Perspectives of hospitalized patients with mental disorders and their clinicians on vocational goals, barriers, and steps to overcome barriers.

BACKGROUND
People with mental disorders experience difficulties with finding competitive jobs. In countries with longer psychiatric hospitalization periods, the vocational rehabilitation process can start during hospitalization. Yet, rehabilitation can be hindered by a lack of focus by clinicians on the patients' vocational goals and a lack of agreement between clinicians and patients.


AIMS
To compare (i) vocational goals, (ii) barriers to employment and (iii) support needed to overcome barriers faced by patients.


METHOD
The paired data-set comprised 733 hospitalized patients and their 279 clinicians. Patients selected their vocational goals and clinicians indicated options that seemed realistic. Patients and clinicians indicated how many barriers exist and what support is needed to overcome barriers.


RESULTS
Almost 45% of patients prefer competitive jobs, while 32% of clinicians find this realistic, indicating a moderate relationship between patients' goals and clinicians' perceptions. Patients and clinicians also differ in their perception of the level of barriers and types of support to overcome them. Patients perceive fewer vocational barriers than clinicians and prefer less intense vocational support options.


CONCLUSIONS
Patients and clinicians have different perspectives concerning vocational recovery. Improving vocational rehabilitation requires a stronger alignment between patients and clinicians' vocational goals and barriers.


Introduction
For many people with mental disorders participation in the community and in employment is an important aspect of their recovery process. Although they experience many difficulties with finding competitive jobs (OECD, 2012;Waghorn et al., 2011), reviews of the Individual Placement and Support model (IPS) of Supported Employment provide evidence that competitive employment is realistic (Kinoshita et al., 2013).
Key elements of this evidence-based practice of IPS are the focus on competitive employment, quick job searches, collaboration with mental health services, and working with clients' job preferences (Becker & Drake, 1994). By focussing on these elements, clinicians foster hope, a feeling of control and involve the client as an equal partner. Consequently, it is more likely that goals are actually achieved and prolonged unemployment is decreased (Hogue et al., 2010;Ryan & Deci, 2008). Thus, the involvement of clients in setting vocational goals and deciding on the process of how to achieve these goals must become a primary element of the early stages of rehabilitation (Anthony et al., 2014;Hogue et al., 2010).
Early involvement of people with severe mental illness (SMI) means that vocational rehabilitation (VR) can be initiated during hospitalization. This is especially important in countries where hospitalization rates are higher, hospitalization periods are longer and few community-based services exist.
At a psychiatric hospital, patients are supported by a mental health clinician. They are the patient's contact person, offer support and care and help the patient to develop a recovery plan. Ideally, this also includes assessing vocational goals, defining barriers to employment and pinpointing areas in which support is necessary (Harris et al., 2014). These activities are important because having knowledge of vocational goals and needs may stimulate the vocational progress (Rampton et al., 2010). However, mental health professionals often consider that employment-related questions fall outside the realm of their clinical services (Taskila et al., 2014). Consequently, the VR process often starts after hospitalization. When the patient has vocational concerns, delaying the VR process can have detrimental effects because a timely implementation of the process and a good working alliance with the mental health clinician indirectly affects subsequent relationships with the VR counsellor and employment outcomes (Catty et al., 2010;Tryon & Winograd, 2011). Not only delayed, but also ineffective VR support may contribute to negative experiences, thereby decreasing the likelihood of a positive VR outcome (Marwaha & Johnson, 2005).
Previous studies mainly focused on clients with SMI who receive care through community-based services. Little is therefore known about hospitalized patients. The aim of the article is to study (i) vocational goals, (ii) barriers to employment and (iii) perceived ways to overcome barriers faced by hospitalized patients with mental disorders and their clinicians. The present study also sought to obtain information about potential differences in perspectives between patients and clinicians.

