What is the Evidence for Offense-Specific Group Treatment Programs for Forensic Patients?

abstract Evidence-based practice (EBP) advocates that treatments offered to patients should be empirically supported and effective. Group-based treatment is offered in forensic mental health services as a way of assisting forensic patients (FPs) to address their offending behavior. However, little research exists examining how research and practice are integrated in these interventions, consistent with the principles of EBP. This study examined the utilization and evaluation of offense-specific group treatment programs with FPs, with a specific focus on interventions for substance misuse, sexual offending, firesetting, and violent offending. The results highlight that despite frequent use of offense-specific group interventions for FPs in UK forensic mental health services, evaluation is lacking regarding both published evaluations and through routine clinical practice. To ensure consistency with principles of EBP, recommendations are made surrounding the implementation of routine evaluations (e.g., follow-up studies) of offense-specific group treatment programs with FPs.

Individuals detained in forensic mental health services often have complex presentations including a range of mental health issues, serious behavioral problems, and specific criminogenic needs. Forensic mental health services are tasked with reducing forensic patients' (FPs') risk of reoffending, assessing, and treating their mental health needs, and promoting reintegration and recovery (Vojt, Slesser, Marshall, & Thomson, 2011). Offense-specific interventions form an essential part of treatment in forensic settings, through addressing both the psychological and criminogenic needs of patients (Duggan, 2008); consistent with the Risk Need Responsivity Model (Andrews & Bonta, 2010). Interventions abiding to these principles have been shown to have positive results at reducing recidivism (Hanson, Bourgon, Helmus, & Hodgson, 2009).
Qualitative research has highlighted that group interventions can positively contribute toward FPs progress and recovery (Clarke, Tapp, Lord, & Moore, 2013;Mead & MacNeil, 2006;Tapp, Warren, Fife-Schaw, Perkins, & Moore, 2016), suggesting that they have the potential to provide both a therapeutically and fiscally effective form of therapy through providing treatment for multiple FPs simultaneously (Duggan, 2008;Rees-Jones, 2011). Due to the potential therapeutic and fiscal benefits of group treatment, offense-specific treatment groups are frequently utilized in UK forensic mental health services to address factors related to FP's offending behavior. Although qualitative research suggests that there are potential benefits of group treatment for FPs, understanding the utilization and effectiveness of offense-specific group interventions for FPs is of critical importance (Rees- Jones, 2011;Sturgeon, Tyler, & Gannon, 2018).
EBP is considered the "gold standard" for the delivery of clinical care and psychological therapies and represents the effective implementation of scientific evidence into practice (British Psychological Society, 2011;Gannon & Ward, 2014;National Institute for Health & Care Excellence, 2014). EBP comprises of three key tenets: (1) identifying and integrating the best research evidence as to whether and why a treatment should work (e.g., using empirically supported therapies, using the best theoretical and empirical evidence to guide treatment content/targets, evaluating the effectiveness of the intervention); (2) using clinical experience and expertise when applying research to practice, particularly in under-researched areas; and (3) incorporating client's values and therapeutic preferences (e.g., responsiveness, engagement, whether individual or group treatment is more appropriate) (Lilienfeld, Ritschel, Lynn, Cautin, & Latzman, 2013). Evaluation of interventions is a key component of EBP, and is essential for informing practitioners of the types of treatment that provide positive results with FPs; thus, ensuring patients receive the best treatment available and advancing clinical practice and knowledge in the field (Blackburn, 2004;Glorney et al., 2010;Tapp, Perkins, Warren, Fife-Schaw, & Moore, 2013;M€ uller-Isberner & Hodgins, 2000). Assessing the effectiveness of offense-specific interventions is of particular importance for practitioners working in forensic settings as ineffective interventions have the potential to be detrimental for patients or elicit adverse outcomes. For example, risk of retraumatizing individuals through discussion of adverse childhood experiences or offense-related information (Mueser, Rosenberg, Goodman, & Trumbetta, 2002;Doob, 1992), or risk of increasing recidivism rates for completers and/or non-completers through not effectively targeting key risk factors associated with their offending (Jones, 2007;McMurran & Theodosi, 2007;Mews, Di Bella, & Purver, 2017).
