A tale of two cultures: examining patient-centered care in a forensic mental health hospital.

Several questions remain unanswered regarding the extent to which the principles and practices of patient-centered care are achievable in the context of a forensic mental health hospital. This study examined patient-centered care from the perspectives of patients and providers in a forensic mental health hospital. Patient-centered care was assessed using several measures of complementary constructs. Interviews were conducted with 30 patients and surveys were completed by 28 service providers in a forensic mental health hospital. Patients and providers shared similar views of the therapeutic milieu and recovery orientation of services; however, providers were more likely to perceive the hospital as being potentially unsafe. Overall, the findings indicated that characteristics of patient-centered care may be found within a forensic mental health hospital. The principles of patient-centered care can be integrated into service delivery in forensic mental health hospitals, though special attention to providers' perceptions of safety is needed.

Introduction

Forensic mental health services must contend with tensions that result from intersecting health and criminal justice policy objectives. Each perspective generates a unique understanding of both the people who use forensic mental health services and the values that underpin forensic mental health systems. From a health perspective, forensic mental health service users are ‘patients’ and the purpose of the system is to provide treatment and support services in order to assist in their recovery. Conversely, from a criminal justice perspective, forensic clients are ‘accused persons’ and the purpose of the system is to detain potentially dangerous individuals, and to reduce risk for violent and criminal recidivism. Forensic mental health services must balance the interests of many parties, including the public, the state, the staff and their professional unions, as well as the patient. Indeed, reconciling these differences and balancing divergent needs is a daunting task – especially in the complex, dynamic environment of a forensic mental health hospital.

Over the past decade, there has been a strong push internationally toward models of mental health service delivery that are patient-centered and recovery-oriented (Mental Health Commission of Canada, 2009; U.S. Department of Health and Human Services, 2009). Patient-centered care describes an approach to providing health care in a manner that emphasizes and respects the needs, values, and choices of patients (Australian Commission on Safety and Quality in Health Care, 2010). It has been delineated into five key dimensions: (a) viewing health in a holistic manner, (b) seeing the patient as a person with needs that extend beyond their illness, (c) sharing power and responsibility with patients and helping them collaborate in their own care, (d) building therapeutic alliance with patients and maximizing the therapeutic value of the patient-provider relationship, and (e) understanding how the personal qualities of the provider influences quality of care (Mead & Bower, 2000). Recovery-oriented care shares many of these same features (U.S. Department of Health and Human Services, 2009). Published research on patient-centered care and the related topics of patient engagement and recovery-oriented care have begun to demonstrate that involving patients in care processes, including the planning and the delivery of health services, will improve outcomes (Resnick & Rosenheck, 2008; Sidani, 2008; Warner, 2010). However, there is a dearth of forensic-related empirical research in this particular area.

Certain characteristics of forensic mental health hospitals, such as the correctional and public safety-oriented milieu, create serious challenges in relation to providing services that are inclusive, collaborative, and egalitarian. Additionally, forensic mental health services are typically mandated by law, which may involve an individual being detained and treated on an involuntary and indeterminate basis. Such a scenario complicates processes aimed at building engagement and facilitating power-sharing. Also, individuals who have been hospitalized for a lengthy duration may become accustomed to being passive service recipients without ambitions for being empowered and actively collaborating in their own care. Additional characteristics of some forensic mental health patients, such as antisocial personality disorder, criminal or violent histories, susceptibility to mental decompensation, poor illness insight, treatment non-adherence, suicidality, and risk of aggression or violence, can present serious impediments to adopting a patient-centered care approach in a forensic mental health hospital (Green, Batson, & Gudjonsson, 2011).

Much of the research literature regarding patient-centered care and related constructs has been built on the foundation of shared decision-making between service providers and patients, the latter of which are assumed to be self-determining agents (Pouncey & Lukens, 2010). In a forensic mental health hospital, this principle must be balanced with a myriad of opposing priorities that relate to safety and security, as well as the fact that some patients are unable to realize their own treatment needs. Forensic mental health services predominantly have been professionally driven and organized according to principles of a bio-medical paradigm, which often is oriented toward impairment, biological reductionism, and physician-led treatment decision-making (Borrell-Carrio, Suchman, & Epstein, 2004; Ghaemi, 2006; Pouncey & Lukens, 2010). For the aforementioned reasons, patient-centered care historically has not been prioritized in forensic mental health services. This is problematic, since there is growing recognition that participatory and collaborative approaches to patient care are important for achieving high-quality and effective mental health services. While the context and nature of service provision in a forensic mental health hospital poses problems for delivering patient-centered care, it simultaneously underscores its importance.

