with experience of sexual violence. Given the limited literature focusing on the economic evaluation of services and programmes to address sexual violence and abuse, the study provides useful data to inform decision-making in this area.

The main limitation of the analysis was the level of missing data on resource use and QoL at all time points. Accordingly, multiple imputation methods could not be employed because of the extent of missing data and a complete case analysis needed to be undertaken. 116 Excluding those who did not complete the resource use questionnaire at all timepoints may have introduced bias and some level of inefficiency. The bias may be based on the finding that certain groups (including more disadvantaged groups) were more likely to be lost-to-follow-up. Furthermore, among those who were retained, the resource use section of the interview was the section more often omitted by field workers in situations where the person reported lacking time or that the interview was too difficult for them to complete. In this sense, the study may have underestimated service use. Caution is also needed in interpreting the level of service use at baseline; while it covers a period of approximately 3 months (median was 104 days), there was a high variability in the timeframe between SARC access and the baseline interview. Another limitation was that the follow-up was only for 12 months, which may not have allowed longer-term impacts to be fully captured.

Recommendations for research

The data collected in the study suggested that participants needed to access a wide range of healthcare, third-sector and other public sector services. There is a gap in understanding about the extent to which SARC service users access voluntary sector care and support in the aftermath of sexual violence and abuse. The study reports on important aspects of resource use in the voluntary sector for SARC service users, specifically, charity-led SARC services and voluntary sector-provided ISVAs. However, a full analysis of the voluntary sector resource use by SARC service users (e.g. helplines, support groups, counselling, advocacy services, domestic violence services) was beyond the scope of the current research. This is a gap our research will fill in future analyses. There is a need for greater understanding of the costs and outcomes for SARC users compared with those who do not access SARC services, which was not possible in this study. It is also important to assess the longer-term service use and health outcomes for those who access SARC services (beyond 12 months). In addition, there is a need to analyse health outcomes and costs for different population groups, which was beyond the scope of this component of the study.

Study design overview

We conducted semi-structured interviews with CYP about their experiences of attending a SARC and what their recovery looked like afterwards. Qualitative inquiry was important to enable CYP to articulate their own experiences. A focused approach to study recruitment was used (see below), recruiting CYP who attended one of two paediatric SARCs in two geographical regions in England. These were sites where the team had built relationships to work closely with three linked third-sector organisations, ensuring that CYP were supported before, during and after the study.

Changes to protocol

The intended sample size of 40 as proposed in Protocol V3.3 was not fulfilled. Challenges to recruitment are discussed later in this chapter.

Theoretical approach

Our theoretical approach was underpinned by the United Nations Convention on the Rights of the Child,132 whereby all children have the right to make their voices heard in matters that concern them, while having the right to protection from harm and exploitation. 133 We recognised the importance of ensuring that the research did not lead to children experiencing further harm, in light of the potentially long-term and wide-ranging detrimental impacts of experiencing sexual abuse during childhood,134–136 while not wanting to exclude CYP from the research opportunity. 131,137 The approach was also informed by theoretical developments in the sociology of childhood138 viewing children as those able to make meaningful contributions towards issues that affect them, rather than being ‘objects’ of enquiry,139 thus researching with children, as has been acknowledged in inter-related fields. 140

This study was underpinned by Critical Realism141 acknowledging that individuals or societies socially construct independent worlds and there are different yet valid perspectives about ‘reality’. 142 Critical Realism recognises that there is an independent reality as well as a subjective interpretation of it,143 which is influenced by individuals’ interpretation of this reality at individual and social levels.

Sampling and recruitment

Recruitment relied on SARCs and third-sector organisations identifying eligible participants and connecting them to the research team. When appropriate, the professional introduced the study to the CYP and provided the leaflet, participant information sheet (PIS) and directed them to the recruitment video. With the consent of CYP interested in participating (and the parents of those aged 13–15 years), contact details were securely shared with the research team who contacted CYP to discuss the study further. Those who agreed to participate (and the parents of those aged 13–15 years) provided informed consent. Eligible individuals were 13–17 years old in person SARC service users between August 2019 and December 2022. CYP were not approached if the assault occurred ˂ 1 month ago and/or if, in exercising clinical judgement, the professional at SARC or third-sector agency anticipated the CYP might encounter difficulties in providing informed consent or understanding the interview content due to mental or physical health issues, cognitive impairment, learning disability or poor English language skills. CYP were not approached if they were known to be in an unstable or unsafe situation. We were prepared to overcome barriers to participation, for example, through the use of interpreters. It was discussed with third-sector organisations that it might be appropriate for them to provide additional support to CYP as someone who was present in the research interview if the CYP requested, or being available for a debrief or follow-up support afterwards. Two SARCs and three third-sector organisations passed on details for 31 CYP, and we enrolled 12. The first CYP research interview occurred on 3 August 2021, and the final interview was conducted on 27 January 2023. Of the 19 we did not enrol, some no longer wished to participate (n = 6), some arranged the interviews but later changed their minds (n = 3) and we received no responses or communication from others (n = 10).

Recruitment challenges and strategies

The study initiation was delayed due to COVID-19. As the SARCs were NHS-based, NHS studies that were not yet ‘live’ or not primarily focused on COVID-19 were not allowed to progress. Furthermore, the third-sector organisations initially had no capacity to assist due to the impacts of COVID-19 pandemic on their services. Once the SARCs and third-sector organisations had capacity, the research team conducted training sessions to equip sites in their understanding of participant eligibility. The research team worked with organisations to identify strategies and appropriate times for them to facilitate conversations about the research with CYP. There were occasions when SARCs and third-sector organisations regularly shared contact details; however, there were often periods when this was less frequent. This meant that recruitment was slow, as there was a small pool of individuals from which we could approach at any given point in time. Some CYP were quickly enrolled into the study while it took longer to enrol others.

Recruitment challenges also related to the service landscape of third-sector services. While one organisation experienced major staffing shortages exacerbated by COVID-19 pandemic, staff found it helpful to meet with the research team every 6 weeks to discuss recruitment progress. As another organisation had staff members join their team at various junctures, we attended their team meetings at different time points to explain the research opportunity. In the final year of recruitment, a third-sector organisation had capacity to assist, enabling a wider pool of CYP to be approached.

Participants

One boy (cis-gender) participated along with 11 adolescent girls (cis-gender). At entry to the study, three were aged 13–15 years, two were aged 17 years and seven had just turned 18 years. At access to SARC, seven were aged 13–15 years and five were aged 16 or 17 years. Ethnicities included White British (n = 9), Indian British (n = 1), Black British (n = 1) and White British and Black Caribbean heritage (n = 1). While no CYP was known to have a physical disability, two were neurodiverse (attention-deficit hyperactivity disorder and autism diagnoses). At the time of participating, all CYP attended secondary school or 6th form/college. Eleven participants lived with non-abusive parents and one lived independently. All CYP were referred to SARC by the police and seven had a FME. Some also received remote support from the SARC. At the time of accessing SARC, the time since the index offence took place varied among the sample: ˂ 7 days (n = 3), between 7 days and 6 months (n = 4), more than 7 months (n = 1), and over 12 months (n = 4). We spoke to CYP for 2 months to 3 years after they interacted with the SARC.

Information provided by SARC indicated that across our sample, the index offences involved nine rapes and three sexual assaults and the perpetrator included a family member (n = 4), peer (n = 7), and someone in a position of authority (n = 1). Criminal justice outcomes included one withdrawal (due to the perpetrator being re-interviewed), eight cases were dropped by the Crown Prosecution Service or police and three were under police investigation. There were no recorded prosecutions in our sample.

