Introduction

Improving treatment for individuals experiencing psychosis begins with providing adequate training for direct care providers. Adequate training not only involves improving knowledge and skills, but also addresses the providers’ assumptions and perceptions about individuals who experience psychosis. Cognitive Behavioral Therapy for Psychosis (CBT-p) is one approach that has shown benefits for working with psychosis (Wykes et al. 2008) and is currently recommended for individuals diagnosed with schizophrenia (NICE 2014). While the extent of benefits achieved via CBT-p has been somewhat contested (for current commentary on this debate, see McKenna et al. 2019), service users tend to be very satisfied with this treatment (see Lawlor et al. 2017).

With an interest in increasing access to services for psychosis, frontline providers have begun to receive training in methods and strategies that align with CBT-p principles (see Hazell et al. 2016). These low intensity, CBT-p informed strategies often emphasize engagement strategies as well as straightforward ideas for coping with psychosis embedded within a recovery-oriented treatment philosophy. A focus on collaboration, optimism and shared decision-making is essential in this work (Brabban et al. 2017).

One element of treatment particularly salient for working with psychosis relates to the service provider’s attitudes about recovery. Mental health providers across a variety of settings tend to report a mix of positive and negative views about individuals who have mental illness (Schulze 2007). Wahl and Aroesty-Cohen (2010) extended this initial work and reported that, in general, attitudes of providers tend to be more positive in relation to public views of mental illness. At the same time, other studies reported negative attitudes of mental health professionals toward individuals with psychosis (see Alshahrani 2018). For example, Nordt et al. (2006) reported that the psychiatrists in their study held more negative stereotypes about schizophrenia than the general population. Additionally, Nordt et al. (2006) found that mental health providers (nurses, psychiatrists and other mental health professionals) and the general public endorsed similar levels of desire for social distance from individuals with psychotic disorders. This is especially important, as the desire for social distance from individuals experiencing psychosis is often much higher than is the desire for social distance from those with other mental health problems, such as depression (Link et al. 1999; Nordt et al. 2006).

McLeod et al. (2002) reported that many mental health workers prefer not to work with individuals diagnosed with severe mental illness. Mirabi et al. (1985), found mental health workers (psychiatrists, psychologists, social workers, psychiatric nurses, caseworkers) felt little satisfaction in working with this group. Prytys et al. (2011) in a qualitative study, reported that care coordinators (nurses, social workers, and occupational therapists) often held pessimistic attitudes about individuals with schizophrenia as “chronically ill” and requiring “ongoing, long-term support” (p. 54). And when Sullivan et al. (2015) asked nurses and physicians working in VA settings to evaluate a clinical vignette of a person with medical issues, they found that when the diagnosis of schizophrenia was added to the vignette, it adversely affected the providers’ rating of the person’s adherence, ability to use education materials, and competence to make treatment decisions, even when the description included several indicators of stable functioning.

Van Dorn et al. (2005) assessed the attitudes of mental health providers (majority of whom were psychiatrists and clinical social workers, however included other providers) who work with individuals with schizophrenia using a case vignette. Approximately one third of clinicians in this sample responded to the vignette about schizophrenia with concerns about violence (30% indicated somewhat or very likely) and a preference for social distance (36% indicating very or somewhat likely). Magliano et al. (2017a, b) reported that the perceptions of dangerousness in a sample of general practitioners was correlated with other beliefs about discriminatory treatment of those affected by schizophrenia, need for social distance, and beliefs about the need for medication for life.

Attitudes about schizophrenia are also likely to have an influence on selection of treatment goals, strategies, and outcomes. Foundational to CBT-p is the idea that it is vital to develop a positive alliance with the individual and to pursue personally meaningful goals. For example, it is not uncommon in CBT-p to encourage clients to pursue goals such as long-term relationships and to contribute meaningfully to society. This can be challenging when some mental health providers favor social restrictions for those with schizophrenia. For example, Magliano et al. (2004) asked nurses and psychiatrists to respond to a variety of prompts about social restrictions for individuals with schizophrenia. They reported that 63% of nurses and 43% of psychiatrists agreed (i.e., answered completely or partially true) people with this disorder “should not get married”; and 34% of nurses and 17% of psychiatrists agreed they “should not vote.” (p. 325).