In the forensic mental health system, the notion of patient-centered care (and its related constructs) faces several paradoxes, some of which are considered here. For instance, patients in the forensic mental health system (e.g. those found ‘Not Criminally Responsible on Account of Mental Disorder’ (NCRMD) in Canada) have often been excused from full legal culpability because their mental illness created a situation in which they committed a crime for which they were not full agents of self-determination. Such people in many societies, including Canada, are treated with greater compassion and humanity by diverting them toward treatment and away from traditional criminal justice processing which includes punitive sentences. This presents a quagmire for philosophies of care that prioritize self-determination and rational choice: as is articulated by Pouncey and Lukens (2010): ‘how can we advocate treating persons with mental illness as full moral agents for the purpose of providing social goods, while simultaneously treating them as compromised moral agents when the same illness earns them social sanction?’ (p. 102). Additionally, individuals are typically under the jurisdiction of a forensic mental health system because an untreated or ineffectively treated mental illness led to the commission of a criminal act. Combined with the prevalence of poor illness insight, this suggests that for some people in the forensic mental health system, the patient-centered approach of emphasizing self-directed treatment is neither appropriate nor desirable.

Another paradox that arises with the notion of providing patient-centered care in a secure forensic mental health hospital involves the fact that inpatients are detained and, in some cases, are given treatment against their will. Feelings of powerlessness and oppression are part of the subjective experience of receiving forensic mental health services (Livingston & Rossiter, 2011; Livingston, Rossiter, & Verdun-Jones, 2011; Mezey, Kavuma, Turton, Demetriou, & Wright, 2010). Indeed, the concept of patient-centered care does not typically invoke images of court-ordered treatment, locked gates, highly secure treatment wards, and security guards. One might question whether power-sharing, collaboration, and non-coercive practices can be realistically encouraged in such a context (Green et al., 2011).

Lastly, the principles of social inclusion and social justice underpinning the patient-centered care approach do not apply neatly to the forensic mental health system. There is certain unwillingness for a segment of society to accept that forensic mental health patients, especially those who have committed heinous crimes, should be given an opportunity to reintegrate back into society. Public intolerance and social stigma create a challenging environment for prioritizing approaches that promote equality, inclusivity, and full citizenship for people who receive forensic mental health services.

A growing body of literature is beginning to grapple with these philosophical and pragmatic issues (Adshead & Sarkar, 2005; Coffey, 2006; Green et al., 2011; Gudjonsson, Webster, & Green, 2010; Hillbrand, Hawkins, Howe, & Stayner, 2006; Hillbrand & Young, 2008; Hillbrand, Young, & Griffith, 2010; MacInnes, Beer, Keeble, Rees, & Reid, 2010; Mezey et al., 2010; Pouncey & Lukens, 2010; Simpson & Penney, 2011; Spiers, Harney, & Chilvers, 2005). The current study contributes to this body of knowledge by shedding light on the perceptions of patients and providers towards patient-centered care in a forensic mental health hospital.

Method

The study protocols were approved by the research ethics committees of the University of British Columbia and the BC Forensic Psychiatric Services Commission. All participants provided informed consent.

Setting

The setting for this study is a forensic mental health hospital (hereinafter called ‘forensic hospital') that serves British Columbia, Canada. This province is situated on the West Coast of Canada and has a population of approximately 4.4 million. The 190-bed hospital provides tertiary-level mental health assessment, treatment and support to adults with severe mental illness who are adjudicated by a court to be ‘Unfit to Stand Trial’ or ‘NCRMD'. Individuals who are adjudicated NCRMD in British Columbia are typically male (85%), White/Caucasian (80%), and are on average 36 years of age (Livingston, Wilson, Tien, & Bond, 2003). The hospital contains nine units that vary from high, medium, and low security levels. Patients detained in the forensic mental health hospital have been deemed by a Criminal Code Review Board to be a significant threat to public safety and are, therefore, inappropriate for being managed in the community.