As highlighted earlier, FPs present with a complex range of psychological, criminogenic, and mental health needs, which are likely to interact in terms of initiating and maintaining offending behavior (Nagi & Davies, 2010). Subsequently, standardized group treatment programs developed for non-mentally ill offenders (e.g., in prisons) may not be suitable or adequate to address the needs of FPs and may require adaptation to provide specific focus on issues pertinent to this client group (i.e., the relationship between mental health and offending or risk behaviors). To address this, many forensic mental health services have taken to devising their own group treatment programs or adapting existing prison group treatment programs to target the specific needs of FPs (Sainsbury Centre for Mental Health, 2008). However, little attention has been given to examining the effectiveness of adapted interventions with FPs (Barnao & Ward, 2015;Davies, Howells, & Jones, 2007;Grubin, 2001;Howells, Day, & Thomas-Peter, 2004). Given clinicians should be cognizant of the need to apply the tenets of EBP when developing or implementing treatment with FPs, a potential deficit may exist in its application with this group.
Despite the frequent use of offense-specific group treatment programs in forensic mental health settings it is clear that there is a pressing need for evaluation research to guide rehabilitation with FPs (Barnao & Ward, 2015;Duncan, Nicol, Ager, & Dalgleish, 2006;Knabb, Welsh, & Graham-Howard, 2011;Sturgeon et al., 2018). Previous research in the area predominantly consists of reviews synthesizing existing published evaluations of interventions with FPs. However, these have tended to focus on either particular groups of FPs' (e.g., high secure patients; Sturgeon et al., 2018), interventions utilizing specific rehabilitation models or treatment modalities (e.g., Barnao & Ward, 2015;Knabb et al., 2011), or skills based interventions (e.g., Duncan et al., 2006). Further, previous work in this area has focused solely on published studies, and therefore may neglect good quality evaluation research which has taken place as part of standard clinical practice. There has been no research to our knowledge that has focused specifically on examining both published and practice based evidence for offense-specific group interventions for FPs.
This study aims to synthesize the evidence on the utilization and evaluation of offense-specific group treatment programs with FPs in UK forensic mental health services. The current study aims to extend previous research in this area in two ways: (1) by examining the extent and quality of the existing published evidence in relation to offense-specific group treatment programs for FPs; and (2) exploring practitioners' work regarding the delivery and evaluation of existing and planned offense-specific group treatment programs in UK forensic mental health services.

Method and results
To understand the available evidence for group-based offense-specific treatment programs with FPs, the current study consisted of two phases. Phase 1 consisted of a review of the literature to identify all published evaluations of offense-specific interventions addressing substance misuse, sexual offending, firesetting, and violent offending for inpatients in UK forensic mental health services. These offending behaviors were selected as they are the most common reasons reported by admission to forensic mental health services in the UK Rutherford & Duggan, 2008;V€ ollm et al., 2018). To compliment this, Phase 2 consisted of a national survey examining practitioners' preexisting and planned future use of group-based offense-specific interventions and any existing or planned evaluations of these. Both published evaluations (Phase 1) and survey responses (Phase 2) regarding individuals with Intellectual and Developmental Disabilities (IDD) were excluded from the current study as the presence of an IDD is considered to be an additional responsivity factor that needs to be accounted for in treatment programs. Consequently, interventions for individuals with IDD are often "adapted" to meet the literacy, social, and cognitive needs of this group to ensure their accessibility (Melvin, Langdon, & Murphy, 2017;Keeling & Rose, 2006;Williams, Wakeling, & Webster, 2007). As a result, this group are often examined separately to those who do not have an IDD.

Ethical considerations
Formal ethical approval was not required for Phase 1 as this consisted purely of data extraction from existing published studies. Consistent with the Department of Health (2012) governance arrangements and similar research in the area (e.g., Tapp et al., 2016), ethical approval was not required for Phase 2 as all survey respondents were healthcare professionals and the information collected focused solely on general service provision, and no information was collected about individual patients or staff. However, best practice principles, consistent with APA (2017) ethical standards, were applied.

