Alternatives to custodial remand for women in the criminal justice system: A multi-sector approach

Throughout the world, women involved in criminal justice systems often present with substantial needs and vulnerabilities. Diverting vulnerable people away from prison is government policy in England and Wales, but full psychiatric and social assessments are expensive and hard to access. A screening and quick response initiative - alternatives to custodial remand for women (ACRW) - was implemented across three areas of London (West, South and East) to sup-plement existing court liaison and diversion services, to assess the feasibility of a supplementary custodial remand service as part of a women's specialist service pathway in the criminal justice system in England. Three mental health trusts and two voluntary sector providers offered this service enhancement – a screening and service link provision in three London boroughs between 2012 and 2014. We conducted a service evaluation using routinely collected service use record data. The service made 809 contacts, of whom 104 had contact on multiple occasions. Many were identified as at risk of self-harm (46%) or had histories of hospital admission for mental disorder (36%), but few were referred either to the liaison and diversion service or specialist mental health services. The largest group of referrals was to women's community services outside the health service (e.g. counselling, domestic violence or sexual abuse services). 180 women had dependent children and 22 were pregnant, increasing the urgency to find non-custodial alternatives. As well as confirming high levels of need amongst women entering the criminal justice system, this evaluation confirms the feasibility of working across sectors in this field, providing an extra layer of service that can complement existing liaison and diversion service provision. The service was responsive and most women using it were kept out of custody. Research is now required to understand the appropriateness of the referrals, the extent to which women follow them through and the impact on their mental health and desistance from offending. input within an already over-stretched system; our second question was: does this approach lead to a reduction in custodial remands?