Beliefs about the responsiveness to treatment of individuals with schizophrenia can dictate how interventions are delivered. Manuel et al. (2013) examined the attitudes of Assertive Community Treatment (ACT) teams in relation to level of intrusiveness of interventions. In this review, the authors found that staff who held stigmatizing beliefs were likely to use more intrusive interventions such as conditional involvement and reporting clients to authorities. In a separate study, Lecomte et al. (2018) found that negative attitudes of clinicians included beliefs that people with psychosis have no ability to develop insight and are too ill to be able to benefit from CBT-p treatment.

Inflexible beliefs about treatment approach may also reflect underlying attitudes regarding individuals who experience psychosis. For example, providers (e.g., general practitioners; psychiatrists; psychosocial staff; and nurses) who reported greater certainty about the necessity of long-term medications, also more firmly believed that people with schizophrenia are dangerous and not able to develop trusting relationships with staff (see Magliano et al. 2017a, b). In this same work, providers who reported greater optimism about recovery gave more favorable views about the usefulness of psychosocial interventions. Relating attitudes to outcomes, O’Connell and Stein (2011) demonstrated that clients of case managers who held more optimistic expectations about the internal resources of individuals with schizophrenia were employed significantly more days compared with clients of case managers who held lower expectations about the personal resources or efficacy of individuals with schizophrenia.

In summary, the above studies show that negative attitudes about working with people who experience psychosis are commonly held by mental health providers and can influence treatment. Negative attitudes/expectations (especially the perception of threat) may inhibit capacity to develop empathy (see Gilbert 2014) and undermine collaboration which are both essential to developing a strong therapeutic relationship when working with people who experience psychosis (Evans-Jones et al. 2009). When negative attitudes are present, they are likely to exert their greatest effect on the therapeutic alliance. In this regard, it is widely acknowledged that therapeutic alliance is perhaps the factor most associated with positive outcomes (Ardito and Rabellino 2011; Laska et al. 2014; McCabe and Priebe 2004). In relation to schizophrenia, Svensson and Hansson (1999) reported that a stronger initial therapeutic alliance between patient and therapist resulted in better treatment outcomes. Similarly, Goldsmith et al. (2015) reported that therapeutic alliance in the treatment of early psychosis was associated with better outcomes. In addition, they found that when therapeutic alliance was poor, continuing treatment had detrimental effects.

CBT-p Training to Influence Staff Attitudes About Working with Psychosis

Based on the above findings, it seems reasonable to propose that training for staff who work with people with schizophrenia would benefit from offering targeted strategies to address staff perception and attitudes in addition to providing skill-based training with a significant focus on improving working alliance. It follows that the use of training strategies that lead to greater acceptance and understanding of schizophrenia are desirable (Wahl and Aroesty-Cohen 2010). Effective training would also increase a sense of confidence and efficacy when working with individuals who experience psychosis.

There is evidence that training in CBT-p may help staff develop more positive attitudes about working with psychosis. For example, Berry et al. (2009), in an uncontrolled pilot investigation, taught staff members (i.e., mental health nurses and support workers) how to think about client problem behaviors using a formulation approach (i.e., using a CBT model). They found that using a CBT-formulation helped staff perceptions in the following ways: feeling more empowered in their interventions, reducing blame and increasing optimism in treatment, increasing knowledge of service user problems and more positive feelings toward service users.

Other studies providing intensive training in CBT-p have also led to improvements in key provider attitudes. For example, trainees who participated in 3 days of CBT-p training reported improved levels of empathy for the experience of psychotic symptoms following training (Mcleod et al. 2002). Staff who were provided Recovery Oriented Cognitive Therapy training in an inpatient psychiatric unit reported improvements in trainees’ perception of treatment and therapeutic milieu (Chang et al. 2014). Interestingly, staff attitudes about working with schizophrenia did not change pre to post-training at the 6-month follow-up.

The BeST Center developed a comprehensive technical assistance package to guide the implementation of CBT-p and CBT-p informed skills in practice settings. One aspect of the implementation was a 2-day intensive training based on the CBT-p model of work with psychosis which provides an education in recovery concepts, strategies for engaging clients with psychosis, methods for understanding symptoms, and strategies for coping with psychosis. Because training included providers with sometimes vastly different roles (e.g. therapists, case managers, nurses, psychiatrists), the intensive training emphasized foundational skills for working with psychosis informed by CBT-p. Following the intensive training, providers received additional training and consultation specific to their scope of practice. The 2-day intensive training sessions involved a mix of didactics, videos of individuals with lived experience of schizophrenia who are in recovery, and a variety of experiential empathy exercises. These exercises were intended to help mental health providers get a better sense of what individuals with psychosis experience. Empathy activities and behavioral rehearsals provided trainees with a type of exposure to the experience of psychosis to allow for desensitization as well as experiential practice.

