A chance to work represents an important goal for many people with mental disorders and it is one of the main elements protecting them from the possibility of developing, or worsening, problems of mental health, as well as it being a real possibility of experimenting an adequate level of community integration [1].

Innovative experiences demonstrated that offering employment reduces patients risk of new hospitalizations and of their becoming chronic, establishing positive dynamics that improve their social fitness and clinical recovery [2]. Nevertheless, the lack of employment represents one of the more relevant problems for people with mental disorders [3, 4]. It is estimated that about 85% of people with serious mental disorders are unemployed [5-7], and that the rate of their access to the labor market is approximately half the rate of people with other types of disability: physical, intellective, sensorial [8, 9].

Studies on this topic are cluster able into three main areas of interest: 1) studies that analyze which variables predict a return to work, following prolonged absence; 2) studies that assess the elements associated with absenteeism; 3) studies examining variables predicting job maintenance by people with mental illness.

Studies on return to work show that only 50% of people who are absent from work for more than six months return to their previous employment afterward [10]. The most predictive factors are the presence of good cognitive functions, good medical conditions, the presence of social support and skills, previous work experiences, and the presence of an established good therapeutic relationship, both with the key reference practitioner, and with the practitioner for work integration [11, 12].

Studies that have assessed absenteeism revealed that being a male, having an early onset of mental disorder [13], receiving little support from colleagues, and being in a general condition of reduced social support [14] are elements associated to greater absence from work. In addition, a heavy workload, risky behaviors, and the existence of some specific disorders, such as major depression, dysthymia, phobia, and use of street drugs are associated to repeated and prolonged absences from work [15]. Moreover, these studies have highlighted that returning to work after six months was predictive of a subsequent decline of depressive symptoms.

Studies that have assessed job maintenance by people with mental illness have shown that about 70% of those participating to a program such as Individual Placement and Support (IPS) keep their work longer over time [16, 17]. The IPS model was developed to provide supported employment services to people with severe mental disorders. IPS used a team approach to integrate mental health and vocational services. Mental health centers hired employment specialists who, in order to ensure coordinated services, directly joined the participating clinical teams. Rather than providing pre-employment assessment and training for job-related activities, IPS employment specialists began helping their patients to find a job immediately and, after having secured their employment, they trained them and followed-along as needed. IPS employment specialists assumed that their patients would have learnt job skills and made-up their minds about their work preferences, whilst on the job, rather than through pre-employment counselling and training [18].

A longitudinal study, based on an IPS that studied a cohort of people for ten years, demonstrated that most people were still working at follow-up and that about one third of them had been on the job for at least five years, whenever working part-time jobs, and that they were competitively paid [19]. A further study, which assessed the state of occupation after 8-12 years, showed that all the subjects involved were still working at follow-up time: 82% were in a competitive job, whilst 71% of them had worked for more than half of the follow-up years taken into consideration [20]. The most predictive variables were symptoms remission, having had previous work experiences and experimenting a good level of work satisfaction. Another study, carried out on patients with psychotic disorders, showed that having a specialist assigned to the patient work place predicts maintenance time in a competitive job at the 24 months follow-up [21].

All studies reported above show that the presence of previous work experience and higher cognitive levels of memory, attention, and executive functions are associated with a greater number of working hours and higher payoff. On the contrary, cognitive impairment is associated with the need of greater support at the workplace and with the request for more contacts with the mental health care providers.

Other studies, carried out on patients with psychosis, show that coping skills and social support are more important for the management of their illness than for facing any other problem at their workplace. Conversely, the latter mostly benefit by supporting company staff and by colleagues sympathetic to mental health problems of their peers [22].

In another study, which investigated the working reality of people with schizophrenia in UK, France and Germany it emerged that the probability of having and keeping a job is associated with factors such as: younger age, higher education, reduced symptoms prevalence, and family life. Conversely, the probability of being employed is hampered by disorder chronicity, early onset, substance abuse [7], and receiving supplementary incomes, linked, or not, to any disability [23, 24].

Beyond the above-mentioned ones, there are further causes, not always due to mental health problems, or any other characteristics of the work environment, which limit occupational projects and which are. The Italian condition is likely mainly characterized by insufficient integration among care programs, mental health rehabilitation services and the employment ones. On stage, there is also: a poor support, by psychiatric services, toward individual work integration, likely due to the absence of a specific culture in this area; the lack of specific resources for work-related projects; the lack of systematic assessment procedures of the existing vocational integration programs, which precludes the possibility of meaningful comparison in scientific studies; the poor competence of the psychiatric services in the work rehabilitation fields and the workforce shortage of the functional and multidisciplinary teams specialized in this field.

All the above address the problem of disciplinary integration and the consideration that rehabilitative approaches and their response patterns at work are neither univocal, nor definitive present themselves today [25]. Their heterogeneity influences the clinical practice, which in Italy is not always clearly defined. Data are lacking on both the outcomes of the vocational placement services and on the professional and organizational models at work in this field of Italian society.

This is a retrospective study, conducted at vocational integration service of a town in northern Italy; it had two major aims: a) to assess programs for vocational integration active from 1^st January 2004 to 1^st January 2007; b) to identify job tenure-associated predictors. We hope that the results of this study could help implementing new strategies and promote a reflection on the development of good standard procedures to improve joint ventures on vocational integration between mental health services and the public and private/social employment-dedicated activities.

This study dealt with the daily practice of vocational placement service with the aim of identifying any variables predictive of its outcomes, whose main one is employment maintenance. The research shows that about 65% of the subjects involved in a supported work program kept their employment over time. This study also stresses the validity of using supported employment as an effective strategy to help people with mental disorders to obtain and keep competitive jobs.