‘Conversations’ with children and young people

We conducted interviews that we referred to as ‘conversations’. We were mindful about language as we did not want CYP to compare the research to a ‘police interview’ as the research was not about them providing an account of the abuse experiences. We also wanted dialogue that was accessible and comfortable. By adopting qualitative participatory methods, we wanted CYP to have the opportunity to articulate their experiences in their own voice. 144,145 CYP were encouraged to challenge misrepresentations, clarify and interpret comments they and/or others made, thus increasing collaboration. All CYP chose to engage in dialogue during the ‘conversation’, and one young person additionally shared a drawing, as she felt this better demonstrated aspects of her journey. Conversations lasted for 30–85 minutes (mean: 49 minutes).

Analytical approach

We adopted Braun and Clarke’s Reflexive TA approach,56–58 which can be used across a range of research questions, sample sizes and data collection methods. 146 The approach complemented our Critical Realist perspective. 141 We used Braun and Clarke’s six phases of TA: (1) familiarisation with the data, (2) coding, (3) generating initial themes, (4) developing and reviewing themes, (5) refining, defining and naming themes and (6) writing-up. 56 Analysis was recursive, enabling movement back and forth between phases. Transcripts were coded by hand. GC and FL who were involved in speaking with CYP remained reflexive throughout the research, considering how preconceptions, values and beliefs impacted how the study was conducted, and how data were interpreted and presented. 147 The iterative process of collecting and analysing data meant that GC and FL checked the correspondence between the developing analysis and the perspectives of CYP during conversations, as well as exploring unexpected findings. 148 GC, FL, ELW and one expert-by-experience (see Chapter 2 for lived experiences methodology) coded the transcripts, generated themes and refined them and drafted the study write-up.

Children and young people’s experiences of SARCs

Children and young people had positive experiences at SARC, acknowledging its key role in enabling holistic care through conducting FMEs, providing medical care that provided reassurance, and offering referrals to organisations for further support. Many CYP told us that they did not have prior awareness of what a SARC was until they disclosed to a trusted individual and reported to the police. One young person had prior awareness through social media, where they had learned of others’ experiences of what SARCs were like:

I had heard of it before. I got most of the information actually from Tik Tok, where others were all telling their own stories, but both good and bad. I would say my experience is better than a lot of the ones I’d heard of.

Serenity

Children and young people reported that although they had been anxious when they first attended the SARC, the staff made them feel comfortable and welcomed: ‘At the start I felt a bit nervous, I didn’t obviously know what was going on properly. I didn’t really want everyone to look at me … but then by the end I felt a bit more comfortable’. (Ben)

Children and young people appreciated SARC staff communicating clearly to them before and throughout the FME:

They explained it very well what was going to happen … like the whole swabs and how it’s going to work. She said to me, ‘It’s not going to hurt but you might feel a little bit uncomfortable’.

Aurora

I had a full body examination done for bruises and marks. I had a couple of STD tests. I had another internal examination for bruising on the cervix. They’d show me what they were using, and they let me know every step of the way what they were doing.

Serenity

Children and young people valued having opportunities to make decisions during the FME:

They gave me the option for the support worker to stay in the room as well, like behind the curtain to talk to you, to distract you, so that was really good, that helped.

Somaya

The SARCs in our study were based in hospitals, and the CYP had mixed opinions about how suitable this was. Some appreciated the hospital environment, as it provided a sense of anonymity. Others thought it felt clinical and could have been more welcoming:

When I kind of turned up to the SARC, it made me nervous because it just gave me hospital- prison vibes. I think I would give it more kind of homely, I know obviously they have to be like professional and medical, but I don’t know if there was something more kind of welcoming.

Lola

Children and young people highlighted the importance of being able to take literature from SARC home that they could read when they felt in a better place to engage with it. They also valued receiving a care pack ‘It’s like a massive relief the way they told me that I can go home with this really good quality stuff. And I could relax myself, I could clean myself.’ (Serenity)

The service provision landscape after SARCs

Children and young people believed their recovery was impacted by the service landscape in which support services were provided post-SARC; this led to many experiencing continued poor mental health and further trauma symptoms. Long waiting lists for counselling delayed CYP accessing support. Some suggested that time spent on waiting lists could be better managed by services, so that their mental health did not worsen. One CYP has been presented as an example of this.

There was this lady that would call me once a month when I was on the waiting list. That happened 4 or 5 times and she sent me workbooks. They had lots of quotes, powerful messages and activities to do, like naming things you’ve achieved since what happened to you, getting you to think about positives. I found them helpful.

Lola

Some CYP described how they accessed support through digital apps, which they engaged with at their own pace particularly while waiting for counselling. CYP often found these apps themselves and thought they could help them manage physical and mental health impacts associated with their trauma.

For some, long waiting lists for counselling from third-sector organisations led to them starting counselling provided by their school. However, this still had its limitations:

I think they said that it would be a long wait, I think it was like a year and a half. However, the school gave me counselling, but they couldn’t be very helpful because I was in my last year, and I didn’t know the counsellor and I didn’t even do that for long because I was in the last year of school.

Somaya

Those who had positive experiences of support often felt that its duration was too short. This was particularly difficult when CYP had developed positive relationships with those supporting them. CYP felt frustrated about being unable to share details about the abuse with their therapist when their case was going through the CJS. Thus, long waiting lists, short-term support and criminal justice guidance impacted their health and well-being.

Children and young people felt that there should be considerable financial investment into services to educate people about sexual abuse and support those who had experienced it:

I think there needs to be more funding put into recovery groups for people who have been through this kind of thing. People who have been through this kind of thing need a way to rebuild their identity and give them that sense of community. More projects that work around the stigma of boys who have been sexually assaulted. Projects to educate people and then, projects to help rebuild people and their self-esteem after this has happened.

Clara

Organisations and systems working together

Children and young people emphasised the importance of professionals working well together within and between organisations at SARC and post-SARC, as this had implications on the support and care they received.

The staff seemed to work well as a team. They understood each other and the comfortableness I felt, or I could see with them, was put onto me. I felt like I could very much trust them. They were very patient with me. They really understood how it was for me.

Serenity

One young person explained how a staff member in school worked with a third-sector organisation to ensure she had a safe space to engage in virtual therapy sessions during school hours. Working together in this way enabled her to receive support seamlessly. CYP highlighted the importance of joint working within education settings, particularly when the topic of sexual violence and abuse was part of the school curriculum.

The well-being officer was messaging my teachers for me, she was like, ‘Oh she’s feeling a bit like this today, maybe not do this, or maybe not pick on her as much in class, like don’t get her to put her hand up all the time’. She helped me to work with my teachers, especially as I do health and social, and we had to learn about sexual abuse.

Yolanda

Many described the critical role of their ChISVA who provided them with updates and explained criminal justice processes because of their good relationships with other organisations such as the police. CYP felt that their ChISVA fulfilled their role as their advocate.

However, CYP told us how there were missed opportunities for support even when professionals tried to work with other agencies.

I remember there was one time I was with my detective. She casually said in conversation, ’Oh you should tell your social worker this, this and this’ and I was just like, ’What social worker?’ And she was like, ’When I set up your case, I arranged for a social worker to be in contact with you. You should have had one from the start’. I was like, this whole time she thought I had a social worker because she had said to them, ’You need to go do this’. And nothing ever happened.

Clara

Children and young people emphasised the importance of inter-agency communication by professionals following a disclosure of sexual abuse. Aurora shared how her teacher responded by following safeguarding processes, but it was not communicated how quickly her parents and the police would be notified.