Phase 1 -literature search
Procedure A search was conducted in 2017 of several online databases (e.g., Google Scholar, HeinOnline, PubMed, PsychINFO, Web of Science) as well as hand searches of relevant reference and citation lists, and doctoral thesis abstracts, to identify all available published evaluations (in English) of offense-specific interventions addressing substance misuse, sexual offending, firesetting, and violent offending for inpatients in UK forensic mental health services. Search terms used included: evaluation, AND [substance use, firesetting, arson, sex offending, violent offending], AND treatment OR intervention, AND secure forensic OR mentally disordered offenders OR forensic mental health. Articles were only included if they reported information about the effectiveness of an offense-specific group treatment program (descriptive, qualitative, or quantitative in nature) which had been delivered with patients in UK forensic mental health services for either substance misuse, sexual offending, firesetting, or violent offending. Evaluations of interventions for those with an IDD were excluded to focus the review on those interventions specifically designed and evaluated for those with a mental illness. Further, interventions which had been delivered and evaluated with individuals in prison who had a mental disorder or which focused on factors proximally related to offending (e.g., anger management, social skills, cognitive skills, psychoeducation or CBT for mental illness) were also excluded from the review so as to concentrate on interventions directly targeting offending behavior.
Once retrieved, either the first or second author reviewed the content of the article to ensure that it met the inclusion criteria. Once all relevant articles had been identified each study was coded using the Maryland Scientific Methods Scale (SMS; Sherman et al., 1997). The SMS is a five-level classification system which has been extensively used in criminology to assess the evidence base for interventions according to the quality of the methodological design of evaluation studies, based upon its internal validity. Levels range from (1) basic correlation between pre-and post-measures, to (5) randomized control trials Either (a) a basic correlational design between an intervention and an outcome measure of reoffending or, (b) a cross-sectional comparison of treated groups with untreated groups, not matched on time or other relevant variables (e.g., comparing the reoffending rate of the treatment group to previously reported base rate of reoffending) 2 Either (a) within treatment change (e.g., pre-post intervention outcomes) for a treatment group with no inclusion of an untreated comparison group, or (b) comparison of reoffending rates for a treatment group and a comparison group without direct statistical comparisons being made between the groups 3 Comparison of outcome measures for a treatment group and a nontreatment comparison group. Statistical evidence presented on the comparability of the two groups (i.e., basic quasi-experimental studies). 4 A comparison study incorporating both a treatment and a non-treatment comparison group where relevant factors are controlled for or the groups are matched on these relevant factors (i.e., controlled/matched quasi-experimental designs). 5 A comparison study where participants are randomly allocated to either a treatment or non-treatment comparison group, direct statistical comparisons are made between the groups on outcome measures (i.e., randomized control trials).   Table 1). Qualitative and descriptive studies do not form part of the SMS, however, they would form tiers below Level 1 (basic correlation) evaluations. Level 3 studies (within treatment effectiveness with a control group) are considered as good quality preliminary evaluations and the minimum evidence required to draw conclusions about the effectiveness of an intervention (Hollin, 2008).

Results
The results of Phase 1 are presented in Table 2.
Twenty-three published studies were identified that had evaluated the efficacy of offense-specific group interventions for substance misuse, sexual offending, firesetting, or violent offending with either male or female forensic mental health patients in the UK. The majority of published evaluations identified were for group-based interventions targeting substance misuse (n ¼ 12), followed by violent offending (n ¼ 5), then firesetting (n ¼ 4), with interventions for sexual offending having received the lowest number of evaluations (n ¼ 2). Five studies represented evaluations of two programs for substance misuse.
In terms of quality, the majority of evaluations were classified as either falling below the lowest SMS tier (e.g., basic intervention descriptions or qualitative evaluations; n ¼ 5) or basic pre-post treatment assessments of those completing treatment with no control group (Level 2; n ¼ 13). There were no quantitative evaluations of group treatment for sexual offending and only two for firesetting. Only four studies were considered as meeting the criteria for a Level 3 evaluation, one for substance misuse, one for firesetting, and two for violent offending. Only three of those studies classified as Level 3 evaluations included a follow-up period, two for violence and one for substance misuse. Sample sizes across the majority of evaluations were small ( 30; n ¼ 16, 69.57%) and comparison groups (where used) were unequal in size.

