Who are the victims?

The relationship between violent women offenders and their victims follows a different pattern to men. Violence by women is commonly characterized by interpersonal conflicts. The evidence suggests that the victims of homicide perpetrated by women are more likely to be close family members (Cooper and Smith, 2011; Flynn et al., 2011; Hakkanen-Nyholm et al., 2009). Findings from Flynn et al. (2011) for England and Wales have shown that the victims of women convicted of homicide were overwhelmingly immediate family members, either a spouse/partner/ex-spouse partner (36%) or a son or daughter (19%).

Mental illness in women who commit homicide

The association between violence and mental illness has been the focus of much academic and clinical research. In a recent study of serious violence in England and Wales over a 12-month period, Flynn et al. (2013) found 8% of serious violent offenses in the general population were committed by women and 10% of these women had been recently under the care of mental health services. That study was based on a national sample, but other research has been undertaken in specific populations. For example, findings of Monahan et al. (2001) from the MacArthur Risk Assessment Study on violence by patients following discharge from psychiatric hospitals found no overall significant difference in offending by gender. However, when specific diagnoses were examined, Fazel and Grann (2006) revealed that compared to men, women with psychosis had a greater attributable risk of violence in all age groups. Serious mental illness has been shown to be an important feature in homicides by women. Flynn et al. (2011) showed 28% of women compared with 20% of men had symptoms of mental illness at the time of offense and half had a prior history of mental disorder compared to less than a third of men (50% vs. 30%).

It has been argued that research relying on pretrial forensic evaluations of the perpetrator as the main measure of mental illness can be subject to bias and potentially overestimate mental illness in female offenders (Eronen, 1995; Nielssen, 2007; Schanda et al., 2004). Consequently, there is evidence to suggest a leniency towards women. Although there is evidence to support potential bias, particularly in sentencing (Hoyt and Scherer, 1998; Flynn et al., 2011), it is difficult to prove due to numerous confounders that may influence the judicial outcomes. The clinical differences between men and women with mental illness committing violent acts may be moderate but this does warrant further research to ensure women are receiving services that are appropriate to their needs.

Systems of care for women offenders

The system of care for women offenders is constantly changing. Sexual segregation was introduced into prisons by the middle of the nineteenth century doing much to reduce prison pregnancy and the exploitation of vulnerable women incarcerated alongside men (Zedner, 1998). This was echoed in the first provision of secure hospital care for women in the UK in 1860. The current, interconnected system of care, treatment, supervision and punishment is designed to ensure that women who offend are correctly placed in terms of the gravity of their actions, the presence of treatable mental disorders and the need for supervision, in order to protect the public and to prevent re-offending. The particular roles of health, social care and criminal justice agencies are both specific and different (Bartlett and Kesteven, 2009), but in practice much more similar than many practitioners might wish to admit.

Currently there are only about 50 women in high secure hospital care and 1,314 in enhanced medium, medium and low secure care in the UK (Bartlett et al., 2014b). These units undertake assessments and treatment of women with and without criminal convictions who are thought to require secure hospital environments. A small proportion of women remanded to prison are transferred to hospital care under the 1983 Mental Health Act (revised 2007). Both the National Health Service (NHS) and the Independent Sector health care providers have secure hospital beds and undertake prison health care, now commissioned by NHS England, in line with the changes in the Health and Social Care Act (2012). Third-sector providers also have roles in liaison and diversion schemes whose intention is both to avoid unnecessary remands in custody, particularly of women with major mental illness, and also to foster rehabilitation and prevent family disruption.

Ways of understanding the needs of women offenders

There is no single agreed way to conceptualize the needs of women offenders. There are a number of cogent frameworks evident in clinical work, research, policy initiatives and in funding for service innovation. These include:

None of the frameworks considered here are exclusive but there is considerable work to be done to establish a shared understanding across and within agencies with responsibility for women offenders.

Epidemiological studies (e.g., Coid et al., 2000; O’Brien et al., 2003) have tended to rely on psychiatric diagnostic systems based on standardized assessments and, where possible, large-scale projects. These have generated a comprehensive account of the psychiatric problems of women in prison and to some extent in secure hospital care, but are limited by using gender-blind approaches. Studies on self-harm in women and suicide in both prisons and hospitals (Marzano et al., 2010; Shaw et al., 2004) have been revealing, and environments and practice have changed in women’s prisons in response to concerns about preventable deaths (NOMS, 2013).

Trauma-informed formulation both of individual women and of the environments in which they live have influenced policy in this area and led to specific interventions (Covington, 1998; Najavits et al., 1998). This discourse highlights the impact in adult life of earlier sexual and physical abuse and creates a narrative for individual women to understand adult problems, particularly repetitive self-harm and pseudo-hallucinatory voices. It allows practitioners in both custodial and clinical environments to grasp that systems of care and supervision (e.g., male staff, restraint) can echo, inadvertently, unwanted earlier experiences of disempowerment and coercion, especially by men (Heney and Kristiansen, 1997; Henderson et al., 1998, Aitken and Logan, 2004; Benda et al., 2005). The fact that many women offenders have involvement in sex work and are further abused as adults (see US prison literature on this especially) reinforces the relevance of this approach.