Method
The Belgian context Belgium has the second-highest number of beds per capita in the world, with 179 psychiatric beds per 100 000 population (Samele et al., 2013). Moreover, patients are often hospitalized for periods exceeding 10 weeks (Schoevaerts et al., 2014;Umbach & Vanrillaer, 2014). At the time of the study, the Belgian mental health care context was characterized by an on-going reform process. More community-based services are being implemented and the number of psychiatric beds is decreasing.
Mental health clinicians are rarely trained in traditional or evidence-based VR practices. There exists a segregation of employment and mental health services and there remains a strong focus on Sheltered Employment (Knaeps et al., 2012;Shima et al., 2008).

Participants
Self-developed questionnaires were completed by 733 (62%) of 1160 hospitalized patients, consistent with expectations. Patients were on average 42 years old (SD: 12.13, range 18-64 y), and, according to the clinicians, 39% have multiple mental health problems (often a combination of psychotic symptoms and substance abuse or personality problems). Patients stayed at different types of psychiatric wards (longstay, acute or in-patient day hospital programs) (93%). The remaining patients (7%) receive services of day activity centres (DAC) or Sheltered Housing. These patients were included as they receive intensive care by psychiatric nurses at locations owned by the psychiatric hospitals. Client characteristics are shown in Table 1.

Questionnaire development
The items of the questionnaire were based on items concerning vocational goals, barriers and possible support options published in previous studies (Ali et al., 2011;Secker et al., 2001;Secker & Gelling, 2006). In order to make the items valid for the Belgian context, the questionnaire was updated using a process of continuously requesting feedback from two clinician groups and two patient groups. One clinician group comprised four VR counsellors, the other three vocational therapists and one psychologist. Patients of the patient groups were hospitalized at different types of wards. In total, 14 patients were involved in three patient groups. Each item was checked for comprehensibility, clarity and whether it was stigmatizing. Patients requested the addition of ''work experience program'' (unpaid internships with off-the-job counselling by a mental health clinician) as an alternative vocational goal. They also requested to let their key clinical worker indicate which mental health problems are present instead of asking them for official diagnoses. When asking the patients how many barriers they perceive, patients needed examples of such barriers and three were chosen (i.e. sensitivity for stress, hindering symptoms, and lack of adequate transportation). Lastly, questionnaires were sent to the participating hospitals' ethical boards.

Measures
A first set of multiple-response questions asked patients to indicate their short-and long-term vocational goals, including competitive employment, no activity, Sheltered Employment, voluntary work, education, day activity centre, self-employment, domestic work, other (e.g. pension) and ''work experience program'' (Secker et al., 2001). In a next question, patients indicated how many vocational barriers they perceive on a continuum from ''none'' over ''some'' and ''multiple'' to ''many'' (Secker et al., 2001). A final question explored which kinds of vocational support were perceived as necessary. This is a multiple response item with increasing level of support options, i.e. ''no support'', ''administrative and informative support'', ''support with job search and solicitation trainings'' and ''on-the-job support''. Clinicians received the same list of vocational options, barriers and supports and they indicated which they consider realistic, present or needed respectively for a particular patient.

Data collection
Each psychiatric ward chose a start date on which questionnaires were offered to all hospitalized patients and the key clinician with whom a particular patient had the most contact. No questionnaires were offered to patients younger than 18 or older than 64 years or to those for whom participation was expected by the clinician to result in worsened well-being. Questionnaires were paired; each questionnaire had a unique code so that clinicians could complete the questionnaire with a specific patient in mind. Some clinicians filled in multiple questionnaires depending on how many patients they were offering support. All participants provided written, informed consent. Patients could request help from a mental health clinician who did not fill in a questionnaire for that particular patient.

