Dr. Cougnard, Ms. Monello, and Dr. Verdoux are affiliated with Institut National de la Santé et de la Recherche Médicale (INSERM) U657, University Victor Segalen Bordeaux 2, Bordeaux, France. Dr. Goumilloux is with the Maison Départementale des Personnes Handicapées de la Gironde Bordeaux, France. Send correspondence to Dr. Verdoux at INSERM U657, University Victor Segalen Bordeaux 2, 146 rue Léo Saignat, Bordeaux 33076, France (e-mail: helene.verdoux@u-bordeaux2.fr).

Helping a person with a psychiatric disorder to keep or find a job is a major therapeutic goal that is generally shared by the patient, her or his family members, and mental health professionals. Employment is socially valued and may be associated with better quality of life, social inclusion, and better outcomes (1,2). However, this goal is rarely reached by persons with schizophrenia. Recent studies carried out in developed countries showed that employment rates of these persons range from 10% to 20% (1,3,4,5,6,7). In the Clinical Antipsychotic Trials of Intervention Effectiveness study, which was carried out in the United States, only 14.5% of persons with schizophrenia had competitive employment and earned income in the previous month (8). Labor market characteristics probably have a limited impact on occupational status, because the same percentage was found in a study carried out in three European countries (3). Schizophrenia is hence a leading cause of disability worldwide (9), with an impact particularly marked in the age group of 15 to 44 years (10).

Thus individuals with schizophrenia, their family members, and even health professionals frequently have to go through a mourning process and admit the disabling impact of the psychotic illness on working activity. In the disease history of a person with schizophrenia, the first contact with mental health services is hence often followed by a key second step—that is, the request for a disability benefit, at least in countries where disabled persons may receive such a benefit. The criteria for allocating social benefits for people who are unable to work as a result of disability are generally the same in developed countries (11). Contrasting with the large amount of research on duration of untreated psychosis and pathways to care of persons with early psychosis (1,12,13,14), little is known about the duration of the period between onset of illness and first request for social benefits. Better knowledge is required of this period over which persons with schizophrenia have to cope not only with the symptoms and treatment of an incipient disease but also with its social and financial consequences. Indeed, a large proportion of persons with schizophrenia are unemployed at the onset of illness, and this proportion dramatically increases in the following years (1,15).

The aims of the study were to examine a sample of French persons with schizophrenia or schizoaffective disorder requesting disability status for the first time to explore the occupational and income history between onset of schizophrenia and first request, the duration of this period, and the demographic and clinical characteristics associated with a long duration.

Setting

This study was carried out in collaboration with the Commission Technique d'Orientation et de Reclassement Professionnel (COTOREP) (Technical Commission for Occupational Guidance and Rehabilitation of the Disabled). COTOREP is a state agency implemented in each French département (subdivision of Metropolitan France). It allocates noncontributory social benefits to people who are unable to work as a result of disability. To obtain this benefit, called Allocation Adulte Handicapé (AAH) (Allowance for Disabled Adults), the person must be aged between 20 and 60 years, live in France, and be a European Union citizen or foreigner with legal authorization to stay in France. The monthly amount of the AAH in 2007 was 620 euros (U.S. $860) for persons with no income; in addition, persons living alone received an additional 100 euros (U.S. $139). COTOREP also assesses the working ability of disabled persons and can attribute the status of disabled worker (Reconnaissance de la qualité de Travailleur Handicapé) (RTH) (qualification for the status of disabled worker), allowing access to employment (sheltered or not) for persons with reduced ability to work. Persons qualified for RTH may still be able to get full AAH funds until they receive a salary.

The study presented here was carried out at the COTOREP in Gironde, a département of 1,287,334 inhabitants located in South-Western France, including the city of Bordeaux. The person applying for disability status is assessed by a COTOREP practitioner who estimates the degree of disability. The study was conducted in the COTOREP of the Gironde because in this area patients presenting with psychiatric disorders as the main diagnosis prompting the request for disability benefit are assessed by a psychiatrist, whereas in most other COTOREPs, the assessment is made by a general practitioner.

Participants

All persons consecutively assessed by the COTOREP psychiatrist were included over a ten-month period (January 2006 through October 2006) if they fulfilled the following criteria: they provided written informed consent to participate in the survey after the procedure had been fully explained, it was their first request for disability status (AAH or RTH), they were older than 20 years (minimum age for a disability status request), they presented with a diagnosis of schizophrenia or schizoaffective disorder according to DSM-IV criteria (16).

Written informed consent was obtained after the study was fully described to the participants. The study conformed to the French bioethics and clinical research and data protection legislation. No advice from the bioethics committee is required in France for this kind of study, because there was no supplementary medical procedure mandated by the study and there was no biomedical intervention. All data were rendered anonymous before analysis.

Assessment

Information was collected with a standardized structured questionnaire designed for the study presented here. The first section of the questionnaire was completed by the COTOREP psychiatrist during the medical assessment and collected information on demographic characteristics, clinical diagnosis according to DSM-IV criteria, type of request (AAH, RTH, or both), and decision regarding the level of disability.