Procedure
To examine current practice with regard to the utilization and evaluation of offense-specific group treatment programs in the UK, a national survey was conducted. The survey ran across low, medium, and high secure UK forensic mental health units between September 2016 and January 2017 to examine the existing and planned use and evaluation of group-based offense-specific offending behavior programs for substance misuse, sexual offending, firesetting, and violent offending. An online questionnaire was developed for this purpose which was adapted from one used in previous research in this area (i.e., Palmer, Caulfield, & Hollin, 2005. The online questionnaire consisted of a series of questions which captured information regarding services current and future planned provision of group interventions for substance misuse, sexual offending, firesetting, and violent behavior, with a particular focus on the availability, development, content, implementation, and evaluation of each type of group offending behavior program. The survey collected demographic information about participants and their service and then was split into two further sub-sections on their previous and future use of offense-specific group treatment for each type of offending behavior (i.e., substance misuse, sexual offending, firesetting, violent behavior). Please see Appendix 1 for an overview of the survey questions included in this study.
A total of 85 forensic mental health inpatient services were invited to participate in the survey. This included the four high secure hospitals in the UK, as well as low and medium secure services identified through the Royal College of Psychiatrists Quality Network for Forensic Mental Health Services (QNFMHS). The survey platform Qualtrics was used for online distribution of the survey. The survey link was distributed via several professional forums and newsletters, in addition to being emailed directly to lead psychologists at inpatient forensic mental health services. Participants were presented with information about the study prior to completing the survey and provided informed consent before participating.
A total of 56 surveys were recorded by Qualtrics as having been started with at least one full section completed. Survey forms returned from services who provided mental health care within prisons (e.g., mental health inreach; n ¼ 6) were removed for analysis since the focus of the research was on forensic mental health services. Survey responses from services who only provided treatment for individuals with IDD were excluded from the current study (n ¼ 5). Further, surveys which were duplicates (i.e., returned from the same service; n ¼ 16) were removed for analysis. 1 The final sample therefore consisted of completed questionnaires from 29 secure units (34.12% of those units contacted). This response rate is similar to that of previous research using similar methodologies (Palmer et al., 2005(Palmer et al., , 2007Tapp et al., 2016, 22.7%). In terms of respondents, 21 of the 29 surveys completed were done so by a member of psychology (i.e., assistant, trainee or qualified psychologists), 3 by psychiatrists, and 4 by nurses (including a nurse therapist). Responses were distributed across all three security levels: low (n ¼ 8, 27.59%), medium (n ¼ 19, 65.52%), and high security (n ¼ 2, 6.90%). Fifty-five percent of services (n ¼ 16) reported that they provided care for just male FPs and 44.83% (n ¼ 13) reported providing care for both male and female FPs.

Results
Substance misuse group treatment programs. Of the 29 forensic mental health units, from which responses were included in the final analysis, 25 (86.21%) reported having provided group treatment for substance misuse in the past 5 years. The majority of group treatment programs were reported as having been developed "in-house" (n ¼ 20, 80.0%). Those services who had bought in manualized programs for group substance misuse treatment were predominantly using programs that were not originally developed for use with forensic patients (e.g., developed for prison populations, adolescents, dual diagnosis patients). All substance misuse group treatment programs were reported to be underpinned by a combination of empirically supported therapies (e.g., Cognitive Behavioral Therapy, Motivational Interviewing, Dialectial Behavioral Therapy, Psycho-Education). In terms of evaluation, fifteen out of the twenty services who had developed and delivered group treatment programs "in-house" for substance misuse reported that they had not conducted any evaluation of the intervention's effectiveness (75.0%). Of the five services which had conducted evaluations of their group intervention, two reported that this had consisted of examining within-treatment change on pre-post treatment psychometrics for group attendees. The other three services had either examined within-treatment change including a post-treatment psychometric follow up (n ¼ 1), examined abstinence rates following treatment (n ¼ 1), or had examined patients satisfaction with the group (n ¼ 1) (see Table 3). Two of the five services who had conducted evaluations had published these.
Future substance misuse group treatment programs were reported to be planned in 17 of the services surveyed (58.62%). The majority of these reported that their future planned groups will be consistent with those previously run in their service (n ¼ 13, 76.47%). Three services reported that they were currently developing new interventions for substance use "in-house" and one service reported that they did not know what their future substance misuse program would look like. Five services reported that they were planning evaluations of their future substance misuse groups. Three of these services had not previously offered group-based substance misuse treatment, while the other two services had not previously evaluated their existing group treatment program. Evaluation designs considered by these services were predominantly audits of pre-post treatment psychometric data, with one service also considering including post-treatment follow up data, and another including an evaluation of participant satisfaction.