A screening and quick response initiative -alternatives to custodial remand for women (ACRW) -was implemented across three areas of London (West, South and East) to supplement existing court liaison and diversion services, to assess the feasibility of a supplementary custodial remand service as part of a women's specialist service pathway in the criminal justice system in England. Three mental health trusts and two voluntary sector providers offered this service enhancementa screening and service link provision in three London boroughs between 2012 and 2014.
We conducted a service evaluation using routinely collected service use record data. The service made 809 contacts, of whom 104 had contact on multiple occasions. Many were identified as at risk of self-harm (46%) or had histories of hospital admission for mental disorder (36%), but few were referred either to the liaison and diversion service or specialist mental health services. The largest group of referrals was to women's community services outside the health service (e.g. counselling, domestic violence or sexual abuse services). 180 women had dependent children and 22 were pregnant, increasing the urgency to find non-custodial alternatives. As well as confirming high levels of need amongst women entering the criminal justice system, this evaluation confirms the feasibility of working across sectors in this field, providing an extra layer of service that can complement existing liaison and diversion service provision. The service was responsive and most women using it were kept out of custody. Research is now required to understand the appropriateness of the referrals, the extent to which women follow them through and the impact on their mental health and desistance from offending. Women represent a minority of people in criminal justice systems across the world and tend to be there only briefly in any one episode. When they are justice involved, they are likely to present with substantial needs and vulnerability. Of the nearly 800,000 arrests made by police in England and Wales every year, only around 15% are women and when women are arrested it is typically for lower level offences (Home Office, 2018). In a third of cases, arrests relate to theft and handling stolen goods. Although about half of all cases relate to episodes of violence, most of these are represented by minor assaults. Women charged and convicted in lower (Magistrates) or higher (Crown) courts are most likely to receive a fine but a substantial minority are remanded into custody to await trial and may go on to receive custodial sentences. In England and Wales, there are 4,000-5,000 women in prison at any one time.
Increases in the severity of sentencing between 1995 and 2010 has led to a doubling of their number. Around 20% of women prisoners are on remand awaiting trial, compared to fewer than 10% of male prisoners. Over the course of a year, over 3,500 women pass through prison as remand prisoners, with many released at the point of sentencing as a result of time served (Prison Reform Trust, 2018). Women's prisons have high levels of turnover as sentences of less than 6 months are given in around 75% of cases, compared to around 60% for men (Ministry of Justice, 2016).
Both men and women in the criminal justice system have high levels of need but women are likely to have more pronounced vulnerabilities arising from many sources. Although some of these vulnerabilities do not relate to gender, others relate specifically to issues faced by women and criminal justice agencies are often unresponsive to these needs.
The landmark Corston Report highlights most of these (Corston, 2007). First, women have often experienced domestic abuse and violence, sexual exploitation or rape, and these experiences can have a profound and ongoing traumatic effect. Secondly, mental health problems are known to be highly prevalent in this group both in prison, where women are at higher risk of depression and self-harm than men (Fazel, Hayes, Bartellas, Clerici, & Trestman, 2016), and at earlier stages of the criminal justice system (Forrester, Samele, Slade, Craig, & Valmaggia, 2017;McKinnon, Srivastava, Kaler, & Grubin, 2013;Shaw, Creed, Price, Huxley, & Tomenson, 1999). Third, women often face socio-economic disadvantage and come from backgrounds characterised by poverty and social isolation, which can contribute to offending behaviour and prevent desistance from crime.
Prison conditions in many countries are considered to be unfit for purpose. In England and Wales, there have been serious problems with overcrowding and reductions in staffing across the prison estate of up to 30% between 2010 and 2016, leading to increasingly restricted regimes with higher levels of cellular confinement (Ismail, 2020;National Audit Office, 2017). According to the Prison Inspectorate, some women's prisons are unsafe and the UK Government has failed to meet spending commitments to improve poor living conditions and care (Advisory Board on Female Offenders, 2018). As well as problematic living conditions, people in prison also have to deal with stress, fear, boredom and separation from family. Women are commonly sole caregivers for their children, so this can be especially problematic (Corston, 2007). Prisoners describe such difficulties, alongside the uncertainty related to being on remand, or adjusting to a long custodial sentence, as challenging (Nurse, Woodcock, & Ormsby, 2003). Due to the small number of women prisoners, there are only 12 women's prisons in England and none in Wales. Thus, each prison has a large geographical catchment area and may receive prisoners from courts over 150 miles away. This means that relatives may face difficulties visiting, potentially straining relationships, and community service links may be severed.
Combinations of the difficulties of prison life and the vulnerabilities of criminal justice involving women mean that custodial sentences may be unproductive and may further traumatise the already vulnerable (Mollard & Brage Hudson, 2016). For women with mental disorders, availability of treatment in prison is limited, and if hospital inpatient treatment is needed there are often long transfer delays (Bartlett, Somers, Reeves, & White, 2012;Brooker & Gojkovic, 2009;Hales, Somers, Reeves, & Bartlett, 2016). For these reasons, imprisonment of women should be used only in special circumstances. Influential independent reviews (Bradley, 2009;Corston, 2007) have led official government policy to provide guidance that custodial disposals should only be used for the most serious of crimes and/or when protecting public safety (Ministry of Justice, 2018). The extent of this policy success is, however, unclear.
Court appearances are a key stage in the criminal justice pathway. Interventions there could reduce custodial remand and imprisonment. In North America, mental health courts which consider the mental health needs of offenders and aim to use community-based treatment rather than custody have become popular, and there is developing evidence that they can reduce rates of reoffending (Lowder, Rade, & Desmarais, 2018). Substantial changes to the set-up of the criminal justice system may, however, be needed in order to set up such courts, and their involvement to date in mandating and monitoring treatment has been criticised (Seltzer, 2005). In the UK, the availability of community services and liaison and diversion services means that less substantial changes to the criminal justice system may be needed and enhancement of existing services may be sufficient to increase use of community alternatives. These enhancements could focus on known weaknesses of current liaison and diversion services (Srivastava, Forrester, Davies, & Nadkarni, 2013), by improving identification of vulnerable women and assisting courts to order community disposals.
Our aim was to conduct a preliminary assessment of an "alternatives to custodial remand for women" (ACRW) service and describe the women using it. The service was designed to complement existing liaison and diversion and voluntary sector services to identify vulnerable women early in court proceedings, improve coordination of their care and promote use of community sentences for them. Our first question, therefore, was about feasibility of additional input within an already over-stretched system; our second question was: does this approach lead to a reduction in custodial remands?

| METHODS
The study was approved by appropriate bodies to cover each site and partner; approvals from a senior District Judge and the National Offender Management Service were also obtained. The appropriate research and service evaluation review group confirmed that as the study was designed as a service evaluation of routinely collected data, consent from participants was not required.