At an international level, it has now been agreed since some time, that people receiving a program, like the IPS, perform on the job more effectively even in competitive employment, and do it longer, than those who undergo traditional professional rehabilitative approaches. Moreover, it would seem that receiving IPS could lower the probability of suffering more hospitalizations and could even lower the risk of dropping out from work [27].

This present study shows that the total number of interventions, received before job placement, is the only variable somehow associated with the program outcome. This result indicates the tendency of job maintenance to be, albeit weakly, negatively linked with the number of interventions received after the beginning of the program, but before the users’ placement in a company, or in social cooperative. This result could be at odd with the results by Leff and collaborators who, in their study [26], found that the amount of job support received was significantly associated with the number of months that their patients spent in their first competitive job. However, these authors did not consider whether job support was given, or not, before users’ job placement, opening the possibility that their patients had received more job support just because they had spent more time on the job. Of course, this would invalidate the hypothesis that job support could have been a causal factor associated with job maintenance. In this study the causal role between job support and job maintenance is, therefore, questionable.

A negative relationship between prior support and job maintenance was found by John and collaborators. In their study they proved that two different categories of financial support (social security disability income and supplemental security income), before the beginning of supported employment experience, were negatively associated with future work status [28]. Other studies have explored the impact of support on job maintenance, in the IPS model, but the results are still controversial. In fact, little direct evidence exists supporting the hypothesis that ongoing support contributes to job maintenance, even though it is one of the core tenets of the IPS model. In a recent study Bond and Kukla [29] found a moderate correlation between intensity of employment specialist support and duration of employment, but an important question not answered by their study is whether other sources of support, such as mental health case managers and natural supports, such as family members, can substitute for or augment the assistance of employment specialists. In another study, McGuire and collaborators [30] found that the total IPS contacts were positively associated with weeks worked and that the client background characteristics did not predict service intensity. However, one relationship indicated that participants who were better performers had received more services. Conversely, other studies had found that having more severe symptoms was associated with more intense support.

Also in our study, the results could be attributed to the number of interventions received before entering the work environment, as it could be a proxy of seriousness. When users show higher levels of psychopathology, – or lower levels of psychosocial functioning, and/or a worse background of work experience – they need greater ongoing support by practitioners and that could hinder the attainment of a positive outcome.

Anyway, the above reported studies show that further research on job support is needed, to improve our understanding of variables associated with employment outcomes.

Finally, we have to acknowledge that the possibility of deriving conclusions from this study is hindered by a number of limitations, as follows. Lack of information on subjects previous work experiences, missing the mean follow-up period of job maintenance, lacking information about subjects’ psychopathology and, last, but not least, lacking information on personal and social functioning and cognitive profile of tested users. The lack of these data makes it, therefore, impossible to estimate what effects the above-mentioned variables would have had on, both, the choice of entering a specific vocational placement program, and on job maintenance. We cannot be sure of which factors really led to the premature loss of their employment by those people who had a negative outcome. There are many ways that disabilities can affect the capacity to perform effectively on the job. It might simply be that not all people with severe mental disorders can be led to obtain and maintain a job.

All these shortages undoubtedly represent critical elements of this study, but they were unavoidable because they are the consequence of the outlook of Italian services dealing with the mental health patients and/or with their work integration. In Italy, mental health practitioners working in this field do not usually assess, through standardized instruments, the severity of symptoms, the level of disability and the neuropsychological functioning of their patients. In addition, has we were dealing with a retrospective study; we were able to assess just the data that had been collected in a routine service.

We considered satisfaction important and chose to investigated as the dependent variable in our study, as users are becoming more and more active whenever it comes to take decisions about their future and their care. Therefore, they have greater influence, than they had in the past, to change and improve the quality of services by reporting the satisfaction of their experiences. Data show that the area where, both, users as well as their caregivers and company contacts too, express a greater level of satisfaction, is the support received by the practitioners of the WIU. It is possible that this result had been influenced by a social desirability bias, but such satisfaction could also depend on the good support supplied by service practitioners and be the main variable positively affecting the relative effectiveness of the program [31]. However, no other variables, which could be associated with a lower, or greater, level of satisfaction, and to the main outcome as well, were collected. Candidate variables would be, e.g., salary, the number of working hours, the type of work, etc.

Finally, a few words must be spent on the satisfaction by the practitioners of the WIU, which scored the lowest in the area of the collaboration received by the dispatching mental health services. This result opens an important reflection on the need to increase the collaboration between the different services that, at various levels, take care of mental disorders. No matter how beneficial the effect that job status and work experiences play on the progress of mental illnesses, insufficient interest still seems to persist toward the role that work rehabilitation could play in the treatment of people with mental disorders. It still seems too rare, for mental health practitioners, to consider unemployment a significant stressor for people under their care; but this is a factor that can influence, not just clinical recovery, but also the probability of relapse. Working can indeed be a decisive element for the development of an adequate feeling of self-esteem and for the attainment of a minimum amount of social status by people suffering from mental disorders. It is, therefore, necessary to promote and implement a system of care foreseeing an effective integration of all the services, which deal with mental health, realizing an overall total charge of the person with psychiatric disabilities. Only through a psychosocial rehabilitation thus conceived it would be possible to boost the real social reintegration of people suffering from these disorders.

The authors declare they have not conflict of interest regarding this paper. This work was supported by the Italian Institute of Health (ex article 56, Law 289/2002, N° 534M/7).