Assessing Participants’ Attitudes Before and After Intensive Training

As part of a larger continuous quality improvement approach to implementing a range of CBT-p services and technical assistance, trainees completed the Psychosis Attitude Survey (PAS) before and after the 2-day, intensive training. The intent was to help assess if participating in the intensive training demonstrated changes in attitudes, levels of empathy, and confidence about working with people who experience psychosis. Participants were instructed to fill out the PAS using an anonymized number and to use the same number on the pre- and post-training surveys to match participants’ pre- and post-training responses. Aggregated data from each training group were reviewed to determine response to the training and to guide follow-up consultation. The procedures were reviewed by the IRB of Northeast Ohio Medical University and were considered consistent with a quality improvement/program evaluation project.

Methods

The Psychosis Attitude Survey (PAS) is a 19-item survey in which trainees are asked to respond to statements (e.g., “I want to work with persons with psychosis”) on a 7-point Likert scale from “strongly disagree” to “strongly agree” with neutral in the middle. Items were written from both positive and negative directions and recoded for analysis. The PAS was developed based on the measure used by McLeod et al. (2002; adapted from a measure by Cartwright 1980; and Gallagher et al. 1991); modified with items from a survey of CBT-p experts by Morrison and Barratt (2010); as well as items developed based upon common attitudes expressed by case managers (See Tables 1, 2).

Table 1 Psychosis Attitude Survey (PAS) Scores pre and post training
Full size table
Table 2 Percentage of trainees endorsing PAS consistent with positive working attitude
Full size table

In addition to the overall total (minimum 19, maximum 133) and individual items, the authors identified three subscales within the PAS representing major content areas relevant to working with psychosis. Empathy representing belief in one’s ability to understand a person’s experience of psychosis (items 2R, 3, 7R—min 3, max 21). Adequacy/Confidence representing belief in one’s ability to work with psychosis (items 5R, 6, 11, 12, 15—min 5, max 35). And Optimism representing hopeful attitudes about an individual’s ability to recover; possessing strengths on which to build further skills (items 1R, 17R—min 2, max 14).

Participants

The PAS was completed by participants in BeST Center CBT-p Intensive Training sessions conducted between 2016 and 2018. Sixty-nine trainees completed at least one PAS. Due to inability to match forms (either no anonymized number was used or no corresponding anonymized number existed between pre- and post-training surveys), a total of fifty-three trainees completed both pre and post training surveys and were included in the analysis.

The information was collected as part of a program improvement project and not designed to test hypotheses. Therefore, specific characteristics of the training group were not collected. Individuals who participated in the training represented a cross-section of many community mental health centers. A non-random, convenience sample of counselors, case managers, employment specialists and to a lesser extent nurses and prescribers participated in training.

Results

Our primary intent was to determine if PAS scores shifted in a direction associated with more favorable attitudes about working with psychosis following CBT-p training. Wilcoxon Signed Rank test was used to assess differences between pre- and post-training PAS responses given the ordinal nature and non-normal distribution of the responses. Fifteen of the nineteen items showed a statistically significant shift in participants’ attitudes consistent with a favorable increase in attitude toward working with psychosis. Similarly, the PAS Total Score (mean rank 7.5 versus 22.07, Z =  − 5.29, p < 0.0001), as well as the Empathy (mean rank 13.96 versus 22.06, Z =  − 2.96, p = 0.003), Adequacy/Confidence (mean rank 9.63 versus 22.75, Z =  − 5.17, p < 0.0001), and Optimism (mean rank 7.50 versus 17.43, Z =  − 4.58, p < 0.0001) subscale scores showed statistically significant shifts. The greatest shifts in response were observed for items targeted in the intensive training: I have a clear idea of how to interact with people who have psychosis; and I have adequate training (see Table 1).

Consistent with methods employed by both McLeod et al. (2002) and Gallagher et al. (1991), the authors interpreted the Total PAS and the subscale (Empathy, Adequacy/Confidence, and Optimism) scores higher than “neutral” (mean greater than 5.5) as a Positive Working Attitude. Using this method, approximately 40% of trainees fell in the positive working attitude range prior to training and approximately 71% of trainees’ scores fell in this range following training based on PAS Total. We conducted the same breakdown by the subscales with the largest shift seen in the Adequacy/Confidence subscale (see Table 2).