| Service description
The ACRW service was introduced in 2012, after specific funding was made available by the Department of Health.
Its key aim was to divert women from custodial remand, where appropriate, and to resolve factors underlying crimi- All women attending Court were eligible to be seen in the ACRW service. A voluntary sector forensic mental health practitioner with a first degree in psychology or a registered health care practitioner with a nursing, psychology or social work background was the first level of contact. They undertook reviews of court documents, including prisoner escort records and any available clinical notes, to determine whether specialist services might be required.
Other agencies could also raise a concern or refer a woman to the ACRW service, upon which a forensic mental health practitioner would provide an initial assessment of needs and vulnerabilities. No standardised assessment package was used.
Women thus identified as having mental health problems requiring probable specialist mental health care could then be referred to the second level NHS court liaison and diversion service for further assessment, women with housing needs could be referred to the accommodation-focused voluntary sector organisation, and some both. Any of these services could also refer women to a range of other more specialist services, including domestic violence services, exit prostitution projects, human trafficking specialists, debt and finance advisers, housing services, social services, community mental health teams and drug and alcohol services.

| Data sources and analysis
The ACRW service practitioners electronically recorded information on contacts, referrals and recommendations to the court. Information on need and vulnerability was also collected by them, using an agreed pro-forma that included demographic and mental health profiles, social care needs, engagement with services, interventions, activities and support provided, sources of referrals and accommodation outcomes. Mental health histories, including self-harm, were ascertained both from self-report and other documentation when it was available. If intellectual disability was reported or suspected by staff, it was further assessed using the Learning Disability Screening Questionnaire (McKenzie, Michie, Murray, & Hales, 2012). Additional data on court disposals were also collected, when available, for those who had contact with the ACRW service at the point when these disposals were made.
Data were analysed using SPSS (Version 20) and Excel, employing descriptive statistics.

| Operational information
At the first level of assessment, there were 809 contacts with voluntary sector forensic mental health practitioners; 104 women had contacts on multiple occasions. An assessment was completed in 659 contacts. Most contacts were initiated after the practitioners reviewed prisoner escort records as part of an informal screening process (618; 76.4%) with the remaining contacts coming from referrals from magistrates or District Judges (32; 3.9%), solicitors (13; 1.6%), probation (62; 7.6%) or police custody (19; 2.3%). At the second level, NHS liaison and diversion teams made 103 contacts with 96 women, with 81 referrals coming directly from the voluntary sector forensic mental health service (58) or from court officials (23) (sometimes due to recommendations from practitioners); the remaining 22 (21.4%) contacts were referred by prison in-reach teams, police mental health teams, probation officers or solicitors.
The accommodation-focused voluntary sector partner received 70 referrals from the forensic mental health service, mostly at the South and East London sites where formal partnerships were most embedded.

| Needs and vulnerabilities
Details of needs among the women referred to the ACRW are shown in Table 2. Although 529 (65%) of women selfreported a diagnosis of mental disorder, voluntary sector forensic mental health staff recorded that none was present in 85 (16%) cases or were unable to confirm the diagnosis in 114 (21.5%) cases, the latter largely due to being unable to access prior clinical records. Nevertheless, it was confirmed that about half of the women had at least one mental health need and, overall, 44% had a substance misuse related need. Ninety (17%) women had both. There appeared to be some variation between sites in these characteristics, but the differences were not significant. By contrast, there was a substantial and significant difference in attitudinal or behavioural related needs between the areas, 45% Overall, about one in five women had an accommodation-related need, 129 (16%) had financial needs, 112 (4%) had needs relating to children or families, 105 (13%) reported needs relating to domestic violence, and 77 (9.5%) had physical health needs. Seventy women (9%) had vulnerabilities relating to their involvement in prostitution, while only 46 (6%) reported having skill-based or employment needs. Again, there were marked differences between areas in social needs, with a lower proportion of women in South London generally rated as having each social need. One hundred and eighty-one women reported having dependent children, but only 35 of them had their children living with them at the time of arrest. Fifty-eight had children in care, the rest were with others in the family or the paternal family. Twenty-two women were pregnant at the time of arrest. One hundred and twelve of the referred women reported active needs relating to children or caregiving for other dependent relatives.
More detailed information on mental health needs was extracted where information on diagnosis could be retrieved. Depression was most common (140/529, 26%) followed by psychosis (49/