Obviously knowing about the whole sexual assault she [teacher] couldn’t keep it to herself. And I think that she had to call safeguarding and obviously safeguarding had to call the police and then she called my mum about it. I knew that she would have to tell someone because obviously I’m under 18. But I didn’t realise it was going to be like, super, super quick.

Aurora

Children and young people valued healthcare professionals working together so they did not have to explain their experience of assault multiple times. Somaya shared how the SARC communicated with her GP, but as the GP did not acknowledge her sexual abuse when prescribing her medication, this led to her feeling as though the GP was not concerned with her overall health.

SARC had told my GP. It says on like my history or something now. They just kind of without questions sort of prescribed the medication. Like they could have asked me and gave me the option and like tried to check my overall health or tried to like support before prescribing. I don’t know, like if they maybe like cared about my overall health rather than just being like, ‘Oh this is for you’.

Somaya

Children and young people expected professionals to work together to help them build their support system by signposting them to other local services, ranging from housing to sexual health. However, CYP and their parents sometimes found it overwhelming and confusing when many services contacted them at once. Thus, CYP recommended a more joined-up approach between organisations.

Principles underpinning a ‘trauma-informed’ approach

Children and young people’s healing was influenced primarily by their experience of interacting with professionals. CYP provided examples of positive interactions and ones that felt damaging. CYP described the following principles that can be seen to be promoted in ‘trauma-informed’ approaches to supporting survivors.

Ideal characteristics or qualities of professionals

Children and young people identified key qualities that they believed professionals should have, all stemming from the principles identified by CYP which could be described as ‘trauma-informed’. While CYP felt that professionals could be trained in these, they recognised these could be inherent qualities within a person, which should be considered when recruiting professionals into job roles.

Children and young people’s recovery outcomes

Children and young people shared how abuse impacted their lives and how their experiences of accessing support helped to improve these effects. Factors such as being unable to access support or re-victimisation exacerbated these effects. One important outcome was CYP gaining understanding of sexual violence and abuse and how it affected them, knowing it was not their fault and they were not to blame. By receiving appropriate support, CYP found it beneficial to recognise and confront what happened to them.

When your family blame you, your school blame you, it feels like it is your fault, but then when somebody finally goes, ‘it wasn’t your fault what happened’ you kind of feel at peace with yourself.

Ashleigh

This led to CYP experiencing improvements in their physical and mental health, their education, interpersonal relationships and being able to re-engage in a wide range of areas in their lives. CYP valued being provided with tools and techniques they could use once formal support ended.

I think the support has made me more capable of looking after myself. I can notice more in myself, I can identify my emotions and the real reasons behind them. I think I’m more calm. Before I would have outbursts a lot and I’d get really angry quite quickly.

Yolanda

Recovery is ongoing

Children and young people explained how their healing was ongoing, and this was not the same for everyone. CYP recognised that there were time points when they felt ready to start receiving support and times where they did not. Some felt tensions between not feeling ready to engage with support, while being afraid that support may no longer be available if they did not take it up. Although CYP knew their rights to access support at another point, some felt unable to ask for that support again if they had previously refused it. CYP also highlighted the significance of being prepared for formal support to end and having a sense of closure as part of their recovery:

We really got on and then it sort of feels a bit like a break up that you’re never gonna see her. But then, when she sent the letter that was nice because it sort of put it all together, like how much we’d done and accomplished.

Lola

Motivation to make changes for others

Children and young people’s experiences of SARC and support post-SARC motivated them to engage with opportunities where they could be heard. Some support services provided CYP opportunities to participate in a consultation group once formal support ended, where they could influence service delivery. This enabled CYP to remain connected with services and make changes for others.

Children and young people told us that they felt motivated to participate in our research study for a number of reasons. Because of CYP’s positive experiences with services, they wanted to ensure that others received positive support.

I feel like my motivation has just been to help others. When it happened I was in a very down place and I’m now a much happier, more creative, just independent person now. I want people to know that it’s not gonna be like horrible for the rest of their lives.

Polly

For others, negative experiences of support or areas where they felt could have been improved fuelled their motivation to ensure changes were made for CYP in the future:

I hate the thought of someone I know to go through this and have the same experience, where actually one small change would have made the broader difference. I’m keen to give my opinion because I wish things were different for me in some areas.

Sophie

I know that I will be at least making some difference and I’m giving my time, my thoughts and my energy to this research. I want to campaign for change. My own experiences of the criminal justice system made me passionate about helping people.

Clara

Summary of main findings

This study is the first of its kind to explore CYP’s experiences of SARCs and their journeys post-SARC. Our findings demonstrated the uniqueness of recovery in this group of survivors. Recovery was influenced by how quickly CYP accessed support, its content and duration, the location and context of care and how well services worked together. Healing was also affected by life events that CYP experienced which included sexual re-victimisation. Re-exposure to abuse served to aggravate trauma symptoms and undermine engagement with services.

Children and young people had positive experiences of SARCs and valued their role in offering them a FME and connecting them with onward support services. As the majority of CYP in the study did not have prior knowledge of SARC, this raises questions about SARC awareness among younger populations and the importance of professionals’ awareness of SARCs when responding to disclosures of sexual violence and abuse. Among the majority of those we spoke to, SARCs played a role in responding to recent sexual offences, which relied on CYP disclosing abuse. In light of the costs to the NHS of delayed disclosures of sexual violence in childhood and lifelong impacts for survivors,149 it is paramount that professionals are prepared to recognise and manage disclosures effectively.

Children and young people felt that their recovery was impacted by the limitations associated with the broader service provision landscape of services accessed post-SARC. Accessing delayed and short-term support was the norm. There continues to be an urgent need to invest in resources for services to minimise the negative health and well-being impacts for young survivors that are exacerbated by not receiving timely and long-term support. Our findings support other research that has shown young survivors managing their own mental health and well-being while waiting for support. 130 While adequate resourcing of services may take time, CYP in the study suggested that there were benefits of services managing waiting list times such as providing psychoeducation.

Children and young people identified principles that were key to their recovery which could be described as principles underlining a ‘trauma-informed’ response to supporting survivors. 44 Recovery outcomes not only were dependent on timely support but were also impacted by CYP’s interactions with professionals at SARCs and across sectors, including specialist sexual violence services, statutory health, education, social services and the CJS. Thus, trauma-informed principles and certain qualities associated with professionals such as showing empathy and compassion, not judging CYP, providing a personalised response and empowering CYP, all play a role in CYP achieving long-term gains in the aftermath of sexual violence and abuse. These principles and qualities reinforce key messages from our qualitative evidence synthesis (see Chapter 3) about what factors impacted survivors’ engagement with interventions and recovery outcomes.

Notably, the lack of a ‘trauma-informed’ approach was associated with poor or delayed healing. Therefore, early intervention and a consistent trauma-informed approach across professionals could prevent long-term negative health and well-being impacts for young survivors. Further exploration is needed to distil what ‘trauma-informed’ practice looks like across sectors that are specialist SAAS agencies and those that are not (e.g. education and NHS services) and how professionals can be trained. This supports observations that demonstrate that adult survivors’ contact with support services alone does not reduce trauma symptoms; rather, it is the quality of services and survivors’ assessments of their interactions with professionals that appear relevant in promoting recovery. 150

The study also observed that when the opportunity is provided, CYP valued being part of informing and developing services to help others. For some, this aspect played a key role in their recovery journey. As noted by Bovarnick et al. ,151 choosing appropriate levels of participation for CYP’s engagement is important going forward, carefully considering how opportunities will empower CYP, whose needs their involvement serves and whether participation can be enacted ethically and meaningfully.