Data analyses
Because multiple vocational goals could be selected, data are offered as percentages of cases as was done in Secker, Grove, & Seebohm (2001). These percentages indicate how many respondents select a particular vocational option.Binary logistic regression analyses with deviation contrasts were used to assess how patients' correlates (age, gender, time not employed, time hospitalized, mental health problems, and highest attained educational degree) are related to competitive employment interest. Differences between patients' competitive employment goals and clinicians' perspectives were checked using Chi 2tests, kappa with linear weighting and Cramer's V which is an effect size measure for contingency table analyses (Cohen, 1968;Gibbons et al., 2005). Kappa and Cramer's V were interpreted as recommended; 0.20 ¼ poor to slight agreement, 0.21-0.40 ¼ fair, 0.41-0.60 ¼ moderate, 0.61-0.80 ¼ substantial and 0.8-1.0 ¼ almost perfect agreement (Landis & Koch, 1977). To study the trend of disagreement, McNemar-Bowker's test for symmetry (Bowker, 1948) was used as it checks for a significant difference in frequencies below and above the diagonal in the cross table. The item concerning vocational support was a multiple response item.
To compare patients' and clinicians' perspectives, the most intense vocational support option of each person was compared.

Results
In the short term, 35.5% of the patients preferred competitive employment and 21.8% preferred voluntary work (Table 2). Both vocational options were also the most favoured options in the long term by 44.6% and 14.6% of patients, respectively.
According to clinicians, realistic short-term options were day activity centres (34.6%) and domestic work (33.9%). Competitive employment was realistic according to 24.1% (short term) and 32.4% (long term) of clinicians. Some patient characteristics were associated with clinicians' perspectives, 2 (14) ¼ 278.328, p50.001, Nagelkerke R 2 : 51.1%, correctly classified cases: 82.1%. Compared to the overall group, clinicians were more likely to find competitive employment realistic on the long term for patients (i) who have been not working less than two years (OR: 1.594, p: 0.020) or (ii) who have substance (OR: 2.732, p50.001) or emotional (OR: 2.158, p: 0.003) symptoms. It was also considered more realistic when the patient is less than one year hospitalized (OR: 2.868, p50.001). Competitive employment was considered less realistic for patients who have been not working more than two years (OR: 0.346, p50.001), are older (OR: 0.953, p50.001) or who have multiple mental health problems (OR: 0.596, p: 0.013). It was also perceived less realistic for patients holding only a primary school degree (OR: 0.333, p50.001) or hospitalized more than one year (OR: 0.607, p: 0.031) (Supplementary Table 1).
Competitive employment was the option most frequently selected by patients and is also focused on during IPS services. Therefore congruence with clinicians' perspectives was evaluated. There was a moderate relationship between clinicians' perception of competitive employment and patients' competitive employment goals in both the short ( 2 (1,688):128.05, p50.001, weighted Kappa: 0.42, CI: 0.35-0.50, Cramer's V: 0.431, p50.001) and long term ( 2 (1,704):176.00, p50.001, weighted Kappa: 0.48, CI: 0.42-0.54, Cramer's V: 0.50, p50.001). In 18.3% of pairs (n: 126), patients opted for competitive employment whereas clinicians did not find this realistic. If the patient was willing to work in the short term, the odds of clinicians indicating competitive employment was 9.03 (OR) times higher than if the patient was not willing to work. The same trend was true in the long term (OR: 11.2) ( Table 3).