| Referrals, court outcomes and service interventions
Seven hundred and sixty-two (94%) of cases were referred to other services or were given information about other services they could contact. Details of onward referral after the first level of assessment are shown in Note: Totals in column headings indicate the number of assessment contacts. Percentages indicate the proportion of women referred and will sum to more than the number of women assessed due to co-occurring need. a At the West London site, 47 contacts were with women subject to extradition proceedings and were not eligible to receive referrals to other services. They have been removed from the overall number of contacts. referred to women's services including, for example, locally available counselling services for women (e.g. for domestic or sexual abuse). There were also a number of referrals between the partners within the ACRW service after this first assessment. Referrals were also made by the NHS liaison and diversion team and voluntary housing sector partner to community health and housing services. The West London site was, uniquely between sites, responsible for 47 extradition cases in London and such women were not eligible for onward referrals. They have been removed from calculations in Table 3, so could not have accounted for apparent differences in nature of onward referrals between sites.
The ACRW service provided courts with 694 reports over the duration of the project, with 598 reports originating from the voluntary sector forensic mental health service and 96 reports from NHS liaison and diversion team.
Each report would contain recommendations on disposal to the court, taking account of the information on each woman's circumstances, including the offence, level of risk to others, risk of victimisation, and perception of their own needs. In most cases (68%) the ACRW service recommended bail, and almost all of these recommendations were supported by advice that the women should continue existing service contact and/or be referred for additional support from new services. During the project, diversion to psychiatric hospital from court was recommended for 22 women, while further court-mandated psychiatric assessment in the community, including Mental Health Act assessments, were recommended for 58 women. The ACRW service did recommend custodial remand in some cases where risk or other factors indicated this would be appropriate; in some cases they also recommended reconsideration of the charges or early decision in favour of a community disposal.
Recommendations were followed by courts in most cases. In the 378 cases where bail was recommended, the court granted this in 240 (63.5%) cases. Just over one fifth were remanded (153/692, 22%) or sentenced to custody (76/692, 11%): community sentences were handed to 33 contacts (of 692, 5%), mostly with a supervision requirement.
Over time, there was a downward trend in the use of custodial options for women who had had contact with the service. In the first quarter after implementation, custodial disposals were used in 43% of cases, reducing steadily to 21% of cases in the final quarter.