In 2009, the BC Forensic Psychiatric Services Commission consulted with patients, staff, and service partners to develop strategic directions that would articulate a blueprint for moving forward over the next five years. The strategic initiatives included engaging families and clients as partners in the design, delivery, and evaluation of services, and moving toward clinical service models that are firmly patient-centered. This offered a window of opportunity to examine the concept of patient-centered care from the perspectives of patients and providers at a forensic hospital.

Participants

Patients and providers were recruited through flyers posted throughout the forensic hospital and through presentations at patient and staff meetings. Patients were eligible to participate if they met the following criteria: (a) at least 19 years of age, (b) could speak and understand English, (c) were receiving treatment services for at least one month at the forensic hospital, and (d) were approved by their psychiatrist to participate in the study (e.g. cognitively capable, not a current risk to staff). Thirty patients participated in the study. Most were men (n = 24, 80%), White (n = 26, 87%), and born in Canada (n = 27, 90%). The average age was 40 years (n = 30, SD = 11.1). More than half had completed high school or equivalent (n = 17, 57%). The patients self-reported the following psychiatric diagnoses: schizophrenia (n = 18, 60%), schizoaffective disorder (n = 5, 17%), bipolar disorder (n = 3, 10%), other psychotic disorders (n = 2, 7%), and unknown (n = 2, 7%). A history of comorbid substance use problems was reported by 53% (n = 16) of participants. On average, patients had three prior admissions to the forensic hospital (n = 30, SD = 3.6, Mdn = 2, Max = 15). Median length of stay for the current admission was 23 months (n = 30, M = 46.1, SD = 54.8).

Providers were eligible to participate if they had worked in a clinical or therapeutic role at the forensic hospital for at least six months. Twenty-eight providers completed a questionnaire. Half were men (n = 14, 50%), most were White (n = 24, 89%), and 78% (n = 21) were born in Canada. The average age was 43 years (n = 28, SD = 10.8). Professional roles included nursing and other direct care (i.e. healthcare worker) (n = 13, 46%), psychosocial rehabilitation (e.g. occupational therapy, counseling) (n = 9, 32%), psychiatry and psychology (n = 3, 11%), and social work (n = 3, 11%). On average, providers had been working at the forensic hospital for 10 years (n = 27, SD = 7.7) and reported working an average of 22 hours (n = 28, SD = 12.2) per week of direct patient contact.

Materials and procedure

Patient interviews

Semi-structured interviews were conducted in private by a graduate-level research assistant at the forensic hospital. Interviews lasted approximately 60 minutes. Patients were compensated $10 for their participation. The interview guide contained several self-report measures (described below) that were selected for their ability to assess a range of constructs related to patient-centered care.

Recovery-oriented care. The degree to which services were aligned with the principles of the recovery model was measured using the ‘person in recovery’ version of the Recovery Self Assessment Scale (RSA) (O'Connell, Tondora, Croog, Evans, & Davidson, 2005). The RSA has 32 items and five domains: life goals, service involvement, diversity of treatment options, choice, and individually tailored services. For the current study, internal consistency (Cronbach's α) for the RSA ‘person in recovery’ version was .88.

Therapeutic milieu. Hospital milieu was assessed with the Essen Climate Evaluation Schema (EssenCES), which was designed for use in forensic mental health inpatient settings (Schalast, Redies, Collins, Stacey, & Howells, 2008). The EssenCES is a 15-item scale with three domains: patients’ cohesion, experienced safety, and therapeutic hold. Internal consistency (Cronbach's α) for the EssenCES was .77.

Personal recovery. The personal recovery process was assessed using the Mental Health Recovery Measure (MHRM) (Bullock, 2005). The MHRM contains 30 items and seven subscales that reflect different stages of recovery: overcoming stuckness, self-empowerment, learning and self-redefinition, basic functioning, overall well-being, new potentials, and advocacy/ enrichment. Internal consistency (Cronbach's α) for the MHRM was .86.

Engagement. Patient engagement in mental health services was measured using the Singh O'Brien Level of Engagement Scale (SOLES) (O'Brien, White, Fahmy, & Singh, 2009). The SOLES contains 16 items and assesses two domains: acceptance of need for treatment and perceived benefit of treatment. Internal consistency (Cronbach's α) of the SOLES was .91.