There has been a range of policy and practice initiatives in the field of Personality Disorder (PD) since 1999. The impetus behind the Dangerous and Severe Personality Disorder (DSPD) programme (Eastman, 1999) was the management of mainly male violence and the only tangible outcome for women was the creation of the Low Newton 12 bedded DSPD unit. Changes in the direction of policy and the funding of initiatives by government have meant that there is now an offender PD strategy for women (Department of Health/NOMS, 2011). The epidemiology established that patterns of PD in male and women offenders are different and that Emotionally Unstable Personality Disorder (Singleton et al., 1998), for which the evidence base for treatment is slightly stronger, is more common in women. The direction of policy is not to deliver treatment per se, but to promote interventions and practical support and enhanced understanding within the Criminal Justice System (rather than in the health sector) for those likely to have a PD. National Offender Management Service (NOMS) commissions a range of providers to undertake this work including, but far from exclusively, health providers. “PD: no longer a diagnosis of exclusion” (Department of Health, 2003) does not seem to have led to comprehensive PD services in the community and a range of nonhealth providers are responding with innovative provision of services (e.g., Women in Prison).

There is no accepted definition of the term “complex needs.” In practice, it appears to have become shorthand for multiple health and social care, gender-specific needs (children and separation, the avoidance of cycles of deprivation and the management of disrupted attachments) in women offenders. A formulation of complete needs incorporates social roles as well as psychological issues and has a practical focus on accommodation, training and educational opportunities.

In the UK, there is a large gap in conceptual frameworks around drug and alcohol problems. Both are very common in women offenders, in hospital and prison populations, often in combination with other difficulties (Bland et al., 1999; Coid et al., 2000; Singleton et al., 1998; O’Brien et al., 2003; Abram et al., 2003). The separate commissioning structure in substance misuse services, including in prison, could be argued to have impaired a comprehensive, gendered understanding of women’s use of legal and illegal substances in the UK. Detoxification services and psychosocial programs focus on the addiction itself, rather than the person with the addiction and have protocol driven approaches. Within prisons, until recently, psychosocial programs neglected alcohol in favor of street drugs. Therefore, the reasons why women use alcohol and drugs, their patterns of use and the links with trauma and mental health problems can be poorly operationalized in clinical practice, despite a robust evidence base in the United States (Peugh and Belenko 1999; McCellan et al. 1997). This undermines a coherent understanding of the reasons for abuse/dependence and also the consequences in terms of well-being (Chapter 15).

These competing but overlapping frameworks may provide a degree of flexibility in terms of models of care but risk leaving frontline practitioners at sea and talking past each other. Multiple reviews of women offenders have made recommendations for treatment (Henderson, 1998; Baletka and Shearer, 2001; Nicholls et al., 2001). However, treatment based on these concepts has been poorly anchored by evidence of efficacy per se or of cost effectiveness; this is curious given the size of the populations and the cost of institutional rather than community care and supervision.

Evidence that treatment and interventions work?

There is, therefore, a sense of despondency in some writing about women offenders. Authors note repeatedly the absence of interventions in situ (Covington, 1998; Ford et al., 2013) as well as the lack of an evidence base for efficacy (Bartlett, 2001, 2007; Nee and Farman, 2007; Drapalski et al., 2009).

Possible frameworks of evaluation would include: mental health outcome measures, length of stay required for change, quality of life, re-offending rates and community survival. Where possible these would use similar tools of measurement to maximize useful cross-referencing. This would inform commissioners about cost-effective care and improve the chances of sustainable effective services being provided. Patient generated measures have been strikingly absent from the evaluations that do exist but would warrant a place in contemporary research (Bartlett et al., 2014c).

Bartlett (2001) reviewed the evidence base for treatment studies in secure settings and found sparse information on women leaving secure hospital care and those women who had received care and treatment in medium and high secure care had contested its value at the same time as it being far from clear what was delivered to them. The exception to the general statement about poor evaluation was in the US prison system where treatment initiatives directed at trauma symptoms and substance misuse were showing promising findings.

These observations remained largely true 6 years later (Bartlett, 2007) compounded in the UK by a recognition that even the size of the secure hospital population of women in the UK was unknown at that time and needed mapping. Currently, the psychiatric epidemiology of the women prison population is seriously out of date as the last major study was over 15 years ago. This creates real difficulty in designing services and evaluating interventions.