Strengths and limitations

This study provided detailed and novel insights into how SARCs and onward services impacted CYP’s recovery. However, it replicated some of the problems of access to services, with the voices of young participants from black communities and diverse cultural communities, of individuals with physical disabilities and of boys and young men, largely missing in this study. There were also limitations in representing SARC services with only two paediatric services involved. We hope that reflections and learnings from our qualitative study of diverse adult survivors (see Chapter 8) can speak to some of the risk and access issues affecting members of subpopulations missed in the current study.

Implications for health care

It is paramount that a ‘trauma-informed’ approach is adopted across universal and specialist settings, as professionals may unknowingly interact with individuals who have experienced sexual violence and abuse. Therefore, attention should be provided to equipping professionals through training. As delays to receiving support can negatively impact CYP’s health and well-being, urgent investment is needed to resource services adequately. In addition, services should consider ways to manage opportunities while CYP are on waiting lists for therapeutic support which could be beneficial to their recovery. There is also a need for greater education about the role of SARCs and onward support services among any professional a young survivor may interact with, to build CYP’s support network.

Recommendations for research

1. Involving groups that have been under-represented at SARCs is vital to gain increased understanding about barriers to accessing support and how services can be tailored to better meet needs.

2. Further research needs to distil what ‘trauma-informed’ practice means in different sectors and identify barriers to its implementation.

3. The voices of CYP should be part of informing research studies throughout their lifecycle and in the development of services. These opportunities should be safe and trauma-informed spaces.

Study design overview

This qualitative study was embedded within the wider cohort study presented in Chapter 5 and involved interviews with individuals who had and had not accessed SARCs following exposure to sexual violence and abuse at some point in their lives. It employs a story telling approach to gather data with participants and applies narrative analysis. This chapter is underpinned by narrative constructionism; this is an interpretive and socio-cultural approach that positions those participating in research as meaning-makers who use narratives to make sense of and communicate about their experiences. These narratives are seen to have been passed down and co-constructed through participants’ social and cultural worlds. 163 This chapter sets out to explore people’s actions and efforts to help themselves in relation to the occurrence of abuse and violence and the impacts on their lives, in particular, where individuals hold identities associated with marginalisation or ‘minoritisation’. 164 This could relate to identifying as a member of a minority cultural or ethnic group, identifying as transgender, having chronic or long-term mental health problems, involved in sex work through coercion or choice or being homeless. Thus, approaching ‘narratives’ as the cultural and social resources available to participants and upon which they construct the ‘big’ and ‘small’ tales of their lives and experiences of trauma and abuse and help-seeking is viewed as an appropriate orientation for this work.

We acknowledge that without clear explanation, terms such as ‘story’, ‘tale’ and ‘story telling’, which are central to narrative methods could be unhelpful, even harmful, in this context. Indeed, ontological relativism, the basis for viewing knowledge in this study, conceptualises psychosocial phenomena as multiple and ‘created’. This may seem at odds with survivors’ expressed need for being believed and a ‘positivist’ stance that a single truth related to individuals’ victimisation exists. Our research takes the position that all participants are believed, with full understanding of the role of ‘being believed’ in survivors’ healing and recovery. People’s stories are not constructed within a vacuum, this form of narrative enquiry enables the participant to share their experiences in the order of importance that matters most to them. 165,166 This also enables the researcher to look at the whole person and not just that part of a person’s life that is under investigation. 165,166 We also believed narrative enquiry would minimise reductionism and draw attention to power dynamics related to gender, sexuality, race and socioeconomic status operating in everyday life that may affect help-seeking. For us, storytelling is used as a way to gather powerful accounts and authentic voices that centre the survivor as the ‘story-teller’ and the researchers and MESARCH LEG member (or experts-by-experience) as story analysts. Use of the constructionist approach enables the study to advance one of its key aims: to access the critical personal, socio-cultural and political narratives about people’s capacities to keep themselves safe and seek help (e.g. from SARCs) in the aftermath of exposure to trauma and abuse as well as those narratives that influence their recovery journeys while conferring anonymity and contributing to ‘future-forming’ research. 167

Changes to protocol

The study was unable to recruit in the community to the projected target of 20 (Group 2: culturally and linguistically diverse people who had never accessed SARCs) due to the particular barriers of stigma, silencing and abuse faced by marginalised individuals and groups. We engaged with eight willing community-based organisations; these agencies reflected on the difficulties in engaging with their target communities and how they run outreach programmes using workshops about confidence-building and group activities for many months before issues related to safety and abuse can be raised. Despite the challenges, the MESARCH cohort study did recruit 15% of survivors from a range of cultural, ethnic or linguistic backgrounds; therefore, we focused our efforts to understand about barriers for these individuals, expanding beyond our target number of 10 (Group 1). We experienced no difficulties recruiting the planned 20 participants for our third group (Group 3 were individuals in the cohort study with minority characteristics other than cultural or linguistic diversity (e.g. LGBTQ+ identity, people with a disability, men, older people).

Theoretical approach

Reflecting the specific aims of this study, intersectionality168 was employed as a theoretical framework for guiding design, data collection and analysis. It is a valuable lens for interpreting the experiences of people, intentionally accounting for a relational web of factors that may help or hinder their safety, access to helping resources and recovery. This web of factors includes gender and sexual identities, ethnicity, socio-economic status, nationality and immigration and a range of other protected characteristics such as mental health, age and disability. Another important and linked theoretical underpinning is the potential for survivors to be marginalised or ‘othered’ within their own communities. 169 Survivors may be isolated for having poor mental health, for example, or discouraged from seeking help for their health concerns. Similarly, domestic and sexual violence and abuse may be seen as private family matters: victims and survivors attempting to exit abusive relationships or seek help, especially from outside of that community, may be seen to bring shame on families/communities. Othering may also apply where minority populations are treated as ‘them’ by institutions made up of ‘us’ (e.g. inadequate policing responses to a report of a rape by a gay man or by a non-English-speaking migrant woman). Many people who participated in MESARCH would say, in reference to the abuse and what occurred afterwards, that they experienced events that had life-changing consequences. Narratives may serve to rebuild a person’s ‘shattered sense of identity and meaning’,170 and thus, the research was also concerned with capturing these narratives.

Participants and settings

All participants for this story telling and narrative analysis study were aged ≥ 18 years at the time of their qualitative interview.

This part of the MESARCH project set out to engage three sub groups of survivors:

• Survivors that had accessed a SARC and self-identified as being from any cultural, ethnic or linguistic minority community in England (Group 1).

• Survivors in the community who self-identified as being from any cultural, ethnic or linguistic minority community in England but had not accessed SARCs (Group 2).

• Survivors that had accessed SARCs and considered themselves to be from any other minority (e.g. LGBTQ+ community or had a disability) or held a characteristic that may lead to experiencing marginalisation or vulnerability in this context (e.g. men survivors; homeless individuals; those with long-term mental health problems such as psychosis or complex PTSD) (Group 3).

The SARC sample (i.e. Groups 1 and 3) was drawn from the MESARCH cohort study using maximum variation sampling to represent a range of characteristics. The setting of the MESARCH cohort study is described in Chapter 5. Group 2 consisted of people from minoritised groups who had not accessed SARCs and were likely to have experienced delayed help-seeking. 171 This group was more difficult to recruit. We relied on voluntary sector organisations that provide support after sexual assault, rape or abuse and/or those with community ethnic minorities outreach services aimed at engaging seldom heard populations.

Recruitment and consent

See Chapter 2 for a discussion of the ethical principles governing the MESARCH project.

