Awareness of the importance of employment for people who use mental health services has been raised significantly through the recent Social Exclusion Unit report (Social Exclusion Unit, 2004), and the role of health and social services has been described as critical. The key recommendation is that vocational and social support is embedded into the care programme approach (CPA). However, implementation is likely to be complex and challenging, as different parts of the country will be at different stages in their thinking and practice regarding their level of integration. Mental health services have not routinely collected data on how many service users are employed or engaged in occupational options or the extent of occupational care planning within the CPA. We therefore aimed to identify the employment rates of people using local mental health services and to examine any evidence in care plans for vocational or occupational interventions.

Method

Employees from the Clinical Governance Department at the South London and Maudsley Trust obtained all available case notes from several clinical areas (see below) and examined them for the following variables: sociodemographic (gender, ethnicity, age); diagnosis; evidence of any documentation relating to vocational needs in the case notes and more specifically in CPA minutes; employment status; whether patients were engaged in or had been referred for voluntary work, vocational work schemes or projects, education or training courses; any evidence of one-to-one vocational work with primary worker or care coordinator; vocational interventions/discussions (including vocational interests, such as training, education or seeking employment), assessments, referrals; documented reasons if the service user was unoccupied and other related comments.

Data were analysed with Stata version 8 for Windows. Simple descriptive statistics were used; χ² tests were used to compare differences in proportions, and Kruskal-Wallis tests were used to test differences in continuous measures as these were not normally distributed.

Results

A total of 297 case notes were obtained between December 2003 and June 2004. These were from rehabilitation services (n=11), 5 community mental health teams (CMHTs) (n=150), 2 acute wards (n=41), and an out-patients department (n=94). There were 160 men and 137 women who had had a median of 5 years in contact with services (range 0.16-42 years). Their mean age was 42.5 years. There were 125 White patients, 100 Black patients, 14 Asian or Asian British patients, 4 patients of mixed background, and 10 other (with other patients having no descriptions given). There were 136 patients with a clinical diagnosis of schizophrenia, 29 with bipolar disorder, 5 with schizoaffective disorder, 5 with drug-induced psychoses, 2 with delusional disorders, 1 with paranoid psychosis, and 1 with a psychosis not otherwise specified. Of the patients, 28 had a diagnosis of depression, 4 patients had post-traumatic stress disorder, 1 patient had a dissociative amnesia, 3 patients had alcohol misuse, 2 patients had drug misuse, 1 patient had Korsakoff's syndrome and 1 patient had dementia. Diagnoses were not clear from the case notes for the other 78 patients.

Overall, 88% of the patients were unemployed (77% of out-patients, 78% of patients on acute wards and 96% under the CMHTs). There was evidence of documentation relating to vocational needs in the case notes in 10 out of the 11 patients (91%) in the rehabilitation services. However, for out-patients this occurred for only 18% (17 of 94), for 8% of those under CMHTs (12 of 150) and 39% in acute wards (16 of 41).

Table 1 shows the number of patients in work or training schemes according to the service. Overall, 8% of patients were engaged in a range of work schemes, 10% were engaged in educational options, and 9% were engaged in one-to-one work with their primary worker/care coordinator. The evidence indicates that across most service areas very little or in some instances no one-to-one vocational work was being provided by care coordinators/primary workers. Of note, for the case notes available, one CMHT engaged over half its patients (15 of 29) in vocational interventions. The outcomes for this team were also much higher than any other service area: 4 (14%) were in employment, 1 (4%) was in voluntary work, 5 (17%) in work schemes and 4 (14%) in education or training.

Table 1. Number of patients engaged in work schemes or training

Service	Patients in each service, n	Voluntary work	Work scheme	Education/training	Vocational work with primary worker/care coordinator
Rehabilitation	11	2	3	1	4
CMHTs	150	5	12	14	19
Acute wards	41	5	4	3	2
Out-patients	94	4	4	10	0

Any documented reasons why service users were not engaged in occupational options were recorded. The most commonly noted reasons given were drugs (n=21) and alcohol (n=9). Asylum-related problems (n=11) were also frequent, as were problems with child care or parenting (n=9). Physical problems were also cited for 8 patients.

Employment (part-time or full-time) was less likely in patients from Black and minority ethnic groups (OR=0.41, 95% CI 0.17-0.99, P=0.05), and there was a trend for these patients to be less likely to receive specific one-to-one vocational work with a primary worker or care coordinator (OR=0.44, 95% CI 0.18-1.05, P=0.07). We combined employment with engagement in or referral for voluntary work, work schemes, education or training, and specific one-to-one vocational work with primary worker/care coordinator as a ‘ good employment outcome’. This outcome was not more likely in women or men (P=0.5), and did not differ significantly with age (P=0.02) or ethnicity (P=0.11), but the difference did approach significance for years in contact with services (P=0.06), with patients in contact with services for several years having the least chance of a good employment outcome.

In some case notes, staff elaborated more on individual reasons why service users were not engaged in occupational activities. These included engagement issues: ‘difficult to engage’, ‘won't accept mental health problems’, ‘no motivation - depressed’, ‘poor personal hygiene’, and individual functional issues (e.g. ‘poor memory’, ‘cannot concentrate’, ‘decline in physical health’).