Interventions in secure hospitals

Existing descriptions of the women secure population indicate diagnostic heterogeneity (Coid et al. 2000; Bartlett et al., 2007) with high proportions of the women having multiple diagnoses (Bartlett et al., 2014b).

Long is conspicuous in the UK in articulating a model of gender sensitive care (Long et al., 2008) and delineating and assessing hospital interventions in a modest-sized cohort (Long et al., 2010, 2011a, 2011b, 2012, 2013). Hitherto the treatment offered has been poorly described and change, whilst discharge and community survival had not been attributable to specified interventions. Evidence from this cohort suggests that group interventions within secure services reduce symptoms of anxiety and suicidality, that treatment engagement reduces risk and that short length of stay is associated with treatment completion. Noncompletion of interventions is common and has economic as well as individual health implications.

Nee and Farman (2007) reviewed dialectic behavioral therapy (DBT) in hospital settings noting encouraging results in terms of reductions in self-harm and psychiatric symptoms in low and high secure inpatient and outpatient settings. Barr et al. (2013) examined the usefulness of high support housing after discharge for women in secure hospital care, finding it to be linked to improvements in psychological health, without increased risk.

These recent findings are important given that traditionally women have had longer lengths of stay than men in high secure care (Butwell et al., 2000) and have been more likely to be readmitted to medium secure hospital care (Coid et al., 2000). Maden et al. (2004, 2006) followed up women from secure hospital care (who were markedly different from men also discharged) and found they had a reconviction rate of 9% at 1 year, less than the rate for men (16%).

Interventions in prison

The paucity of robust empirical research on the expensive secure hospital population of women sits in some contrast to the larger, although still inconclusive body of work conducted in prison. Almost without exception, this work is conducted outside of the UK and so its importance for the UK is debatable. Much is pilot or small-scale research. Programs that ostensibly target one problem can also be seen to have more widespread benefits. Key outcome measures include recidivism but are not restricted to this.

Substance misuse programmes show an impact on re-offending (Dowden and Blanchette, 2002) better compliance if women remained in contact with their children (Nishimoto and Roberts, 2001) and desistance if women were moved away from their home area (Strauss and Falkin, 2001). Drapalski et al. (2009) noted greater help seeking in women in prison, not least because they sought help with traumatic sequelae. Staton-Tindall et al. (2009) noted that some gender-specific aspects of previous community treatment have potential implications for treatment success in prison.

Trauma programs, which can include substance misuse work, resulted in fewer symptoms of post-traumatic stress disorder (Valentine, 2000; Valentine and Smith, 2001; Zlotnick et al., 2003) in some early studies but negative findings have also been reported (e.g., Bradley and Follingstad, 2003). Messina et al. (2010) took a trauma-informed approach to intervention with substance misuse women in prison. In a randomized trial comparing the gender-responsive treatment (GRT) to a therapeutic community, GRT produced better integration into aftercare, less re-incarceration and greater reduction in drug use. Ford et al. (2013), reviewing this body of work, argued that it was hard to specify what was responsible for the mixed results. Nee and Farnam (2005) evaluated DBT with 30 prisoners. Despite a high attrition rate completers did better than noncompleters and controls. In a pilot study in a UK women’s prison, Gee and Reed (2013) found modified DBT to be helpful in producing fewer incidents warranting adjudication, less use of self-harm management processes in prison and improved mental health. Gussak (2009) evaluated art therapy with male and women prison inmates and found reductions in depression and improved locus of control after a period of group treatment.

Interventions in the community

The population statistics would suggest this is a key area of intervention. However, by comparison with both prisons and secure hospitals, the evidence base for work with women offenders in the community is even more limited. Jolliffe et al. (2011) reviewed the impact of women’s centers and found equivocal support for this new area of investment which is currently subject to transformation and further scrutiny (NOMS, 2013a). Unknown numbers of women in the UK are under the care of community forensic health provision. Van den Bosch (2012) notes the relevance of mainstream treatments for this group.

In sum, there is a need to capitalize on programs of intervention and scale up studies to provide an evidence base that would allow for specification in national services and for individuals to have some degree of personal choice in what would help them.

Future developments

Recent interest in women offenders has improved understanding of both their criminality and associated health and social care needs. It is clear that women are less likely than men to have contact with the criminal justice system and that they will rarely be involved in serious violent offending. However, their involvement is routinely complicated by profound health and social care difficulties, which demand a response from relevant criminal justice and other agencies that includes a gender dimension. In the UK and elsewhere a range of agencies has been developed to meet these needs but until recently there was little evidence to support one intervention rather than another. In the UK, hospital or prison-based services have routinely lacked a theoretically driven approach to care. Encouraging findings for trauma-informed interventions, work on substance misuse and more detailed articulation of some secure hospital programs all need pursuing. A conspicuous gap, which stands out in terms of population statistics, is work with women offenders in the community, where most reside, most of the time.