The second section of the questionnaire was completed by a psychologist blind to the data collected by the COTOREP psychiatrist and collected information on demographic and family characteristics (number of siblings dependent on parents and occupational status of parents). Age at onset of psychiatric illness, defined as age at first psychiatric hospitalization or age at first psychiatric consultation if there was no history of hospitalization, was used to calculate the delay between age at onset and first request. Diagnoses of drug and alcohol use disorders over the prior 12 months were based on the corresponding sections of the Mini International Neuropsychiatric Interview (MINI) (17).

History of employment was established by collecting detailed information on all periods of work and the duration of employment before and after the onset of illness. We used a broad definition of work, including moonlighting and episodic activities, such as babysitting or grape harvesting. The variable "employed at least once" was defined as having worked at least once over the person's lifetime according to the previous definition, irrespective of the duration. The percentage of time spent employed before the onset of illness was calculated over the period between the end of educational pursuits and the onset of illness. Persons who were students at the time of the onset of illness were rated as having been fully employed over this period, and housewives were rated as unemployed. The percentage of time spent employed after the onset of illness was calculated with the same criteria. Information was collected on the type and duration of income received since the onset of illness.

Statistical analyses

Statistical analyses were performed using STATA 9 (18). The duration of the period between onset of illness and first disability status request was categorized as long or short, according to the median duration. Age was a priori categorized into three classes: 20—25 years, 26—35 years, and 36 years and older. The lower age limit was chosen because AAH is the only social benefit for persons who are unable to work and are between the ages of 20 and 25 years, because the minimum income for inclusion in society (Revenu Minimum d'Insertion) (RMI), given to persons with no income who are considered as being able to work, cannot be received before this age. The upper age limit was chosen on the basis of a prior survey carried out at the Gironde COTOREP showing that persons with schizophrenia aged 25—35 years at first disability request are relatively homogeneous regarding clinical and work history (19).

Univariate logistic regressions giving odds ratios (ORs) and 95% confidence intervals (CIs) were used to explore the characteristics associated with a long duration between onset of illness and first disability status request. All variables associated with a longer duration at the p<.1 level in univariate analyses were entered into a multivariate logistic regression model. Age and gender were a priori forced into the model.

Demographic and clinical characteristics of the sample

Of the 116 persons eligible for the study, six were excluded. Two of these persons were mute, one person did not complete the second assessment, and information about onset of illness was missing for three persons. Therefore, the sample included 110 persons with a median age of 29 years (interquartile range [IQR] of 25—37 years). Their demographic characteristics are described in Table 1. Three-quarters of persons (N=84, 76%) had a diagnosis of schizophrenia, and the others (N=26, 24%) had a diagnosis of schizoaffective disorder. Most had a history of psychiatric hospitalization (N=98, 89%). The median was used to categorize the number of hospitalizations as "no or one hospitalization" or as "more than one hospitalization." The median age at first hospitalization was 23 years (IQR of 20—28 years). The 12 persons (11%) without a history of psychiatric hospitalization had a median age of 32 years (IQR of 22—37 years) at first psychiatric consultation. One of four persons presented with a MINI diagnosis of drug or alcohol use over the prior 12 months.

Occupational and income history

The majority of persons worked at least once during their lifetime, but this proportion was reduced to less than half after the onset of illness (Table 1). The median percentages of time spent employed were 64% before the onset of illness (IQR of 29%—100%) and 0% after the onset of illness (IQR of 0%—44%) (data missing for eight persons). Percentages of time spent employed were categorized as long or short, according to the medians. More than one out of three persons received the minimum income for inclusion in society (RMI) at some point between the onset of illness and the first disability status request, and this proportion was one-quarter at the time of the first request. The RMI was 440 euros (U.S. $611) in 2007 for a single person with no children—that is, at least 180 euros (U.S. $250) less than the money an eligible person could receive from AAH. About two-thirds of the sample received financial support from the family between onset of illness and first disability status request, and for 41%, family was the only source of income.

Characteristics of disability status request

AAH was requested by a majority of persons (N=103, 94%), of whom 98% (N=101) obtained it after psychiatric assessment. One of three persons requested the RTH (N=41, 37%), of whom 46% (N=19) obtained it. This low percentage is explained by the fact that this status is given only to persons who are able to work, not to those who would like to work. The median duration of the period between onset of illness and first request was four years (IQR of two to eight years; range of 0 to 32 years).

Characteristics associated with duration

Compared with persons with a short duration (less than four years) between onset of illness and first disability status request, persons with a long duration (four or more years) were older at the time of first request, had a higher educational level, had more psychiatric hospitalizations, and had worked over a longer period before (at trend level) and after the onset of illness (Table 2). Multivariate analyses showed that four variables independently predicted a longer time between onset and first request: older age (older than 36 years compared with 20—25 years; adjusted OR=6.6, CI=1.5—28.9, p=.01), higher educational level (adjusted OR=3.9, CI=1.1—14.8, p=.04), more than one psychiatric hospitalization (adjusted OR=11.2, CI=3.2—39.1, p<.001), and longer percentage of time spent employed after the onset of illness (adjusted OR=3.9, CI=1.2—12.2, p=.02).