Overall, the training appeared to influence the participants’ perceptions about working with psychosis. Specifically, participants’ PAS scores shifted in a desired, positive direction.

Discussion

Service providers working with people with schizophrenia often feel underprepared and ill-equipped to provide psychosocial interventions. Often these providers come to training with a wide range of attitudes, interests, and experiences in working with psychosis. Attitudes about working with individuals experiencing psychotic symptoms appeared to have been positively influenced in this sample. Consistent with the objectives of the training, the survey results suggest that participants in these trainings reported feeling more confident in their ability to work with psychosis and endorsed more optimistic attitudes following training.

Participants were individuals already working in community mental health centers with people who experience psychosis. As reported above, 15 of the 19 items showed improvement in the desired direction. The mean rating for all of the PAS item pre-training scores fell at or above the neutral score (4); with 10 items’ mean pre-training of 5.5 or greater (maximum of 7). It may be that individuals who participate in an intensive CBT-p training start with higher baseline levels to begin with. It may also follow, however, that participants who are dedicated to helping individuals who experience psychosis may also be pre-disposed to increasing their attitudes about working with this population; more open to improving their skills and ability to help in recovery; and more able to increase attitudes and optimism.

The findings also highlighted potential items in need of examination. For example, two items that showed the least movement from pre- to post-training were: It is hard to imagine what it might be like to have psychosis and Relating to persons experiencing psychosis is difficult. It was the goal of the training to increase participants’ empathy for the experiences of the people with whom they work. Lack of significant movement on these items may indicate the training was not able to influence these perceptions. Alternatively, the items may have been unable to detect the way attitudes were affected. While not significant, the items moved slightly in the opposite direction of what was expected. One possible reason was that the training experiences may have resulted in trainees reconsidering their baseline self-assessment (e.g., “I thought I could imagine what it is like however, after this training I realize it is harder for anyone to imagine it unless you experience it yourself”). While this hypothetical response shows improved awareness and empathy, the current items would be unable to detect this type of change.

It is important to note that self-rated attitudes do not necessarily translate into more effective evidence-based care, but they are a starting point. The PAS has been used in aggregate to guide training development and modification. For example, from a program evaluation perspective, trainees in this sample showed relatively low levels of empathy before and even after the 2-day training. While this subscale is not the same as an empathy rating provided by a client, it does measure a provider’s self-assessment of ability to understand psychosis. At an aggregate level, this information was helpful in guiding adjustments to the 2-day intensive training and highlighted important areas for ongoing consultation. In the future, it may be even more helpful to look at individual responses to personalize consultation and follow-up. Finally, more optimistic attitudes about recovery, improved confidence in working with psychosis and improved empathy for the experience of psychosis are all factors that would aid the therapeutic alliance.

One future step may be to refine the PAS and examine the psychometric properties with a larger data set. In this sample, some individuals’ scores did not improve or improved very little. This may be due to high scores on initial survey (ceiling effect) or may be due to other limitations of the measure itself. If a measure like the PAS can be developed to reliably and accurately assess mind-set for working with psychosis, it could be used for multiple purposes (e.g., contribute to learning needs assessment of new staff and guide the selection of training exercises for a particular training cohort).

It is important to note that the data reported here were not collected under controlled conditions and simply reflect training participants’ immediate response to a 2-day intensive training. Long term, post-training data were not collected. It is worth re-stating that changes in attitude ratings do not guarantee subsequent behavior changes. In addition, it is well-documented that attitude and knowledge gains are often not sustained after a training workshop (Sholomskas et al. 2005). In other words, continuous learning processes will likely need to be in place to sustain positive changes. For example, in the Chang et al. (2014) study, positive attitudes did not show a change when measured 6 months post training. It is possible that early positive shifts in attitude may return to baseline without routine review and intentional support of recovery-oriented viewpoints. This is consistent with recommendations for ongoing consultation after training to strengthen and support new skills (e.g., see competence standards provided by the NACBT-p Network 2019). It is recommended that research be conducted in order to identify specific elements of training that support attitude shift and skill development, for specific provider roles, especially when working with individuals diagnosed with schizophrenia. Such information may help to shape future CBT-p training events so that a process of continuous learning may be initiated that serves to support ongoing adoption of recovery-oriented care practices informed by CBT-p.