Analysis

Narrative analysis captures a range of methods that have in common their focus on stories. 173 Visual narrative analysis (focus on how and when an image was created), structural narrative analysis (focus on how stories are told and linguistic features used) and personal narrative analysis (focus on evolving personal life biographies) are all examples of types of narrative analysis. We selected dialogical narrative analysis174,175 because it promotes examining what is told in the story (its content) and how a story is put together in terms of the narrative resources that are deployed. 163 Further, dialogical narrative analysis considers what occurs as a result of telling the story (its effects). Smith refers to narrative as a ‘practical medium and the primary medium for action’ (p. 213)163 making it consistent with the goal of many MESARCH participants: After a history of assault, I think I was more frustrated by it. I just wanted to do something productive and so I contacted SARC about the project. I feel the research is empowering me. I don’t feel the shame I used to feel about. It’s a heavy weight to hold on your own; having this dialogue in a way that will help others is just very good and the wider project’s goals for impact. It is reasonable to blend narrative methods and in fact dialogical narrative analysis blends performative narrative analysis (how the interviewer and interviewee perform a story together and why something is said in an interview) with thematic narrative analysis (what the content of stories is). Story-telling in combination with dialogical narrative analysis potentially achieves a range of functions that are consistent with intersectionality as a theoretical framework for interpreting the findings. 168 They allow us to approach and understand about the various ways in which survivors might experience oppression, risk, barriers and delayed care based on multiple, and often minority, identities.

Characteristics and experiences of participants in the embedded qualitative study

The study involved a total of 41 participants: 20 who fulfilled the inclusion criteria for Group 1 and 14 for Group 3. Seven participants were recruited from non-SARC community settings. See Table 20 for an overview of the characteristics of the participants. Core stories were developed for all 41 individuals, and all but one were integrated into the findings presented here. This individual requested that no direct quotes be included for them. The narrative sample diverged from the cohort study (see Chapter 5) on the characteristics shown in Table 20: there was greater minoritisation in the narrative sample as planned. However, the narrative sample was higher for education, and there was no difference on earnings as the main source of income. The community sample (Group 2) constituted six women and one man. All seven participants described themselves as having Asian heritage, from a range of ethnicities including Indian, Bangladeshi and mixed Chinese-Indian. Religious affiliations were Islam, Sikhism, Buddhism, Hinduism and Spiritualism. All participants in Group 2 reported multiple forms of abuse and violence from multiple perpetrators. Sexual violence and abuse had occurred in the context of domestic abuse (100%) and forced marriage, and many of the women had endured honour-based abuse from their families and communities in the UK and overseas. The research also documented testimonials of violent forms of reproductive control and one participant told of persecution over their sexuality from their family in their country of origin.

Main narrative findings

The 41 core stories developed with survivors shared many narrative plots and arcs central to addressing questions about barriers to access to SARCs, other SAAS, NHS care and justice-related services. We first present typologies of risk in relation to sexual violence and abuse (three sub-themes) (see Table 21), these were not intentionally examined but were a pertinent feature of participants’ narratives, intersecting in important ways with identities about mental health, prior trauma, cultural heritage and social disadvantage. Second, we present typologies that foreground the most significant and urgent access issues identified by our sample (eight sub-themes). We also showcase aspects of best practice in enabling vulnerable and/or minority community survivors access SARCs and other settings. Finally, typologies about recovery and healing are presented. Twelve sub-themes address what defines, promotes and prevents recovery in this context.The narrative study provided in-depth data to complement findings from the process evaluation, and the reporting of clinical outcomes in the cohort study.

‘I feel recovered’

For some participants, they were able to tell a story of more complete recovery. They still expressed experience of trauma and having been on a journey but felt that they had reached a positive place that they could call recovery. In addition to the factors described by those participants above, additional features of these narratives included commentary about recent societal progress in talking about sexual violence, a further range of therapies and support and getting good, consistent support from a range of sources (see Table 21).

TABLE 21 - Typologies of risk, access and recovery

Societal progress, talking openly and online help

One participant talked about the experience of a growing openness in society to discuss the issue ofCSA and found this very helpful for recovery (see full core story in Appendix 3, Box 47). Similarly, the growth in online information to support self-help and learning was identified as beneficial by another participant and a source of empowerment (see Box 39).

BOX 39 - Access to information a powerful tool for self-help

I don’t think I would have recovered as well if I didn’t do my own research and my own self help, but obviously the internet can be a dodgy place and you can end up making yourself feel worse. But I think trying to increase my self-awareness around what happened to me and the effects, like reading into the complex PTSD side of things and stuff, constantly working on yourself, has really helped me get to where I am today. Because what I was like after that incident compared to now is two different people and, like I said, it wasn’t because of therapy or anything. It was a lot because of my own work that I put into it. I think having sort of good resources online and stuff is really helpful.

What works is different for everyone

What was noticeable about the narratives we gathered from participants was how every person’s path to recovery differed. In addition to the commentary about meditation, yoga, exercise and art therapy outlined above, other helpful therapies and strategies included EMDR (see Box 40), spirituality and religion, forgiveness (see Box 41) and a type of visualisation meditation (see Box 42).

BOX 40 - EMDR helped

I ended up having EMDR therapy through; it was free, so it was through the NHS. It was great to have access to that, it changed my life completely. Absolutely completely changed my life. Since my EMDR therapy, I’m fine to talk about everything, it’s not a problem really. Nowhere near as much as it used to be. After the first session, it was really difficult and it bought up all the memories again that you try and forget, but the way they make you, that your brain sort of starts working, instead of being like ‘this triggers me’ you know, you see it out and about and instead of triggering you, you think of yeah, there’s that thing. You carry on with your day, I don’t fully understand how I did it. They talked you through, you know you’ve talked through a particular scenario, not even the worst part of your experience, it could have been any part, and you just let your brain take you where it took you. It was really strange, after two or three sessions seeing that particular scenario that you go over and over, seeing how your brain changes. I was already feeling so much difference. That therapy, it was before the COVID-19 pandemic so two and a half years ago and even still I feel okay.

BOX 41 - Forgiveness helped

That was also sort of a similar time when I decided that I was going to forgive my dad, and I called off the police investigation, and actually, I just managed to close quite sort of that chapter and start working on the forgiveness and forgiving my mum as well and having more of a relationship with her. So for my mental health, that was definitely the best thing. I think outside of the space created by church, there is so much encouragement, and rightfully so, for people to come forward to report these things to make sure that justice is served and that we get justice for what’s happened to us. I feel maybe a teeny bit guilty saying it because I know that other women struggle with the coming forward, but for me, it’s been so much easier and so much more valuable to actually just let go of it and forget about it and I know that’s a privilege that I have that other people don’t have. But for me, it’s been the right thing, definitely.

BOX 42 - Visualisation meditation helped

My mental health team dropped me after the [trainee] left. But luckily there was the ‘spiritual lady’. I don’t know what else to call her. She did things alternatively like visualisation techniques with you. Basically, it was like meditation. Understanding that I can distance myself. Like, if I’m with a group of people and I don’t feel comfortable, at any point, if there were any actions, words, anything like that, I can still pour love to those people without emptying my cup. Like, you are allowed to have connections, from a distance. You’re allowed to take two minutes to yourself. When you’re sat in a group full of people and you know, you’re suddenly in a panic or a whirlwind, you’re allowed to sit there and just go, sorry guys. I’m just out of this conversation for a minute. Um, and just visualise, yourself. It’s things like visualisation, like light coming from your feet up to your head. Like, just your inner lightbulb, where you’re protected, and no one can hurt you.

Consistent support from all sources

A final feature of the recovery narratives among those who seemed to feel the greatest sense of recovery was the consistency of support they had received from a range of individuals and support services. This is illustrated in Steve’s narrative (see Appendix 3, Box 47). Positive, skilled responses from a range of people, support organisations and the wider system can contribute to a sense of strong support, even when criminal justice proceedings are slow or non-existent. There’s a sense of sharing the burden within these narratives, such that the recovery and healing journey is not done alone but in collaboration with range of professionals and agencies, with family and friends and with colleagues and communities.