Empowerment. Personal empowerment was assessed using the Making Decisions Empowerment Scale (MDES) (Rogers, Chamberlin, Ellison, & Crean, 1997). The 28-item instrument contains five domains: self-esteem/self-efficacy, power-powerlessness, community activism and autonomy, optimism and control over the future, and righteous anger. Internal consistency (Cronbach's α) of the MDES was .72.

Internalized stigma. Subjective, internalized experiences of stigma were assessed using the Internalized Stigma of Mental Illness scale (ISMI) (Ritsher, Otilingam, & Grajales, 2003). The 29-item measure covers five domains: alienation, stereotype endorsement, discrimination experiences, social withdrawal, and stigma resistance. In the present study, internal consistency (Cronbach's α) of the ISMI was .91.

Provider survey

Providers participated in the study by completing an anonymous paper-based or online questionnaire. Survey participants were remunerated by being entered into a random draw for books on recovery and patient-centered care. The questionnaire contained the RSA ‘provider version’ and the EssenCES. Among the providers, internal consistency (Cronbach's α) for the RSA ‘provider version’ and EssenCES was .89 and .78, respectively.

Analysis

Descriptive analysis on all variables was performed using SPSS version 14.0. To facilitate appraisal and interpretation of the RSA and EssenCES, we considered mean scores that fell one standard deviation below the mid-point of the scale (RSA = 3, EssenCES = 10) to be a perceived gap in services. Mean scores that were one standard above the scale mid-point were considered to be perceived strengths. Pearson product-moment correlation coefficients (2-tailed) were calculated to investigate associations between the different patient-centered care constructs.

Results

Recovery-oriented care

Table 1 provides a summary of the RSA subscales and total mean scores for patients and providers. All of the RSA subscales were within one standard deviation from the mid-point of the scale. Although this suggested that the level of recovery-oriented services at the forensic hospital was perceived as satisfactory by patients and providers, it also indicated an absence of perceived strengths or gaps in services.

Table 1.

Patient and provider ratings of recovery-oriented care.

	Patients			Providers
RSA Scales (possible range)	n	M	SD	n	M	SD	t	df	P
Life goals (1–5)	29	3.30	.72	27	3.22	.60	.48	54	.635
Involvement (1–5)	29	2.77	.93	26	2.44	.75	1.44	53	.157
Treatment options (1–5)	29	2.76	.85	24	3.22	.70	−2.15	51	.036
Choice (1–5)	29	3.11	.77	25	2.94	.81	.79	52	.432
Individualized services (1–5)	29	2.95	.89	26	3.28	.73	−1.48	53	.145
Inviting (1–5)	30	3.28	1.06	27	3.09	.91	.73	55	.471
Total (1–5)	29	3.06	.61	27	3.03	.60	.21	54	.831

T-tests were performed to compare the RSA subscale and summary scores between patients and providers. Patients had significantly lower ratings than providers on the treatment options subscale, t(51) = − 2.15, p < .05, indicating they perceived that fewer treatment options were available at the hospital. More specifically, patients’ ratings were lower in relation to their opportunity to discuss spiritual needs and interests, t(52) = − 2.40, p < .05, and the extent to which hospital discharge criteria had been discussed with them, t(54) = − 3.19, p < .01. Differences in patients’ and providers’ mean ratings across the other RSA subscales and total score were not statistically significant.

Several other significant differences existed in how patients and providers rated the individual RSA items. For example, providers had more positive ratings on items such as: patients are asked about their interests, t(54) = − 2.06, p < .05, and patients are encouraged to take risks and try new things t(53) = − 2.30, p < .05. In contrast, the ratings of patients were more positive on the following items: hospital staff believe patients can recover, t(55) = 2.80, p < .01; hope and high expectations are encouraged in patients, t(55) = 2.11, p < .05; hospital staff believe that patients can self-manage their symptoms, t(55) = 4.38, p < .001; and hospital staff believe patients can make their own life choices t(55) = 2.06, p < .05.

Average patient ratings of all individual RSA items were within one standard deviation of the mid-point of the scale, indicating no perceived strengths or gaps regarding recovery-oriented care. For providers, average ratings of two RSA items signaled perceived strengths, including the extent to which hospital discharge criteria are discussed with patients (n = 27, M = 3.96, SD = .76) and the diversity among hospital staff (n = 27, M = 4.63, SD = .57). Providers’ ratings also revealed a perceived gap that focused on the degree to which persons in recovery are involved in hospital staff training (n = 28, M = 1.61, SD = .96).