Discussion
While some previous articles have shown that clients of community-based services want to work, few studies focused on hospitalized patients with mental disorders. This study shows that many hospitalized patients still prefer competitive employment although many clinicians do not perceive this option as realistic. In addition, clinicians and patients often disagree on the level of barriers to employment and the appropriate ways to overcome them.
Although the interest of 45% of hospitalized patients for competitive employment in the long term seems to replicate some earlier research concerning people with SMI who were receiving community-based care (McQuilken et al., 2003), this interest is less than found by most of the preceding studies, which reported percentages up to 90% (Rogers et al., 1991;Van Audenhove & Wilmotte, 2004). The low interest in competitive work might be partially attributable to sociocultural differences between countries resulting in more negative attitudes and concerns towards competitive employment in Belgium. Belgian psychiatric hospitals only recently started to focus on VR and clinicians are not educated to initiate a VR process. Besides, there is great reliance on Sheltered Employment and volunteering (Shima et al., 2008) and these are often seen as the principal vocational alternatives to competitive employment (Marwaha et al., 2009). As confirmed by our data, clinical professionals hold the expectation that competitive employment is not realistic for most of their hospitalized patients. This partly parallels the study by Marwaha et al. (2009), which indicated that, although clinicians may overall believe that people with psychosis are capable of working, they also find many of their patients incapable of competitive work. As a result, clinicians may be less encouraging, explore employment goals less or refer to pre-vocational services instead of services that focus on competitive jobs (Marwaha & Johnson, 2005;Shima et al., 2008). According to the interpersonal expectancy effect, clinicians' expectations and behaviours influence those of patients. Patients might expect vocational failure which negatively affects their hope and intentions to pursue employment (Goscha et al., 2013;O'Connell & Stein, 2011). The attitudes and behaviour of both the patient and the clinician will result in a reduced competitive employment rate, which acts as a self-fulfilling prophecy, for it seems to confirm the clinicians' and patients' perceptions that competitive employment is unrealistic (O'Connell & Stein, 2011;Rinaldi et al., 2008).
Consistent with the finding that patients and clinicians differ on employment goals, the results show that there is little agreement on the extent of barriers to employment and the steps possible to overcome barriers. Clinicians perceive more barriers and indicate that more support will be needed to overcome them. Divergent perceptions of vocational goals, barriers and the interventions needed to find jobs indicate a lack of consensus and vague treatment plans. A lack of consensus is problematic, as it may lead to reduced effectiveness of treatment (Tryon & Winograd, 2011) and a poor therapeutic relationship with the clinician. Such a poor relationship can negatively affect the subsequent relationship with the VR counsellor. Yet, such a good relationship with the VR counsellor is essential for achieving high competitive employment rates (Catty et al., 2011). In contrast, when a patient and a clinician reach a consensus and the patient perceives that the clinician believes in his or her possibilities, better outcomes can be expected (Marwaha & Johnson, 2005;Roth & Crane-Ross, 2002). Hence, it is important for the clinician to create a positive relationship by offering hope, asking timely questions regarding the patients' vocational goals and seeking an initial consensus regarding the treatment plan (Drebing et al., 2004).

Limitations and future research
A limitation of the study is due to the method of recruiting participants. Self-selecting bias may be present as all hospitalized patients were offered a questionnaire but patients decided themselves whether or not they wished to participate. Non-response analyses showed that non-responders were older and had more often psychotic symptoms. We hypothesize that those not responding are also less interested in competitive jobs, so the overall population interest may be even lower than reported. Further research must focus on how the clinician's age, employment history or educational level affects patients' vocational goals and employment rates. This could be part of a longitudinal study in which the changes in perspectives and the level of agreement during hospitalization are investigated. Subsequent research will also need to focus on interventions to reduce incongruences between clinicians and patients but also between clinicians of different organizations as they can hold very different perspectives (Knaeps et al., 2015).

Implications
Many people with mental disorders have undetected needs and goals (Gibbons et al., 2005;Waghorn et al., 2014). Policy makers and management have to sensitize and train clinicians concerning their role in VR and the importance of employment in people's lives (Marwaha & Johnson, 2005). Clinicians can be stimulated to discuss vocational concerns early in treatment by offering training on motivational interviewing and shared-decision making concerning the type of job and type of support. In addition, patients need to be made aware that employment is important and that it can be addressed during treatment. Overall, contact between VR counsellors, patients and clinicians need to be stimulated.

Conclusions
To our knowledge, this study was the first to include both clinicians and patients when exploring vocational recovery. Contrary to patients, many clinicians do not consider competitive employment a realistic vocational option. Moreover, the opinions of patients and clinicians differ regarding vocational barriers and what is needed to overcome these barriers. Clinicians need to be aware of their patients' vocational needs and should make work of a positive collaborative engagement. In addition, it is important that Belgium further reduces the number of hospitalizations and increase the number of community-based services.