Summary of main findings

Our analysis of the 41 participants’ narrative data has been presented across three different typologies. We identified narratives relating to (1) narratives relating to risk of experiencing sexual violence and abuse; (2) accessing support in the aftermath of sexual assault or abuse; and (3) narratives relating to recovery aligned to lack of recovery, progress in recovery and hope for the future, and feeling a sense of recovery.

We purposively sampled people from minoritised and underserved communities from our wider cohort study and included a sample of people from outside of our cohort recruited through third-sector organisations to include voices of those who have not accessed support through SARCs. These analyses likely represent the largest scale qualitative investigation to date of the experiences of such a diverse range of minoritised groups’ experiences and journeys in the aftermath of rape, sexual assault and sexual and domestic abuse.

Our findings highlight the fact that recognition of what constitutes rape and abuse may act as a barrier to seeking help, and that many forms of sexual violence and abuse can occur in a complex web of community and family norms which serve to hide the abuse and create additional barriers to seeking help and support, including the opportunity to escape from abusive domestic situations. The health and social care workforce needs to be culturally-competent in understanding when they might need to ask questions, and to do so in safe ways. Disseminating information about how to access support in safe ways to people in all communities who may be suffering or at risk is needed to support better access. Professionals also need to be trained to be confident in intervening when a victim/survivor may not feel able to take the first step, but they are in danger.

In Simran’s story (see Appendix 3, Box 44) it is clear how the structure of the system inadvertently acted in a discriminatory way to block her and her child’s access to safety, justice and the right support to recover. There are additional experiences shared that draw out the positives of a LGBTQ + focused response, and the potential negative influences of intersections of social disadvantage, older age and pre-existing mental health conditions can have. Together these serve to highlight the need for outreach and services that take account of these characteristics. This is particularly pertinent where we have evidence that some minoritised groups may be at greater risk of sexual violence. 152–154

It was interesting that although we did not set out to explore risk of victimisation, our participants presented narratives around risk. In addition to the risk of re-victimisation that experiencing abuse can bring, the risk narratives also drew out some of the ethnically and culturally minoritised experiences of people for whom that characteristic was relevant. Risks included the taboo nature of discussing relationships and sex and deeply entrenched patriarchy as well as cultural practices such as forced marriage and honour-based violence.

Changes in societal (and therefore professional) understandings and attitudes towards victims/survivors of sexual violence was cited by participants as helpful in relation to both access to support and their ongoing recovery journey. There was recognition that people are more open to talking about this issue than in the past, and a greater tendency towards believing survivors and aiming to be trauma-informed. Accompanying this societal change is burgeoning information available to people on the internet which was included in one of the recovery narratives of our participants. The availability of online information and psychoeducation for support can empower survivors to take control of their recovery journey. Continued pervasive societal norms that act in support of perpetrator behaviour was also cited, however, as detrimental to recovery journeys. Hence, our data reflect helpful societal progress, but the need to continue to push to further this empowering and supportive agenda. These findings align with recent discussion in the academic literature about the need to understand better the structural and social causes of sexual and domestic violence in our attempts to prevent it and support those who are victims of it. 178

There were challenges identified in relation to accessing the right kind of care, with many of our participants talking about unsuccessful attempts at therapy and inadequate offers of care that did not suit their needs. Ultimately our narratives where there is hope of recovery demonstrate a level of acceptance that recovery will be a long road, and that a range of different therapies and mind and body connecting activity (such as yoga and meditation) may be useful for that process. These findings reflect those from our two Cochrane Reviews to some extent, which provide emerging evidence of benefit from less traditional interventions and therapies such as trauma-informed yoga (see Chapter 3 of this report). 36,42 Similar to our qualitative Cochrane Review which found that interventions led to a wide range of benefits,42 the narratives of our participants showed recovery as having a range of different outcomes that they viewed as beneficial including reductions in anxiety and flashbacks and the ability to engage in activities that they enjoyed doing such as physical activity and volunteering.

Narratives of access and of recovery encompassed the particular challenges there are for those who have complex trauma to deal with. These are often individuals who have experienced a number of ACEs including abuse and neglect, who have a pre-existing mental health condition and who may be experiencing domestic as well as sexual violence. They may have been victimised many times and be at risk of re-victimisation. They may also have children to take care of. Services are not set up to provide the tailored, bespoke, multi-faceted and multidisciplinary care that survivors with these patterns of abuse need. The narratives of our participants show this poor response to complex cases is likely contributing to the lack of recovery experienced.

Strength and limitations

Large-scale qualitative inquiry into the support and recovery experiences of survivors of sexual violence is challenging to accomplish. A strength of this study is that it has achieved the collation of detailed narratives of these experiences from a large sample of SARC service users as well as survivors who have not accessed SARC. The sample likely represents the most diverse to date in terms of its representation of cultural, linguistic and ethnic groups. It includes people of different genders and sexual orientation, and a range of ages, and levels of socio-economic status and education. It highlights the hope there is for recovery across these diverse groups, but also the gaps in support that exist for those experiencing complex trauma and from the most underserved communities. A further strength of this study is the way it complements the more traditional cohort and health economic analysis of outcomes, valuing multiplicity of knowledge and giving weight to experiential knowledge. Its inclusion within the research process has created opportunities for epistemic justice where silenced or marginalised voices are amplified. 30 As one participant put it, I wanted to really have a voice because I felt in other areas, when I was speaking up before about how I was feeling, it was getting diminished or just ignored. We acknowledge that a limitation was that we reached far fewer participants from non-SARC support contexts than intended. Those we have reached provided detailed narratives and provided insights into the complexity of intersecting forms of marginalisation, how these interact with sexual violence and abuse risk, and the nature of barriers to service access. This part of the study further illustrated the resilience of individuals and communities in the aftermath of abuse, and the creative and personal ways in which people navigate life after trauma.

Implications for health care

Health and care and support services have an important role to play in the recovery of survivors of sexual violence and abuse. Getting a good response from services matters and can contribute to good recovery. The response of wider society matters too though (e.g. see our #whatareyoudoing campaign39), and efforts to improve cultural and societal norms that may contribute to sexual violence and abuse and create the context for recovery are needed as well. Better outreach and training for staff to support those still in abusive environments to escape abuse and get support are needed. Support for those experiencing complex trauma histories, multiple victimisation with complex mental health and personal circumstances are also needed to avoid exacerbating health inequalities and meet the needs of the most vulnerable survivors.

Recommendations for research

Future research should focus in more depth on those with the most complex needs and who do not make it to care through SARCs. Given that minoritised groups are often least likely to access support in the aftermath of sexual violence via SARCs, it is not yet clear whether SARCs are the best care option for all.