Table 2 summarizes the RSA items rated highest and lowest by patients and providers in relation to hospital services. Both groups provided high ratings of perceived diversity among staff (e.g. ethnicity, culture, interests). Patients and providers both rated several common items as low, including lack of involvement of patients in hospital staff training, lack of opportunity for patients to discuss sexual needs and interests with hospital staff, and an absence of patient role models or mentors in the hospital. Consistent with the analysis presented above, there was a striking difference between patient and provider perspectives regarding the extent to which hospital staff endorsed and supported symptom self-management strategies, with patients expressing more positive views than providers.

Table 2.

Strengths and gaps of recovery-oriented care.

	Patient perspectives	Provider perspectives
Strengths	Staff believe in patients’ recovery (n = 29, M = 3.93, SD = 1.10) Staff are diverse (n = 29, M = 3.86, SD = 1.33)* Staff encourage hope and recovery (n − 30, M = 3.70, SD = 1.12) Staff believe in symptom self-management (n − 29, M = 3.66, SD = 1.11)† Staff are welcoming (n = 30, M = 3.63, SD = 1.19)	Staff are diverse (n − 27, M = 4.63, SD = .57)* Discharge criteria are discussed (n = 27, M = 3.96, SD = .76) Spiritual needs are discussed (n = 26, M = 3.85, SD = .97) Plans are made for life goals (n = 28, M = 3.61, SD = .99) Staff ask patients about their interests (n = 27, M = 3.59, SD = 1.31)
Gaps	Staff encourage positive risk-taking (n = 29, M = 2.14, SD = 1.13) Patient role models/mentors are used (n = 29, M = 2.31, SD = 1.31)* Staff help patients give back to the community (n − 29, M = 2.45, SD = 1.40) Sexual needs/interests discussed (n = 28, M = 2.46, SD = 1.26)* Patients are involved in staff training (n = 29, M = 2.48, SD = 1.43)*	Patients are involved in staff training (n = 28, M = 1.61, SD = .96)* Patients can access their records (n = 24, M = 2.42, SD = 1.06) Staff believe in symptom self-management (n = 28, M = 2.43, SD = 1.00)† Sexual needs/interests discussed (n = 25, M = 2.44, SD = 1.12)* Patient role models/mentors are used (n = 25, M = 2.44, SD = 1.00)*

Notes: *Indicates commonalities between patients and staff; †Indicates differences between patients and staff.

Therapeutic milieu

Table 3 summarizes the average scores on the EssenCES subscales. Most of the average subscale scores were within one standard deviation of the midpoint, suggesting a satisfactory therapeutic milieu within the forensic hospital. Providers’ ratings of experienced safety were an exception, which indicates that safety is a perceived gap in relation to providers’ impressions of the therapeutic milieu within the hospital.

Table 3.

Patient and provider ratings of therapeutic milieu.

	Patients			Providers
EssenCES scales (possible range)	n	M	SD	n	M	SD	t	df	P
Patient cohesion (0–20)	29	9.86	3.85	28	8.32	4.03	1.47	55	.146
Experienced safety (0–20)	28	9.11	3.21	28	4.18	3.33	5.63	54	.000
Therapeutic hold (0–20)	29	10.86	4.03	28	11.46	4.19	−.55	55	.583
Total average (0–4)	29	2.01	.49	28	1.61	.51	2.99	55	.004

In comparison to patients, providers had significantly lower ratings on the experienced safety subscale, indicating that they perceived greater tension and threat of aggression or violence in the hospital. Compared with providers, patients had more positive perspectives toward the hospital's overall social climate (EssenCES total). On individual EssenCES items, providers were significantly more likely than patients to perceive a greater potential for threatening situations, t(55) = 4.35, p < .001; greater levels of aggressiveness among patients, t(54) = 3.50, p = .001; more fear among patients of other patients, t(55) = 5.29, p < .001; and greater fear among staff of patients, t(54) = 4.30, p < .001. Lastly, patients were more likely than providers to perceive positive peer support among patients in the hospital, t(55) = 2.12, p < .05.