Sexual offending group treatment programs. Eight out of 29 services (27.59%) reported having provided group treatment for sexual offending in the past 5 years. Individual treatment for sexual offending was reported as being more common than group treatment, with 17 units (58.62%) reporting providing treatment in this format. Of the eight services who reported providing group treatment for sexual offending, six reported having developed their group program "in-house" (75.0%). All sexual offending group treatment programs were reported to be underpinned by a combination of empirically supported therapies (e.g., Cognitive Behavioral Therapy, Motivational Interviewing, Psycho-Education). Only two of the services who had developed and delivered group treatment programs "in-house" for sexual offending had conducted an evaluation of the intervention's effectiveness (25.0%) and only one of these provided details about their evaluation (see Table 3). This service had conducted a qualitative evaluation of participants' experience of the group, however, no quantitative evaluation had taken place.
Nine of the services surveyed reported that they had future plans to run a sexual offending treatment group (31.03%). Three of these reported that their future group treatment for sexual offending would be developed in-house (33.33%). Five sevices (55.56%) reported that they would continue to provide their existing sexual offending group program and one reported that they were planning to bring in a group treatment program that had been developed externally. In terms of evaluation, four services reported that they were planning evaluations of their future sexual offending group treatment programs. Two of these had not previously ran group treatment for sexual offending and the other two had not previously evaluated their existing group treatment program. Evaluation designs considered by these services predominantly focused on examining within treatment change on pre-post treatment outcome measures (e.g., psychometrics, risk assessments).
Firesetting group treatment programs. Nine forensic mental health services (31.03%) reported that they had previously provided group treatment progams for firesetting in the past 5 years and 12 (41.38%) reported that they had provided individual treatment for firesetting. Five secure units did not respond to this section of the survey (17.24%). Seven out of nine services who had provided group treatment for firesetting reported having developed their program "in-house" (77.78%). The two services that reported having bought in manualised programs for group-based firesetting treatment were using programs specifically developed for forensic mental health patients. All group treatment programs for deliberate firesetting, regardless of whether they were developed in-house or bought in, were reported to be underpinned by a combination of empirically supported therapies (e.g., Cognitive Behavioral Therapy, Motivational Interviewing, Dialectiacal Behavior Therapy, Psycho-Education). Four out of nine of services who had developed and delivered group treatment programs "in-house" for firesetting reported that they had had conducted some form of evaluation of the program (44.44%). Of the four services who had conducted evaluations of their group intervention, two had joined together to lead a multi-site quasi-experimental evaluation (i.e., pre-post treatment change using psychometric measures with a treatment group and a comparison group), one had examined within treatment change on psychometric measures for those who attended the group as well as obtaining qualitative feedback from patients, and one other had conducted a qualitative examination of patients' experience of the group (see Table 3). Three services reported that they were planning to evaluate their firesetting group treatment program in the future; all three services reported that they would evaluate their inervention by examining at within treatment change on pre-post treatment outcome measures, with one planning to include a reoffending arm as well.
Eleven services reported that they had future plans to run a firesetting treatment group (37.93%). Three of these services reported that they were looking to implement an existing standardized evidence-based group treatment program that had been developed externally, three reported that they were planning to develop a group intervention "in-house," three reported that they were looking to continue with their existing provision, and two were unsure what their future provision would look like. In terms of evaluation, only two services who reported to be developing "in-house" group programs for firesetting also reported that they had plans to evaluate these.