Pre-SARC

A primary focus of this research was to understand what promotes or deters access to SARCs. Through sharing their personal narratives and experiences, participants shaped where and when their story starts, with most of them taking us far into the past. This lifespan lens underscores how, for many survivors, risk for sexual violence and abuse begins early in life. 180 This risk is reinforced by power dynamics connected with race, gender, socioeconomic status, religion, sexuality and disability, shown by our research team and others. 14,178,181–184

Our qualitative study identified gender-based violence rooted in gender inequality as a risk factor, especially poignant among the racially minoritised community sample of non-SARC attendees. Another theme concerned ‘silencing’. We gathered widespread accounts of risk arising from a lack of early education and everyday discourse about relationships, intimacy and sex. 185 When sexual violence and abuse did occur, the milieu of customary silence rendered survivors ‘voiceless’ and unable to draw on, otherwise available, informal support networks, and certainly not formal support. While there was a clear historical context for silencing among many of our older participants and a clear minoritisation context for many of the culturally diverse participants, silencing and shame about abuse was ever-present in narratives irrespective of time, place or people’s backgrounds. The absence of tools for talking about abuse and seeking help, and the debilitating context of shame was exploited and could perpetuate the abuse. The final area of risk concerns a triad of trauma, poor mental health and insecurity. Individual and composite narratives bear witness to life struggles as people strived to exit cycles of: (1) escaping abuse leading to unstable, unsafe living conditions; and (2) escaping chaotic, risky living thrusting people back into risk/victimisation; and all this occurring in a context of long-term mental health problems, which included substance misuse among the most minoritised participants in our samples. A central tenet is that trauma itself is a risk factor that continues to compound and shape risk (and recovery).

Our quantitative analysis on risk and mental health burden at entry to our cohort study effectively triangulated these qualitative findings. Consistent with the findings of other studies of people affected by sexual assault and abuse and attendees at SARCs,83,186 45% of participants joining the study had a prior history of mental health problems. For survivor-level factors, pre-existing mental health problems were associated with a 7.6-point difference on baseline PTSD compared to those without long-term mental health difficulties; each additional ACE was associated with a 1.4-point greater score on PTSD symptoms; there was strong evidence and an apparent dose-response for PTSD symptoms being higher for those with fewer financial resources with a 7.3-point difference, and strong evidence that being ‘unable to work’ was associated with a 10-point higher baseline PTSD score.

As such, it is critical that we adopt an intersectional framework for considering the implications of MESARCH findings for two reasons. Firstly, we need it to make sense of their complexity; we are witnessing layers of interconnecting power dynamics at play in the lives of survivors that relate to gender, race, socioeconomic status and disability (especially if we consider the burden of mental health as frequently the basis for survivors being unable to work). Secondly, there is the sheer scale of individuals affected by these risks and ‘dynamics’ and thus any measures to address disparities in access across the SARC pathway; enhance treatment options; build workforce resilience, skills and competencies; and achieve effective lifelong care must draw on a framework that centralises parity. Bearing in mind that the quantitative data are provided by participants who have managed to access help, it is concerning to consider the risk profile in communities and subgroups who are unable or prevented from getting help.

We do not wish to lose sight of the fact that abuse crosses all sociodemographic lines,187 and our sample included many individuals and communities without traditional risk factors – two-thirds were educated to A-level or above, approximately half reported few or no financial problems and many people were in work and the vast majority of participants was not racially minoritised. Yet, our careful ‘categorisation’ and analysis of the sample’s characteristics revealed a burden of childhood trauma (ACEs) that was noteworthy. Nearly four in every five participants in the study indicated four or more ACEs, widely considered as a threshold beyond which people are at risk of poor long-term mental and physical health outcomes. 188–190

Sexual assault referral centres: access and experiences of care

We now move beyond discussing risk to presenting insights into people’s experiences of accessing SARCs and then the care and support obtained in that setting. Rich data, from across the project, extend our understanding about access to SARCs as well as to the wider context of health, voluntary sector and social care services and justice, addressed further on (see Figure 10).

FIGURE 10.

Socio-ecological organisation of barriers to accessing sexual assault referral centres pathways.

Survivors in this project had accessed 21 SARCs and delays and barriers to accessing SARC resided at individual, family and community levels, such as people not knowing about SARCs, not being ready to seek help, not being able to: After the incident that happened to me, I was in shock for 4 days. Those in our community sample who had experienced severe partner, family and/or community abuse and control (e.g. arranged or forced marriages) rarely perceived options. They were mainly focused on surviving and escaping abuse such that meeting wider needs from sexual abuse (e.g. sexual health care, specific crisis emotional support and forensic care) were not priorities and were neglected for long periods with pervasive long-term effects. Overall, nearly 45% of the recruited sample had accessed SARCs in the first 10 days since the sexual assault down from a pre-COVID-19 pandemic level of 60%, and the most typical perpetrator type was ‘acquaintance’ (40%), followed by partner-perpetrated sexual violence and abuse (25%) and strangers (20%); 13% of people in the sample had sought help in relation to exposure to CSA. Approximately 40% of people had a FME.

The delivery of health care and support by SARCs was consistently evaluated by survivors as exemplary. This finding was consistent across our cohort study, process evaluation and narrative study of minoritised groups. Our findings revealed no indicators of systemic harm, and half of the survivors accessing SARCs rated them as 90 or above for benefit in meeting needs arising from exposure to sexual violence and abuse. While behind every number there is a person and the research has examined individual-level responses, just four individuals reported SARCs had been harmful to them. This was the lowest rate of harm detected across any sector in this research. Our SARC process evaluation emphasised the importance of cohesive regional partnerships between SARCs and multiple statutory and voluntary sector agencies. This was also observed and articulated by survivors with regard to their satisfaction with on-site care, referrals and follow-up. We view SARCs as a vital gateway to recovery and it is often survivors’ first encounter with care in relation to sexual violence and abuse. As such, it is crucial that SARCs provide a safe, individualised response that facilitates the person to continue on the care pathway, as our research shows that there are multiple factors that can derail a person from continuing this journey.

Post-SARC: access and experiences of care

One of the main features of care after SARC was the ISVA service, a professional to advocate for and support survivors through the criminal justice process. Depending on the local model, some survivors continued care with a SARC-provided ISVA; for others, they were referred to an ISVA provided by the voluntary sector. ISVAs were integral to an effective pathway, and for the most part, survivors did not report difficulties accessing ISVA services and adapted during the COVID-19 pandemic to receiving that care online. 191 Furthermore, survivors reported high levels of satisfaction with the support provided by ISVAs. In the study, there were more than 4000 contacts with ISVAs; at baseline, 85% had an ISVA falling to 40% at 1 year in line with the reduction in open cases from 56% to 20%. Participants expressed difficulties when ISVA care was withdrawn in response to case closure and there appears to be variation nationally in the extent to which ISVAs can be accessed or maintained independent of criminal proceedings. The problem of access was much greater in relation to counselling services in the voluntary sector. Capacity issues did not affect SARCs to the same degree, and many survivors accessed other forms of support at SARCs (e.g. information and advice, structured emotional support sessions) to assist them in the time they were on waiting lists.

Young people did express concerns, especially young people from minoritised backgrounds: ‘This system isn’t very inclusive. It’s not very accessible and a lot of ethnic minorities and young black girls are subject to sexual violence at a young age and you have people in the system who don’t understand those backgrounds, don’t understand the sociology behind them and how they ended up being the victims of what they were. I think it’s just leading to a lack of justice being provided. There’s not an understanding of the lives of the people that are involved’.