| DISCUSSION
Women in the criminal justice system have well documented needs and vulnerabilities and there is broad agreement that alternatives to custodial remand should be used where possible. Provision of community alternatives and support services, however, may be insufficient to alter court decision-making and more proactive approaches may be needed. A particularly striking finding was the number of women with mental health diagnoses, current or previous engagement with mental health services or previous psychiatric hospitalisation, suggesting higher levels of need than have previously been reported (Fazel et al., 2016), and yet few of these women received mental health service treatment following the court appearance.
Our evaluation confirms the feasibility of providing an extra layer of service that can complement existing provision, coordinate the activities of NHS and voluntary sector organisations, and perhaps increase willingness of the courts to act on most recommendations arising from these services. It also indicates that despite the apparent burden of mental health problems, psychiatric input through the courts, aside from expert witness involvement which was outside the remit of this evaluation, is used infrequently. Given the high levels of mental health need and the clinical complexity of this group, however, we recommend that the design of services in this area (NHS England, 2020) should be reconsidered. At present, liaison and diversion teams are staffed mainly with nurses with a mental health background; although other disciplines are represented, and not all schemes are the same, few psychiatrists are involved . In addition, many of the pathways leading from liaison and diversion teams to community-based services lack detailwe argue that they now require further thought and more robust design (Disley et al., 2016). There were also important differences in levels of need between the three court sites, indicating that there is a need to understand specific local contexts at the point when these services are designed and financed.
From an organisational perspective, the ACRW services were successfully added to existing liaison and diversion services and voluntary sectors initiatives and able to provide initial assessments and recommendations for a group of women with high levels of need and diverse vulnerabilities, many of whom would not otherwise have been assessed for health and social needs at all. The first level service was able to identify many women with high levels of vulnerability and need and, in partnership with second level services, provided many recommendations which were taken up by the courts. It is apparent, however, that more research is now required in this area. This supplementary service could only provide interventions at one point in time and stage in the process, it is important now to check whether there could be other useful points for supplementary intervention and, above all, to check practical outcomes of the onward referrals workfor example, whether such women are accepted by receiving teams and whether they then engage with these services. Nationally, we also need to understand whether the liaison and diversion approach is the best way forward, or whether a mental health court model, incorporating a clear element of return and followup, would now offer a better approach. In making this recommendation, we acknowledge that the mental health court model has a higher level of evidential support than these liaison and diversion services (Steadman, Redlich, Callahan, Robbins, & Vesselinov, 2011).
Indirect evidence of the preliminary success of this supplementary service lies in the range of services to which referrals were made, indicating that the needs and vulnerabilities of women were being identified by the ACRW service. Although these records did not include information about whether the women's needs were later met by community-based services, there was a high level of agreement between recommendations made by the service and final disposals and a decrease in the use of custodial remand over the time-period of the project. A likely impact is further supported by qualitative evidence which suggested that recommendations reduced the need for adjournments, were valued by court officials and increased confidence in the use of community alternatives. It is arguable that the services should be making some attempt to record simple outcomes like acceptance into or engagement with services to which referredeffectively auditing an aspect of their work. A higher level of evidence in this area is also required, including use of matched comparison groups. As this initiative has been introduced in only a small number of areas, given modest resource, this should be easy to do without any compromise in ethics or court procedure.
This evaluation used prospectively collected data and included all contacts with the service over a defined time period, but it has several limitations. First, the populations attending inner London courts may have particularly high levels of need compared to other areas of England and Wales and other jurisdictions, meaning that these preliminary findings may not be generalisable. In addition, while we were able to show the feasibility of this service and extract data of reasonable quality from its records, there was no opportunity within these services, and thus our evaluation, to assess any impact on longer term outcomes. While the high level of agreement between recommendations and actual Court decisions and the downward trend in custodial disposals are promising, they are not enough.

| Future directions and gaps in knowledge
The service enhancement evaluated here is a promising example of a way to reduce custodial remand by ensuring that needs of women in the criminal justice system are properly assessed as soon as possible after arrest, but at any rate before the first court appearance. They may then be referred to appropriate services, while, at the same time, recommendations can be made to the courts regarding non-custodial disposal options. Next steps must include collecting data on the effectiveness of these services in terms of improved health, social inclusion and desistance from offending. The current lack of more definitive outcome information is a problem shared by other areas of health research in the criminal justice system. Reasons include the over-arching influence of national policy directives which determine the flow of monies in the field, the general lack of research funding in this area and a perceived need to implement short-term solutions quickly. High quality quasi-experimental designs are feasible, and economic calculations should be included because reducing custodial disposals is likely to reduce public expenditure, but it is necessary to show this. Further, there should be consideration of how interventions can be optimally funded and where evidence can be leveraged most effectively (McDaid et al., 2008). Given the multiple needs among these women, it may be most practical to implement change through health, social and criminal justice services being full partners at governmental level as well as in the field.
Finally, there are a number of stages across the rest of the criminal justice pathway that could be targeted to help reduce the use of custodial remand for women and other vulnerable groups. Research is needed with vulnerable groups across the whole pathway to ensure that the most effective approaches are provided, then further adapted as required.

| CONCLUSION
This study confirms and extends information on the range and extent of need among women entering the criminal justice system. It documents the extent to which a quick-response service run by staff who have training in screening for mental health and substance misuse problems and knowledge of community service options may support other layers of service at the beginning of a woman's pathway through the criminal justice system. The courts appear to take notice of advice directly from this service as well as through the outcome of a next stage referral to the Liaison and Diversion Service. Preliminary findings with respect to the range and extent of service referrals made now require further evaluation, through research projects designed for that purpose. This would help to determine the extent to which women follow through on such referrals and the longer term outcomes for mental health and desistance from offending.