This study shows that nearly half of the persons with schizophrenia or schizoaffective disorder requesting disability status for the first time did not receive any income other than financial support from their families after the onset of illness. The median delay between illness onset and first disability status request was four years. The characteristics independently predicting a long period between illness onset and first request were older age, higher educational level, a longer period of working activity after the onset of illness, and more than one psychiatric hospitalization.

Methodological limitations

Regarding the representativeness of our sample, the occupational and income history of persons assessed in the study presented here cannot be extrapolated to all persons presenting with schizophrenia or schizoaffective disorder, because disability status is requested by persons with reduced working capacity and without any other source of income. In particular, this selection may have overestimated the frequency of unemployment and low income. It may be estimated that around 71% of French persons with schizophrenia have an AAH allowance and that 45% of persons with schizophrenia who do not have this allowance are working (20).

Information was based on self-report; thus memory bias may have affected the data collected on duration of employment or income. Because persons were interviewed when they attended the medical appointment assessing their ability to work, they may have underreported their actual income or working activity. However, they were also interviewed after the medical appointment and were informed that their answers would not influence the decision regarding their disability status.

Age at first admission or first consultation was used as a proxy for age at onset of illness. Considering that working capacity is often reduced in the prodromal and premorbid periods, this may have contributed to an underestimation of the duration of employment before the onset of illness.

Because our sample was relatively small, our study may be limited by lack of statistical power, which could explain, for example, why the association between the longer duration before the request and age was not significant among persons older than 35 years, despite an effect size in that group that was comparable to that found in the age group of 20—25 years.

Interpretation of findings

Almost all persons with psychotic disorder requesting disability status had worked at least briefly at some point in their lives. The high proportion in this study is partly explained by the broad definition of working activity, including student status and all kinds of paid activity, irrespective of its duration. In accordance with prior studies (1,15,21), the proportion of employed persons was considerably reduced after the onset of illness. As a consequence, a large proportion of persons received only limited resources after the onset of illness. The proportion of persons with family support as the only source of income between onset of illness and first disability status request was strikingly high.

The delay between the first contact with mental health professionals and first request for disability allowance was longer than two years for three of four persons and longer than four years for half. A population-based study carried out in Denmark also showed that the probability of receiving a disability pension remained higher among persons with schizophrenia than among those without schizophrenia up to 25 years after the onset of illness (15).

The longer delay between onset of illness and first disability status request among older persons is probably explained by the fact that the disability allowance is the only social benefit in France available for persons between the ages of 20 and 25 years, because RMI is available only after the age of 25. Persons with lower income have been found to be more likely to request disability benefits (22). The same mechanism may explain the association between being employed after illness onset and the longer duration before request. An association between lower educational level and increased probability of receiving disability benefits has been reported in previous studies (8,22). The association between higher educational level and increased duration before requesting disability status found in the study presented here was independent of age and occupational status after illness onset; hence it was not explained by the fact that persons with more educational are able to continue to work for a longer period after illness onset. This may be partly explained by the fact that compared with persons without qualifications, persons with psychosis who have graduated from school or university may remain hopeful for a longer period of finding a job that fits their qualifications.

The longer duration between onset of illness and first disability status request among persons with more hospitalizations is more difficult to explain. An association in the opposite direction was expected, because a high number of hospitalizations is an indicator of illness severity and lower rates of unemployment have been reported in persons with no readmissions (15). It may be hypothesized that persons in treatment are not in a position to press a claim for assistance while they are in a hospital or being assisted in a treatment setting, or it may be that social issues are not considered as a priority when the clinical state is not stabilized. Persons with more admissions are also more likely to present with lower adherence to treatment (23) and are less likely to follow the advice of mental health professionals. This finding may also reflect the partial dissociation between clinical and social outcome among persons with psychosis.

Other factors not explored in the study presented here may explain the lengthy duration of the period between illness onset and request for disability benefit. Mental health professionals and social workers of disability agencies may be reluctant to increase the stigmatization linked to the diagnosis of schizophrenia by adding the label of "disabled person" or because giving access to disability benefits may act as a disincentive regarding the return or access to work, even though there is little evidence to support such a concern (8,24,25). Irrespective of the underlying mechanisms, a long time between illness onset and first request for disability status may exacerbate the social consequences of the disease resulting from lack of resources and increase the high prevalence of poverty among persons with severe mental illness (26).

Because of the long duration between illness onset and first request for disability status, many persons with schizophrenia have no income other than financial support from the family in the years after the onset of illness. It would be of interest to explore further the implicit and explicit criteria used by mental health professionals and social workers to decide whether a person with schizophrenia should request disability benefits.

This study was supported by a grant (convention MiRe n° 05/127) from the program Handicap Psychique et Troubles Psychiatriques (Psychic Handicap and Psychiatric Disorders) of the Mission Recherche, French Ministry of Health. The authors thank Ray Cooke, Ph.D., for his help with the English in this article.

The authors report no competing interests.