Correlations between patient-centered measures

Bivariate correlations were calculated between total scores on measures related to patient-centered care, including patient ratings of engagement (n = 30, M = 6.74, SD = 1.88), empowerment (n = 30, M = 2.86, SD = .24), personal recovery (n = 29, M = 85.07, SD = 12.75), internalized stigma (n = 28, M = 2.07, SD = .39), and the level of recovery-oriented services at the forensic hospital (n = 30, M = 6.74, SD = 1.88) (see Table 4).

Table 4.

Correlations between patient-rated measures.

	RSA	EssenCES	MHRM	MDES	SOLES	ISMI
a. Recovery-oriented care (RSA)	1
b. Therapeutic milieu (EssenCES)	.66**	1
c. Personal recovery (MHRM)	.28	.12	1
d. Empowerment (MDES)	.25	.00	.60*	1
e. Engagement (SOLES)	.55*	.36	.06	.05	1
f. Internalized stigma (ISMI)	−.18	.05	−.79**	−.61*	−.05	1

Notes: *Significant at a .01 level (2-tailed); **Significant at a .001 level (2-tailed).

Patients’ self-ratings of their personal recovery (MHRM) were correlated with personal empowerment (MDES) (r = .60, p < .01) and internalized stigma (ISMI) (r = – .79, p < .01). As such, higher levels of personal recovery among patients were associated with greater empowerment and less internalized stigma. Patient ratings of recovery-oriented care (RSA) were moderately correlated with mean ratings of service engagement (SOLES) (r = .55, p < .01) and therapeutic milieu (EssenCES) (r = .66, p < .01). Therefore, patients who perceived hospital services as being consistent with recovery principles were more likely to be engaged with services and had more positive views toward the therapeutic milieu of the hospital. Similarly, provider ratings of recovery-oriented care and therapeutic milieu were also significantly correlated (r = .54, p < .01).

As discussed previously, concerns about safety are significant and unique features of forensic hospitals because of distinctive characteristics of their patient population. As such, we further examined the relationship between perceptions of safety (EssenCES experienced safety) and recovery-oriented care (RSA). Patients who felt more safe in the hospital were more inclined to believe that their choices were valued and respected by staff (RSA choice) (r = .52, p < .01). Providers who had more positive views regarding hospital safety were more likely to perceive services as being welcoming toward the patients (RSA inviting) (r = .41, p < .05). There were no other significant correlations between perceptions of safety and recovery-oriented care.

Patients’ overall ratings of recovery-oriented care (RSA) were not significantly correlated with personal recovery (MHRM). To explore this further, bivariate correlations were performed on the RSA and MHRM subscales, with several significant relationships emerging from the analyses. Patients who perceived that hospital services had assisted in their development and pursuit of life goals (RSA life goals) demonstrated greater achievement of basic functioning (e.g. self-care, being active, connecting with others) (MHRM basic functioning) (r = .39, p < .05) and self-empowerment (MHRM advocacy/enrichment) (r = .43, p < .05). Patients who felt more involved in the development and provision of hospital services (RSA involvement) also felt a greater sense of self-empowerment (MHRM advocacy/enrichment) (r = .48, p < .01). Patients with greater belief that their choices were valued and respected by hospital staff (RSA choice) demonstrated higher levels of quality of life, including greater overall sense of well-being (MHRM well-being) (r = .39, p < .05) and greater desire to reach new potentials of higher functioning (MHRM new potentials) (r = .37, p = .05). Finally, patients with higher ratings about the degree to which hospital services were individually tailored (RSA individualized) also demonstrated better achievement of basic functioning (MHRM basic functioning) (r = .43, p < .05) and personal empowerment (MHRM advocacy/enrichment) (r = .37, p < .05). However, personal recovery was not significantly associated with patients’ perceptions of the welcoming nature or diversity of services at the forensic hospital.

Discussion

The present study used several complementary measures to examine how patient-centered care was perceived by patients and providers in a forensic mental health hospital. Of particular importance was identifying perceived strengths and gaps in hospital services as they related to patient-centered care. Findings indicated that patients and providers viewed the services of the forensic hospital as meeting a satisfactory level of recovery-oriented care. This suggests that, despite their inherent paradoxes, the forensic mental health and recovery paradigms are not incompatible. Although patients and providers shared similar overall views regarding the recovery-orientation of hospital services, patients had more positive ratings concerning how they were viewed by hospital staff (e.g. hopeful for their recovery, supportive of their autonomy).