Violent behavior group treatment programs. Seventeen services (58.62%) reported having provided group treatment for violent behavior in the past five years. Six units did not complete this section of the survey (20.69%). Ten out of seventeen services reported having developed their group treatment program for violent behavior "in-house" (58.82%). Seven services reported that they had bought in manualized programs for violent offending group treatment from an external source (41.18%); four of these reported using programs that had been specifically developed for use with FPs (57.14%). Violent offending group treatment programs were reported to be underpinned by an array of empirically supported therapies (e.g., Cognitive Behavioral Therapy, Motivational Interviewing, Dialectial Behavioral Therapy, Acceptance and Commitment Therapy, Mentalisation Based Therapy, Psycho-Education). Five services who had developed and delivered group treatment programs "in-house" for violent offending reported that they had evaluated the intervention's effectiveness (50.0%) (see Table 3). Of these five, three had examined within treatment change on psychometric assessments for group completers and one had examined if there was a reduction in violent incidents post One service did not provide any detail of the methods used to evaluate their intervention. All programs which had been bought in had been subject to external evaluation. Future delivery of violent offending group treatment was reported to be planned by 14 services (48.28%). The majority (n ¼ 10, 71.43%) of these reported that their future groups will be consistent with those previously delivered. Two services reported that they were currently developing new group interventions "inhouse" and one further service reported that they were in the process of introducing a standardized violent offending group treatment program which had been developed externally. Only three services reported planning to evaluate their future violent offending behavior groups; details of planned evaluations were provided by just two of these services. One of these services had not previously evaluated their existing violent offending group treatment program and one had conducted an evaluation historically but was looking to extend the quality of this. Evaluation designs considered by these services both focused on examining pre-post treatment psychometric data and violent incidents post treatment.

Discussion
The current study aimed to examine the utilization and evaluation of offense-specific group treatment programs with FPs. In Phase 1 existing published studies evaluating the effectiveness of offense-specific group treatment programs with FPs were collated and examined. In Phase 2 a national survey was conducted examining existing practice in the development, delivery, and evaluation of current and planned offensespecific group treatment programs in UK forensic mental health services. Analysis of the existing published literature highlighted only 23 published evaluations of offense-specific group treatment programs with FPs. The majority of these were small in sample size and ranked low in terms of methodological quality on the SMS (Sherman et al., 1997). The results of the national survey showed that group interventions are commonly used to address different types offending behavior in UK forensic mental health services. However, the majority of these programs are developed "in-house" and only around 25-50% of these were reported to have been subject to some form of evaluation. When "in-house" offense-specific group treatment programs had been evaluated they typically involved examining pre-post treatment change on psychometric measures or qualitative reports from participants; methodologies ranked low on the SMS. When examining the findings by offense-type, group treatment was reported as a common treatment modality for all offenses, however, was most frequently utilized for interventions for substance misuse and violent offending. In terms of evaluation, although group treatment for substance misuse had the largest published evidence base, the majority of services reported that they had designed interventions to target this behavior "in-house" with very few having evaluated the effectiveness of these. Surprisingly, group-based interventions for sexual offending had received the least amount of attention in terms of evaluation both in the published literature and as part of standard clinical practice. In comparison, group treatment for violent offending and firesetting both had a relatively small published evidence base, however, evaluation as part of standard clinical practice was more commonly reported than for other types of offense-specific treatment.
The combined findings from Phases 1 and 2 suggest that at present evidence regarding the effectiveness of offense-specific group treatment programs with FPs is scant, particularly for sexual offending where both published research and evaluation as part of routine clinical practice is scarce (see Figure 1). Further, the findings indicate that evidence pertaining to "what works" with FPs is particularly lacking, with very little attention being given to evaluation of group-based offense-specific treatment programs as part of standard clinical practice. This highlights that more rigorous high-quality evaluations are urgently needed to enable the development of a wider evidence base that is of an appropriate level to inform the expansion and improvement of interventions specific to FPs (Tapp et al., 2016).

Clinical implications
The size and quality of the evidence base for the effectiveness of offense-specific group interventions with FPs is concerning. First, as highlighted in recent reviews, the lack of evaluation of offense-specific interventions for FPs means that practitioners working in these settings have little guiding information as to "what works" with this client group and therefore do not have an empirical evidence base upon which to base their practice (Barnao & Ward, 2015). The current research suggests that practitioners develop interventions using empirically supported therapies found to be effective with general adult psychiatric patients and non-mentally disordered offenders, in the absence of evidence specific to FPs. However, the lack of evaluation (both published and as part of routine clinical practice) means at present it is unclear whether interventions which have been developed and/or adapted for use with FP's are effective in their aims (e.g., reducing criminogenic needs, reducing reoffending, promoting recovery) or whether these could potentially lead to adverse treatment outcomes (e.g., increase in reoffending, decrease in psychological wellbeing). Unfortunately, this means that practitioners working with FPs " … must sail in uncharted seas, doing the best they can to find the most effective treatments for their patients … [until further research evidence accrues]" (Barnao & Ward, 2015, p. 83). Second, due to the limited evidence base, it is not possible to say whether FPs are being offered the most effective treatment for their offending behavior or what treatment works best for whom since there is a lack of studies utilizing different treatment models and methodological designs to draw comparisons.