A significant lack of work history and a lack of individual interest in working emerged as a consistent theme, although the individual reasons were not described in any depth from a service user perspective (e.g. ‘has never worked, does not want to’, ‘never been employed’, ‘ attempted to get on a college course/education - failed’. The available options were also identified as affecting engagement: ‘feels there is nothing to do’, ‘resisting day care’ and ‘ does not want to attend rehab’). Some staff had strong opinions about whether service users could work at all, for various reasons: ‘not well enough to maintain full/part-time work’, ‘lacks insight, tried training’ and ‘finds work stressful’.

A wide range of family-related ties were cited as barriers: ‘husband supports wife, does not want her to work’, ‘lives with parents, no motivation’ and ‘looks after grandchildren, too busy to work’. An array of welfare benefits problems emerged as obstacles to engagement in work: ‘in the welfare trap, ‘happy to live on benefits’ and ‘unable to work because of benefits’.

Discussion

This study confirms previous findings that unemployment rates are very high in patients with psychiatric disorders. The national figure for unemployment among people with long-term mental health problems is 76% (Social Exclusion Unit, 2004). Unemployment rates among those with a diagnosis of schizophrenia may be increasing: in Wandsworth and Merton rates increased from 88% in 1992 to 96% in 1999 (Reference Perkins and RinaldiPerkins & Rinaldi, 2002).

Despite these high rates of unemployment, we found very low rates of occupational care planning, one-to-one vocational work and referrals to work projects documented in clinical case notes. Further research would be required to establish why. There was very little evidence from the case note data of occupational care planning that attempted to address the individual's occupational needs in a structured engagement process. The data show this area of clinical work as problematic and consistent with national literature, in that staff tend to relegate or ignore occupational needs, possibly because they themselves have not received much training in this area.

There are several limitations to this study. First, unfortunately, case notes were not randomly selected. These case notes may not therefore be representative of the patients under the care of the trust in south London. Second, several categories have not been operationalised or standardised (e.g. clinical diagnosis only could be obtained from the case notes). The information was not checked, so the quality of the data has not been verified. However, the Clinical Governance Department is experienced at obtaining information from case notes, and so the quality of the information from a large sample such as this means that the data are worth studying.

The high level of unemployment and the correspondingly high level of social exclusion among people using mental health services in south London are consistent with national research findings across England. The extent of one-to-one vocational work with primary workers is low or non-existent in some areas of service provision. However, our evidence indicates that when it happens social inclusion outcomes are much better. Standard 5 of the National Service Framework for mental health set a target that by March 2002 all written care plans for people on enhanced CPA must show plans to secure suitable employment or other occupational activity. This study indicates that this is still not happening and referrals to work projects are also very low.

The Social Exclusion Unit guidance, to be published shortly by the National Institute for Mental Health in England, will require mental health trusts to negotiate vocational outcomes with commissioners and help modernise vocational services generally. Staff will require training and specialist input from dedicated professionals (occupational therapists and vocational specialists), supported by effective pathways and local vocational structures in line with good practice guidelines.

Models of vocational interventions need to be developed and researched for this group of patients, as increasing their rates of employment will decrease social exclusion and stigma. We also need to establish and clarify the evidence on which group of professionals has the optimum skills for helping a patient with severe mental illness into work. An occupational therapist has the core skills for rehabilitation of a patient with severe mental illness but not necessarily the skills to support a patient back into work, and an employment specialist has the skills for vocational assessment and job matching but not neccesarily the holistic range that may help a patient diagnosed with severe mental illness.

We are currently investigating, with Professor Graham Thornicroft (principal investigator), the effectiveness of supported employment in a randomised controlled trial, the Supported Work and Needs (SWAN) trial, funded by the Wellcome Trust. The traditional model of vocational rehabilitation uses a ‘train and place’ approach, offering training in sheltered workshops and then placing individuals in real-life work settings. The SWAN study uses the IPS model (individual placement and support), and is the reverse, that is, ‘place and train’, so that clients are placed in real jobs and then offered variable amounts of direct personal support by an experienced employment consultant to enable them to retain their work positions. Studies of IPS programmes in North America have been encouraging in terms of increased rates of competitive employment (Drake et al, Reference Drake, Becker and Biesanz1996, Reference Drake, McHugo and Bebout1999; Reference Bond, Drake and MueserBond et al, 1997; Reference Marshall, Crowther and Almaraz-SerranoMarshall et al, 2001; Reference Lehman, Goldberg and DixonLehman et al, 2002).

An alternative model involves initial engagement in voluntary work and education. ‘Although these activities may yield benefits in and of themselves, it is often asserted that engagement in such activities can prepare people for entry/re-entry into employment. The continuing downward trend in employment rates indicates that this has not been the case’ (Reference Perkins and RinaldiPerkins & Rinaldi, 2002). Without comparative longitudinal data we cannot make the same assertion.

Conclusion

This project has demonstrated low rates of vocational interventions recorded in case notes despite the high rate of unemployment in patients with long term mental health problems. We hope that growing awareness of vocational possibilities for patients with severe mental illness will mean that in the future, with skilled help, many patients will be able to return to competitive employment.

Declaration of interest

None.