Patients and providers also had similar perceptions about the hospital's therapeutic milieu. Generally, both groups held positive views about the social climate of the forensic hospital (e.g. supports therapeutic needs, promotes mutual support). A key difference was that providers were more likely to perceive the hospital as being a potentially unsafe environment. Such perceptions by providers have important implications for patient-centered care, as hospital staff may feel more comfortable with practices aimed at containing risk (e.g. seclusion, restraint) rather than engaging and collaborating with patients. This finding is consistent with other research focused on the professional tensions that exist while balancing care and custody in forensic hospitals (Hinsby & Baker, 2004; Mason, 2002; Meehan, McIntosh, & Bergen, 2006). For example, findings of a qualitative study by Hinsby and Baker (2004) suggested that forensic mental health providers tend to give greater weight to maintaining safety and security, as opposed to being caring, as they carry out their clinical duties. Research has also indicated that feelings of fear and abjection toward people receiving forensic mental health services may also encourage the use of restrictive interventions in a forensic mental health hospital (Jacob, Gagnon, & Holmes, 2009). As such, moving toward a patient-centered model of care in a forensic hospital likely requires focused attention to address perceptions of fear and safety among providers. It is likely that many of the aforementioned challenges with incorporating patient-centered, recovery-oriented philosophies of care into a forensic hospital also apply to non-forensic tertiary psychiatric hospitals. The degree to which the ‘forensic’ nature of service provision exacerbates these challenges is an empirical question that requires further study.

Finding that a higher level of personal recovery among patients was related to greater empowerment and lower internalized stigma is consistent with other research involving non-forensic mental health samples (Lysaker, Buck, Taylor, & Roe, 2008; Markowitz, 2001; Ritsher et al., 2003; Yanos, Roe, Markus, & Lysaker, 2008). An unexpected finding was that personal recovery, empowerment, and internalized stigma were unrelated to patients’ overall perceptions of recovery-oriented care at the forensic hospital. This is particularly surprising given the suggestion from the literature that aligning mental health services with recovery-oriented and patient-centered principles is a recommended strategy for reducing stigma (Corrigan, Roe, & Tsang, 2011; Heijnders & Van Der Meij, 2006). The current study did, however, reveal significant associations between several domains of recovery-oriented care and personal recovery, which is consistent with the current literature suggesting that the two are related (Slade, 2009a, 2009b). Furthermore, the current study found that patients’ overall perceptions of recovery-oriented care were linked to greater service engagement and positive ratings of therapeutic milieu. The important implication of such findings is that enhancing recovery-oriented care in a forensic mental health hospital may improve patient engagement in (and adherence to) services. Further research is needed to explore the association between patient-centered and recovery-oriented care with clinical and criminal justice outcomes for people who receive forensic mental health services.

This study has several methodological limitations that should be noted. We used several strategies to encourage participation from a diversity of patients and staff, which included attending numerous meetings to discuss the study and the benefits of participating, providing remuneration to those who participated, establishing broad inclusion criteria, and being flexible around the scheduling of interviews. Despite this, the sample sizes for patients and providers were small, which may place restrictions on the external validity of the study findings. Second, it is possible that the study recruited a subgroup of patients and staff that possessed characteristics germane to the study (e.g. empowered, engaged). As well, there was an under-representation of patients from ethnically diverse backgrounds. Such limitations cast some doubt as to whether these findings are representative of the perspectives of patients and providers in the forensic hospital. As a consequence, scores on the different measures (e.g. recovery, empowerment, and self-stigma) may be skewed in a positive direction. Lastly, the study may also be limited by the subjective, self-report nature of the data collection, as we did not seek participant consent to confirm independently any information (e.g. psychiatric diagnosis) that they provided. Although our study examined the perspectives of patients and providers, future research would be strengthened by also incorporating objective measures of patient-centered care, such as the presence or absence of certain service characteristics, into their study designs.

Conclusion

Forensic mental health hospitals may be viewed as inhospitable environments for patient-centered care; however, our results indicate that the two are not incompatible. Patient-centered and recovery-oriented care is possible to achieve, with some modification, in settings that prioritize risk management. Special attention to safety and risk issues is needed when introducing these collaborative approaches to forensic mental health settings. Evaluating patients’ and providers’ perceptions, and understanding the strengths and gaps in services from their perspectives, is an important way to create a blueprint for moving toward a patient-centered model of care.