The findings of the study suggested that participants accessed a wide range of healthcare and other public sector services (with GP visits and police contacts comprising the highest number of reported contacts). Although service use was higher among our participants at baseline (representing a period of nearly 100 days since participants had accessed a SARC), this decreased over time, for example, NHS contacts reduced from almost four contacts at baseline to less than one contact per person at 12 months (covering to the previous 6 months). While a full embedding of the wider health and social care access experiences within wider literature is beyond the scope of this research, we observed survivors’ mental health as a primary barrier to access in all settings. 192,193 For example, in the NHS, survivors struggled to be provided with psychological care specific to sexual abuse, when they had co-existing mental health diagnoses, because the latter was focused upon. There were parallels here with survivors in domestic abuse services, where there were gaps in care for the sexual abuse that had been perpetrated by partners. 6 Another common scenario was being unable to access therapies on the NHS seen as appropriate for complex trauma. Yet another was NHS referral ‘loops’ – patterns of GP referral to mental health, and discharge back to GPs, who were often seen to be ill-equipped or unwilling to support mental health needs of survivors. 186 Formal mental health expertise to identify and support mental health concerns remains limited within the SARC skill mix; the MiMoS project found that staff lacked clarity about pathways into the mental health services together with high thresholds for accepting service users and long waiting lists. 66 As we did with SARCs, we explored survivors’ experiences of harm and benefit in different sectors post-SARC. The findings from the voluntary sector services replicated what we found for SARC – a median of 90, suggesting that half of the people rated maximum benefit scores across a rich landscape of support provision for survivors. The median score for the NHS was 75, with a notably higher proportion of survivors indicating that they experienced a harmful response from the NHS. Coping and recovery for survivors occur in a justice context,194 and while this study captures widespread evidence of positive action, police and justice lagged considerably behind in its harm-benefit profile. While 50% of participants indicated benefit scores more than 50, a quarter of the sample indicated they derived harm from their interactions and experiences: ‘The whole investigation process has been just as traumatic as the abuse, or in another way, worse’. While health care and other support services have come a long way towards providing safe, effective and compassionate care for survivors, society continues to grapple with its poor performance in meeting justice needs. While an in-depth discussion of findings on these aspects is beyond the scope of this research, the next section distils some of the key findings on the factors influencing recovery.

Effectiveness of SARC pathways for survivors of sexual violence and abuse

Central to our endeavour was to understand what impinges on outcomes for survivors. There were many outcomes assessed in the cohort study; here, we summarise the quantitative evidence for a reduction in trauma symptoms (PTSD-based PCL-5) as our primary outcome, complemented by survivors’ voices on recovery journeys. People entered the cohort study with a high trauma load – 70% scored above the diagnostic threshold for PTSD. There was strong evidence of associations between PTSD at baseline and past trauma burden (e.g. ACEs), health profiles (chronic mental health problems) and socioeconomic factors (e.g. economic deprivation). The study also found weak evidence that sexual violence perpetrated by partners, family members and people in positions of trust aggravated PTSD and there was a small effect of time since trauma (more time to SARC, more trauma detected at baseline).

At 1 year, the sample as a whole had achieved a MCID (mean reduction in 9 points > 0.5 SD69); however, 55% were still above the diagnostic threshold. There was no evidence for differences in PTSD at 1 year related to models of SARC service delivery participants had accessed (i.e. charity, police, NHS, private sector provision). Cost-utility analyses also identified no significant difference in cost per QALY gained between these different models of SARC service delivery. Similarly, we did not detect an effect of the ISVA model (i.e. SARC ISVA or non-SARC ISVA). Although intensity/frequency of service use was associated with trauma at baseline, we detected no effect of level of service use over the year or did we detect an effect of harm-benefit from the police/justice response on the PTSD outcome. Thus, the only predictor for PTSD at 1 year was the baseline burden of PTSD, which has been shown to be strongly determined by several well-defined pre-offence characteristics, and to some extent, non-recent sexual violence and abuse and proximity to the perpetrator. Our research also suggests that those with the highest trauma at baseline had the greatest improvement. Furthermore, while those with four or more ACEs had twice the odds of probable PTSD at baseline, this reduced (OR 1.5) at 6 months and effectively was null (OR 1.01) at 1 year. It would be promising if 12–18 months of care and support could dilute the effect of the ACE profile.

Showing clear parallels with our study, Khadr et al. reported based on a study of young survivors that the presence of a psychiatric morbidity 4–5 months post-SARC was associated with baseline psychosocial vulnerability (defined as previous social services involvement, mental health service use, self-harm or sexual abuse), but not assault characteristics. 129 Khadr et al. recommended that CYP require comprehensive support to address high levels of pre-existing vulnerability and psychological comorbidity, and prevent re-victimisation. Our own CYP study observed that recovery and healing was influenced by how quickly CYP accessed support, its content and duration, the location and context of care, how well services worked together (aligning with findings from our process evaluation) and CYP’s interactions with professionals at SARCs and across sectors.

Our process evaluation and narrative analysis helped contextualise these quantitative findings. On the whole, SARCs working with ISVA services is suggested to be an effective gateway to care for survivors: ‘I feel that the support I’ve had from the ISVA service, from SARC, has given me a better outlook on life’. A vast array of other factors interacting at different levels played a role in survivors’ healing journeys (see Figure 11). Some overlapped with ‘access’ barriers and enablers (see Figure 10). Finding therapeutic fit for presenting issues and, importantly, past or complex traumas and this being available within statutory health services was a gap that arose post-SARC. Furthermore, perceiving that justice had prevailed and holding a belief that the police were thorough, respectful, communicative and honest about the investigation was important in recovery and healing.

FIGURE 11.

Socio-ecological organisation of recovery in the context of the sexual assault referral centres pathway.

Effectiveness of SARC pathways for children and young people

Our CYP study showed that the holistic care and support provided by SARCs played a critical role in young survivors’ recovery. Their access to support at SARC relied on professionals referring them to SARC once they disclosed the abuse. Improvements in young survivors’ physical and mental health, education, interpersonal relationships and re-engaging in a wide range of areas of their lives was affected by how quickly they accessed support, its content and duration, the location and context of care. CYP’s readiness to access support was impacted by factors such as re-exposure to abuse, living situations or the quality of relationship with a professional. This mirrors findings in the context of CYP experiencing domestic abuse where readiness to engage in support involves a complex interplay of individual, relational and contextual factors. 195 Notably, when young survivors in this study experienced cohesiveness of support between professionals across and beyond the SARC pathway, they had an increased sense of safety. Importantly, interactions with professionals that lacked a ‘trauma-informed’ approach at any stage of their recovery was associated with poor or delayed healing. This supports findings that show how accessing a service in itself does not necessarily improve trauma symptoms. Rather, it is the quality of services and survivors’ interactions with professionals that lead to long-lasting positive effects. 131,150

Effectiveness of psychosocial interventions for survivors of sexual violence and abuse

Sexual assault referral centres represent a gateway to pathway of care for survivors that ideally opens up options to various psychosocial interventions. These may be accessed through NHS services, including Talking Therapies, for anxiety and depression programme (formerly, Improving Access to Psychological Therapies, IAPT) and community mental health, where many of the traditional trauma-focused interventions are offered. Such therapies are also available in the voluntary sector, which additionally offered survivors in our studies a wide range of arts-based, body-based, counselling and social interventions. There are many examples of people accessing therapies through private routes due to being unable to access support they felt they needed or access that in a timely way. Our Cochrane Review, synthesising findings from 36 randomised controlled trials from various parts of the world, found a large reduction in PTSD symptoms and depression at post-treatment following psychosocial interventions compared with controls. 36 Our review also suggested that CBT with a trauma focus such as Prolonged Exposure and Cognitive Processing Therapy and other exposure-based therapies (e.g. EMDR) probably benefit survivors of sexual violence and abuse over non-trauma-focused intervention. However, trauma-focused interventions may result in higher treatment non-completion and some survivors may continue to experience a high symptom load post-treatment. In response to the limits of trauma-focused interventions, there have been calls for more effective approaches to the management of PTSD,196 especially for sexual violence and abuse exposure. 197 One contribution of our review is the synthesis of several novel treatments across a range of promising new areas such as RTM; trauma-sensitive or trauma-informed yoga; Lifespan Integration (LI); and neurofeedback. Some of these treatments have been shown to be effective among survivors of other types of trauma (e.g. RTM),198 while other therapies have scarce evidence in any population.