Together the findings of the current study suggest that at present offense-specific treatment for FPs is not sufficiently consistent with the principles of evidence based practice. While this is concerning, this problem does not appear to be unique to forensic mental health, with research suggesting that practicing psychologists more generally show some ambivalence toward EBP and the utility of research for informing treatment decision making (see Lilienfeld et al., 2013). Given this, it is important to consider what may prohibit practitioners in forensic mental health services from fully engaging in the research/evaluation element of EBP; as without evaluation of care and treatment provision, forensic mental health cannot evolve or improve. Research has identified several common barriers to engagement in EBP including: negative beliefs about EBP and its value in clinical practice (e.g., it restricts individuality), lack of training in EBP and its application, the cost of purchasing EBP materials (e.g., manuals, training packages, new technologies), the lack of inclusion of research in performance targets along with time and funding for such activities, and a lack of awareness of what evidence based products are available (Lilienfeld et al., 2013;Pagoto et al., 2007;Spring, 2007). Given these barriers, it is imperative that commissioners, practitioners, and academics working in forensic mental health come together to address these barriers to ensure offense-specific group treatment is both evidence based and is effective in terms of generating positive outcomes for FPs.

Limitations
Although the current research indicates a deficit in the application of EBP in forensic mental health settings, the results of the research should be considered within certain parameters. First, difficulty recruiting clinicians to complete the national survey led to a relatively low response rate, which may mean that the results of the survey are not reflective of all forensic mental health services in the UK. In particular, responses from low secure services were under-represented in the sample, compared to the number of services available in the UK. However, the response rate in the current study is comparable to that found in previous research of a similar nature (e.g., Palmer et al., 2005Palmer et al., , 2007Tapp et al., 2016). Further, the majority of surveys were completed by psychology practitioners, those professionals who would generally be responsible for providing offense-specific group treatment, so the results are likely to accurately reflect provision across the services who responded. Second, only a limited number of factors/offense types were explored within the survey (e.g., substance misuse, sexual offending, firesetting, and violence), and FPs may present with a much wider-range of offending behaviors and criminogenic needs (e.g., stalking, acquisitive crime, internet offending). In addition, offending behavior may also be addressed indirectly through group treatment programs targeting specific needs associated with these behaviors (e.g., thinking skills, managing emotions, social skills). However, the current survey focused on the provision of offensespecific treatment groups for the most commonly reported reasons for admission to forensic mental health services (Rutherford & Duggan, 2007) and thus interventions likely to be routinely offered across services. Finally, the national survey focused on UK services only and therefore may not reflect practice internationally.

Future directions
Future research should explore the possibility of conducting multi-site research to increase sample sizes to achieve statistical power. Further, future research evaluations should adopt rigorous high quality methodological designs since evaluations employing lower quality research designs (e.g., below Level 3 on the SMS) have been found to be weaker in terms of internal validity and potentially biased in their outcomes (i.e., more likely to report a favorable treatment effect; Weisburd, Lum, & Petrosino, 2001). While RCT's are considered the gold standard of treatment evaluations it has been debated that the nature of withholding treatment from some patients to ensure randomization creates an ethical dilemma. Further, in forensic settings the feasibility of RCT's are also faced with real-world legal issues related to public safety and risk of reoffending by those in a "non-treatment" control group and the liability of staff and organizations related to this (Harkins & Beech, 2007;Hollin, 2008). Good quasi-experimental designs (either basic or controlled) provide the most feasible high-quality evidence of a program's effectiveness (Prendergast, 2011). Inclusion of a follow up period to examine reoffending rates post-treatment will also enable a more holistic examination of long-term effectiveness as opposed to short-term within treatment change (Friendship, Falshaw, & Beech, 2003). However, it is only through growing the existing evidence base that stronger conclusions can be made about the effectiveness of offense-specific group treatment programs for FPs (Barnao & Ward, 2015;Beech et al., 2007;Davies et al., 2007).